BANK OF CANADA Financial...Reducing financial inefficiencies can, in principle, lead to a better...

B A N K O F C A N A D A

FinancialSystem Review

December 2004

Members of the Editorial Committee

David Longworth, Chair

Agathe CôtéAllan Crawford

Pierre GodinClyde GoodletDonna Howard

Kim McPhailPhilippe Muller

John MurrayGeorge Pickering

James PowellChristopher Ragan

Denis SchutheBonnie Schwab

Jack SelodyRobert Turnbull

Mark Zelmer

Eddy CavéJill Moxley

Lea-Anne Solomonian(Editors)

The significant contribution of individual authors of specific portions of the Developments andTrends section, as well as that of members of the working group mandated with the preparation andorganization of the Review, is gratefully acknowledged.

The Bank of Canada’s Financial System Review is published semi-annually. Copies maybe obtained free of charge by contacting

Publications Distribution, Communications Department, Bank of Canada, Ottawa,Ontario, Canada K1A 0G9Telephone: (613) 782-8248; e-mail: [email protected]

Please forward any comments on the Financial System Review to

Public Information, Communications Department, Bank of Canada, Ottawa,Ontario, Canada K1A 0G9Telephone: (613) 782-8111, 1-800-303-1282; e-mail: [email protected]

Website: <http://www.bankofcanada.ca>

Bank of CanadaDecember 2004

Contents

Developments and Trends ........................................................... 1

Overview......................................................................................................... 3

Highlighted Issues ............................................................................................ 5

The Macrofinancial Environment ...................................................................... 17

The Financial System ....................................................................................... 23

Reports ..................................................................................... 33

Introduction .................................................................................................... 35

A Taxonomy of Market Efficiency ..................................................................... 37

Portrait of the Canadian Hedge Fund Industry .................................................. 41

Policy and Infrastructure Developments ........................................ 45

Introduction .................................................................................................... 47

Bank of Canada Lender-of-Last-Resort Policies ................................................... 49

Research Summaries .................................................................. 57

Introduction .................................................................................................... 59

Basel II and Required Bank Capital .................................................................. 61

Pre-Bid Run-Ups Ahead of Canadian Takeovers: How Big Is the Problem? ........... 67

Monetary Policy, Private Information, and International Stock Markets ............... 71

Developments

and

Trends

Notes

The material in this document is based on information available to 26 November un-less otherwise indicated.

The phrase “major banks” in Canada refers to the six largest Canadian commercialbanks by asset size: the Bank of Montreal, CIBC, National Bank, RBC Financial Group,Scotiabank, and TD Bank Financial Group.

Promoting Financial System Efficiency andStability

The Canadian financial system comprises financial markets, financial institutions,and the clearing and settlement systems. A well-functioning financial system(i) acquires and uses information to allocate resources to the most productiveinvestment projects and (ii) manages and distributes risk to those most willing tobear it. The financial system adds to social welfare and economic growth becauseit improves the allocation of resources and reduces the volatility of consumptionand investment. A well-functioning financial system is also able to better absorbadverse shocks, making the real economy less sensitive to them. As a consequence,economic growth is less volatile.

Frictions in the financial system can cause inefficiencies, which impair the effi-cient allocation of resources and make the economy more sensitive to adverse dis-turbances (i.e., more unstable) with significant welfare consequences (Haldane etal. 2004). Financial inefficiencies can arise for numerous reasons. (See Bauer 2004for definitions of market efficiency.)

Informational asymmetries in both financial markets and institutions can devel-op because borrowers typically have more information than lenders about the po-tential value and risk associated with the investment projects for which they areseeking funds. These asymmetries are exacerbated by factors such as poor qualityof financial information and poor corporate governance. Transactional inefficien-cies, which increase the costs of financial transactions, can occur because of lackof competition in the provision of financial services, regulatory requirement, orpoor legal infrastructure (e.g., weak enforcement of rules, regulations, and con-tracts governing the financial system). Cases like Enron expose extreme instancesof how these types of inefficiencies can play out in the real world. There is empir-ical evidence that these frictions are important from a macroeconomic point ofview, since countries with fewer financial frictions (e.g., better contract law, en-forcement, and greater corporate transparency) tend to have stronger economicgrowth and lower output volatility (Dolar and Meh 2002; La Porta et al., 1997; andCooley, Marimon, and Quadrini 2004).

Reducing financial inefficiencies can, in principle, lead to a better allocation of re-sources, as well as helping the economy and the financial system to better absorbshocks. Policy measures to increase the quality of financial information (e.g., bet-ter accounting and reporting standards, better corporate governance), as well asthe enforceability of rules, regulations, and contracts governing the financial sys-tem can potentially promote both financial system efficiency and stability. Evalu-ating the ultimate net benefit of any given financial system policy, however,requires careful monitoring and analysis. In particular, policies should be re-viewed so that they are, indeed, efficiently achieving the established public policygoal.

Financial System Review

Overview

his section of theFinancial System Reviewexamines the recent performance of theCanadian financial system and the factorsboth domestic and international, that are

influencing it. In each issue, one or more subjectsparticular interest are discussed as highlighted topic.

T

Key Points

• Continued strong global economicgrowth has reduced financial vulnerabili-ties in Canada and elsewhere.

• Higher oil prices have, however, raisedquestions about the strength of globaleconomic prospects going forward.

• The global financial system appears wellpositioned to absorb increases in interestrates over time in several countries,including Canada.

• On balance, conditions in the Canadianfinancial system have improved since thepublication of the June 2004 Review, andmost risks have remained unchanged.

• Overall, the reported financial situationof Canadian non-financial corporationsalso improved through the first threequarters of 2004, despite the apprecia-tion of the Canadian dollar. Profitabilityrose considerably in most industries,aided by stronger global demand.

• Initiatives aimed at enhancing investorconfidence in Canadian capital marketshave been introduced in Canada over thepast few years. Some of these initiativesare reviewed in this issue.

,

ofs

The global economy expanded strongly in2004. In this environment, the steps taken bybusinesses and financial institutions to restruc-ture their balance sheets have bolstered the sta-bility of the international financial system.

Global growth is expected to moderate some-what in 2005, in part because of higher oilprices. Nonetheless, policy interest rates are ex-pected to move up in a number of countries ascentral banks further remove some of the sub-stantial degree of monetary stimulus that hasbeen in place over the past several years.

High levels of indebtedness persist among cer-tain borrowers worldwide (such as households,firms, and some governments in emerging mar-kets). This high level of debt, in conjuction withthe prospect of further increases in policy interestrates, poses some risks for the global financialsystem. However, borrowers appear, well posi-tioned overall to deal with higher borrowing costs.

In particular, corporations and financial institu-tions in most industrial countries have improvedtheir financial positions. As well, with higher in-terest rates being part of the cyclical process ofeconomies returning to full production capacity,higher borrowing costs are not expected to haveadverse effects on the global financial system.

In Canada, the health of non-financial corpora-tions and households is important to thesoundness of domestic financial institutions.Business and household credit account forroughly 30 per cent and 70 per cent, respective-ly, of the credit portfolio of financial institutions.This Review contains an assessment of the po-tential impact of changes in financial marketconditions on both Canadian corporations andhouseholds.

On the corporate side, the analysis indicatesthat, since the late 1980s, improvements in themacroeconomic environment and in corporate

3

Developments and Trends

balance sheets, together with financialinnovations and changes in the debt-maturityprofile of firms, have decreased the corporatesector’s exposure to cyclical movements in inter-est rates. As well, Canadian banks have reducedtheir exposures to large corporate loans.

The financial situation of the non-financial cor-porate sector continued to improve in the firstthree quarters of 2004. Despite the appreciationof the Canadian dollar since early 2003, profit-ability improved considerably over this periodin most industries with a high exposure to inter-national trade. The surge in commodity pricesboosted profits in commodity-producing sec-tors, while weighing on the profitability of somecommodity-consuming sectors. The furtherstrengthening of the Canadian dollar in the pastfew months is expected to place additional fi-nancial pressure on some sectors and firms thatare highly exposed to international trade. Over-all, risks to the financial system arising from thenon-financial corporate sector are currentlyconsidered to be small.

The indebtedness of Canadian households hascontinued to increase, reaching record levels. Inthis Review, the ability of households to servicetheir debts in the event of either a cyclical in-crease in interest rates or a fall in house prices isassessed. The analysis suggests that a cyclical in-crease in interest rates should not significantlyaffect the credit quality of household debt, andthat the possibility of a significant reversal inhouse prices in major Canadian housing mar-kets is unlikely.

Financial institutions in Canada are in a soundfinancial position. Indeed, major Canadianbanks reported continued strong profitabilitythrough the first three quarters of 2004, sup-ported by a diversified business strategy. Cana-dian banks also continue to report high capitallevels, well above minimum requirements. Oth-er financial institutions also posted solid finan-cial results. On balance, conditions in theCanadian financial system have thus continuedto improve since the spring of 2004.

On the regulatory front, important initiativesthat aim to enhance investor confidence in Ca-nadian capital markets have been introduced inCanada over the past few years. These have beenintroduced by various governments, regulators,and financial market participants. A descriptive

4

overview of some of these initiatives is providedin this Review.

Although the state of domestic and global fi-nancial systems has improved since the JuneReview, the global financial system continues toface some risks. Even if developments in theyield differentials of emerging-market debt re-flect a lessening of macroprudential fears, andthese differentials thus remain low, a risk per-sists that they will increase dramatically. Thiscould occur if expectations about policy rateswere revised upwards sharply. But the likeli-hood of this remains small.

The global economy also continues to face risks.Adjustment to the large U.S. trade and fiscal def-icits is one risk. Recent renewed broad-basedU.S.-dollar weakness appears to reflect concernsabout these deficits. There are also uncertaintiesrelated to the geopolitical situation and the highlevel and volatility of crude oil prices.


Chart 2 Bank Rate

%

Source: Bank of Canada

-2

0

2

4

6

8

10

12

14

16

-2

0

2

4

6

8

10

12

14

16

1985 1990 1995 2000

Nominal

Real(nominal minus

y/y core CPI inflation)

Chart 3 Non-Financial Business Sector DebtCapacity: Debt-Service Costs toCash Flow

%

Source: Statistics Canada

10

20

30

40

50

10

20

30

40

50

1985 1990 1995 2000

Chart 1 Debt-to-Equity Ratio of CanadianNon-Financial Corporations


0.9

1.0

1.1

1.2

1.3

1.4

0.9

1.0

1.1

1.2

1.3

1.4

1988 1990 1992 1994 1996 1998 2000 2002 2004

Highlighted Issues

Three issues are discussed in this section: thepotential impact of higher interest rates onCanadian corporate balance sheets, the financialposition of the Canadian household sector, andinitiatives to enhance investor confidence inCanadian capital markets.

Potential impact of higher interestrates on Canadian corporatebalance sheets

The health of Canadian non-financial firms isimportant to the banking sector. Indeed, despitethe decreasing exposure of Canadian banks tocorporations, business credit still accounts forabout 30 per cent of the credit portfolio offinancial institutions. In the past, Canadianfinancial institutions have experienced significantlosses on their corporate loans during economicdownturns or periods of sharply rising interestrates. It is therefore worthwhile to assess howthe removal of monetary policy stimulus in thecurrent cycle might affect the corporate sector.

First, the response of the balance sheets of non-financial businesses to the period of monetarypolicy tightening at the end of the 1980s is ex-amined. The subsequent trends and develop-ments that have affected corporate balancesheets are then reviewed, and the potential im-pact of rising interest rates on the Canadiannon-financial business sector is assessed.

It is important to note that this analysis drawsupon aggregate indicators of financial conditionsin non-financial businesses. Although this pro-vides useful information, conditions across dif-ferent industry sectors and individual businessescould also have important implications that arenot captured in this discussion.1

Monetary policy tightening in the late 1980sBoth the macroeconomic environment and thehealth of corporate balance sheets were very dif-ferent in the late 1980s than they are currently.In particular, the corporate sector was more high-ly leveraged (Chart 1). As well, Canadian bankshave reduced their exposures to large corporateloans over the past few years. The late 1980s werealso characterized by entrenched expectationsof moderate inflation. The need to control infla-tion led to a period of rapid monetary policy

1. See the “Industry” section on p. 22 of this Review.

5


Chart 4 Gross Impaired Loans to Total Loans

By category

%

Source: Banks’ quarterly reports and Office of theSuperintendent of Financial Institutions (OSFI)

0

2

4

6

8

10

12

0

2

4

6

8

10

12

1985 1990 1995 2000

All loans,major banks

Non-mortgagebusiness loans,

all banks

Chart 5 Non-Financial Corporate SectorFinancing

Percentage of total credit

%

* Includes non-residential mortgage loans, leasereceivables, commercial paper, and other loans


Bank loans

Bonds anddebentures

Equity

Other*

10

15

20

25

30

35

40

45

10

15

20

25

30

35

40

45

1985 1990 1995 2000

Chart 6 Non-Financial Corporate Sector:Short-Term Credit

Share of total credit minus equity and trustunits

%


40

45

50

55

60

65

70

40

45

50

55

60

65

70

1986 1991 1996 2001

tightening. The Bank Rate increased from 7.2 percent in 1987 to 13.8 per cent in 1990 (Chart 2).

With corporations borrowing relatively largeamounts of short-term debt, rising short-terminterest rates led to a large increase in the debt-service costs of the business sector. In addition,the economic slowdown itself had an importantadverse impact on profitability. As a result, theratio of debt service to cash flow increased sub-stantially for businesses, from 19 per cent in1987 to 47 per cent in 1992 (Chart 3).2

Deteriorating corporate balance sheets also hada negative impact on Canadian banks. The ag-gregate loan quality of Canadian banks deterio-rated as the ratio of gross impaired loans to totalloans rose sharply from 3.4 per cent in 1989 to5.8 per cent in 1992 (Chart 4). Although the datafor business-loan losses is available only from1994 onwards, it appears that corporate loanlosses were an important driver of the total ag-gregate loan quality in the late 1980s.

What might happen with the expectedremoval of monetary stimulus in the currentcycle?With inflation expectations well anchored tothe 2 per cent inflation target, the removal ofmonetary stimulus over the current cycle is an-ticipated to entail relatively modest adjustmentsin interest rates. As well, non-financial corpo-rate profits have been high and are expected toremain robust, acting as a buffer when interestrates and debt-servicing costs rise. The increasein the debt-service ratio of businesses shouldthus be significantly lower than was the case inthe late 1980s. Moreover, the balance sheets ofCanadian banks are much healthier. In particular,the ratio of gross impaired loans to total loansis currently very low (see Chart 4).

The rise in the debt-service ratio is also likely tobe restrained by recent financial innovationsand by changes in the composition of corporatebalance sheets over the past decade. A numberof trends indicate improvement in Canadiancorporate balance sheets. First, equity financinghas become a major financing source, account-ing for 31 per cent of total outstanding businesscredit in 2004, up from 26 per cent in 1987(Chart 5). As a result, the debt-to-equity ratio

2. Business debt service to cash flow is defined as inter-est payments divided by (after-tax profit plus non-cash items).

6


has fallen from 1.17 in the late 1980s to thecurrent historic low of 0.96 (recall Chart 1).Lower leverage implies healthier corporate bal-ance sheets and lower bankruptcy risk, every-thing else being equal. Indeed, for a given levelof asset risk, reduced leverage implies that cor-porate balance sheets will be less affected byhigher loan rates when interest rates rise.

At the same time, there have been major devel-opments in the debt-maturity profile of corpo-rate debt. First, the share of short-term debt intotal credit has fallen steadily from 65 per centin 1987 to 42 per cent in 2004 (Chart 6). Thelower reliance of firms on short-term debt sug-gests a smaller increase in refinancing costswhen debt is rolled over, should interest ratesrise, or more generally, lower liquidity risks.Second, bank loans, which were the major sourceof financing in the late 1980s, have been replacedby the increasing issuance of bonds and deben-tures. For example, bond financing increasedfrom 16 per cent of total credit in 1990 to 28 percent in 2004 (Chart 5).

Greater reliance on market financing and the in-creased use of derivatives, such as interest rateswaps, have offered corporations more flexibili-ty to adjust their debt profiles and to managedebt-servicing costs based on their exposure andinterest rate expectations. Indeed, interest rateswaps allow corporations to effectively unbun-dle funding and interest rate decisions.3 This isa significant improvement from the situation inthe late 1980s, when bank loans, the mainsource of financing, offered fixed terms to matu-rity and derivatives markets were less devel-oped. But the increased use of interest rateswaps and other derivatives also implies thatthe share of short-term credit in total corporatecredit may no longer be a useful stand-aloneguide for assessing corporate exposure to variousmaturities of interest rates. Information aboutthe direction of the aggregate transfer of interestrate exposures from corporations from the useof interest rate swaps would be required. Thisinformation is not, however, readily available.

A more complete assessment of the interest rateexposures of non-financial corporations re-quires an analysis of their defined-benefit pen-sion plans. Although pension plans are technicallyoff-balance-sheet items, the corporate sponsors

3. These instruments do not necessarily eliminate inter-est rate risk but enable the corporation to tailor itsexposure to interest rate risk.

of defined-benefit plans are legally required tomeet pension liabilities. The average corporatedefined-benefit plan in Canada has been in def-icit since early 2002. These deficits arose prima-rily because of weak equity markets, whichreduced asset values, and falling interest rates,which increased the present value of pensionliabilities (Armstrong 2004).

On balance, rising interest rates reduce pensiondeficits. Higher bond yields tend to reduce thevalue of bond holdings (which typically com-prise about 40 per cent of the assets of large pen-sion plans), but also lower the value of 100 percent of pension liabilities (which are calculatedas the present value of future benefits). There-fore, the net effect of rising interest rates on cor-porate pension funding is significantly positive.For example, at the end of the third quarter of2004, Mercer’s Pension Health Index shows anaverage funding ratio (assets divided by liabili-ties) of 89 per cent, up from a cyclical low of83 per cent in 2003. Mercer estimates that an in-crease in bond yields of 100 basis points at allterms to maturity could improve the averagefunding ratio to about 96 per cent.4

Overall, the analysis indicates that the improve-ments in the macroeconomic environment andin corporate balance sheets, including increasedreliance on equity financing, changes in thedebt-maturity pattern, as well as financial inno-vations, have decreased the exposure of the cor-porate sector to interest rate risk. Therefore, theremoval of monetary stimulus in the current cycleshould have only a relatively small impact oncorporate balance sheets, especially when com-pared with developments in the late 1980s.

Financial position of theCanadian household sector:Autumn 2004

The December 2003 Review featured an assess-ment of the potential impact of changes in thefinancial environment on the ability of house-holds to service their debts. That analysis foundthat financial system risks relating to the creditquality of the household sector remained man-ageable.

Since then, household credit has strengthenedfurther (Chart 7). This trend reflects the increasesin both mortgages and consumer credit, which

4. This calculation is only meant to be suggestive, sinceit assumes that all other variables are held constant.

7


Chart 7 Household Credit

Year-over-year rate of growth

%


Totalhousehold

credit

Consumercredit

Residentialmortgage credit

0

2

4

6

8

10

12

14

16

18

0

2

4

6

8

10

12

14

16

18

1990 1992 1994 1996 1998 2000 2002 2004

Chart 8 Distribution of Lending byCanadian Financial Institutions

Source: Bank of Canada. Based on year-to-dateaverage balances as of September 2004

Consumercredit22%

Businesscredit30%

Residentialmortgage credit

48%

Chart 9 Canadian Household Debt Ratios

% %


15.0

15.5

16.0

16.5

17.0

17.5

18.0

70

80

90

100

110

120

130

1986 1991 1996 2001

Household debtto total assets:Market values

(left scale)

Debt to disposableincome

(right scale)

includes vehicle loans, credit card loans, reno-vation loans, and lines of credit. Consistentwith these developments, household spendinghas been a significant contributor to domesticeconomic growth since 2000. This growth alsoreflects the increasingly widespread access ofhouseholds to various sources of financing, butraises the prospect of their increased sensitivityto interest rates.

The financial health of Canadian households isincreasingly important to the banking sector inlight of the greater emphasis that Canadian bankshave placed on retail lending since 2002. House-hold credit now represents over one-half ofthe total loan exposure of financial institutions(Chart 8). About 70 per cent of household creditis mortgage debt, of which roughly half is insured.About 70 per cent of this mortgage debt has a termto maturity greater than two years. Nevertheless,if the ability of households to service their debtsis significantly compromised by higher interestrates, a drop in house prices, or a reduction inhousehold disposable income, the resultant re-duction in credit quality could adversely affectlending institutions. For instance, higher interestrates could significantly increase the debt-servicecosts of households, and lower house prices couldresult in the value of the financed property drop-ping below the value of the mortgage. Both out-comes could reduce the ability and willingnessof households to service their debts.

In the following assessment of the potential im-pact that changes to the financial environmentmay have on the ability of households to servicetheir debts, it is important to note that this anal-ysis draws upon broad-based indicators of house-hold financial conditions. While this providesuseful information, particular conditions acrossdifferent levels of household income couldhave important implications not captured here.

Servicing household debtTotal household indebtedness, as measured bythe ratio of debt to disposable income, has con-tinued to rise, reaching a record high of 119 percent in the second quarter of 2004 (Chart 9).However, this higher indebtedness should beseen against the background of rising house-hold net wealth arising from higher prices forhouses and other assets. Indeed, the new Statis-tics Canada balance sheet estimate of householdindebtedness relative to net assets on a market-value basis has actually eased over the past year.

8


Chart 10 Household Sector Debt

% $ thousands

* Does not include principal repayments (See Box 1.)Source: Statistics Canada and Bank of Canada

calculations

Householddebt-service

ratio*(left scale)

Effective interest rateon household debt

(left scale)

Real disposableincome per family

(right scale)

4

6

8

10

12

14

16

18

41

42

43

44

45

46

47

48

1980 1985 1990 1995 2000

Chart 11 Canadian Financial Indicators

Percentage of populationaged 20 and over %

* 12-month moving sumSource: Statistics Canada, Canadian Bankers

Association, and Bank of Canada calculations

0.1

0.2

0.3

0.4

0.5

0

0.5

1.0

1.5

2.0

1986 1991 1996 2001

Personalbankruptcies*

(left scale)

Residential mortgage loans in arrears3 months or more (per cent of total

residential housing loans) (right scale)

Credit carddelinquency rate

(90+ days)(right scale)

Chart 12 Projections of the Debt-ServiceRatio Based on Different Paths forthe Overnight Rate

%

Source: Statistics Canada and Bank of Canadacalculations

7

8

9

10

11

12

13

7

8

9

10

11

12

13

1986 1991

Stress-testingscenario

4.0%interest rate

scenario

6.0%interest rate

scenario

Actual Projection

1996 2001 2006

1981Q1 Peak

1980–2004Average

14 14

In addition, the cost of servicing these higherlevels of indebtedness has remained near historiclows, owing to low consumer and mortgage in-terest rates (Chart 10). Indicators of the degreeof financial stress affecting households also re-main positive. For example, residential mort-gage loans in arrears and the rate of credit carddelinquency have both remained low (Chart 11).

Potential challenges: Rising interest ratesBy making a number of assumptions, it is possi-ble to gauge the impact of higher interest rateson the debt-service burden of Canadian house-holds.5 (See Box 1 for a description of the meth-odology used to conduct these simulations.)Simulations assess the impact on the aggregatehousehold debt-service ratio of nominal short-term interest rates returning to more normallevels of 4 to 6 per cent in 2005. (Such rateswould be roughly consistent over time with anenvironment of two per cent inflation.)

The results indicate that in 2005–07, the debt-service ratio would increase to a range of 9 to11 per cent, from the current level of about7.3 per cent. At these levels, the debt-serviceratio would remain near its 1980–2004 average,but would stay well below earlier peaks(Chart 12).6 Although households may bemore sensitive to a given change in interestrates, a steady flow of income remains the keyfactor in the ability of households to servicetheir debt. In this respect, current prospects foremployment growth, and economic conditionsmore generally, remain supportive.

The same method can be used to assess house-hold debt-servicing costs under an extreme in-terest rate scenario. In the extreme scenario,interest rates are assumed to increase in 2006as they did in 1994–95 (short-term rates rapidlyrise by 400 basis points above the midpoint oftheir historical range, to 9 per cent), beforereturning to 5 per cent in 2007.7 Under such ascenario, the debt-service ratio increases by 2007

5. Similar simulations were initially reported in theDecember 2003 issue of the Review. Simulationsreported here are based on a richer and more com-plex methodology.

6. Other components of household financial portfoliosare also affected by changing interest rates, but theirimpact on household wealth would be relativelysmall. See the December 2003 Review.

7. Thus, short-term rates are assumed to increase by650 basis points from their level on 26 November 2004.

9

10


One measure of the cost of household indebted-ness is the debt-service ratio, calculated as the shareof disposable income required to make interestpayments on existing debt. To estimate the futurepath of all the components of this ratio a numberof assumptions must be made.

The data

Estimating past levels of the debt-service ratio re-quires data on its various components. StatisticsCanada data on personal disposable income andhousehold debt are used. Posted rates on consumerloans are used to calculate consumer debt-servicepayments. The use of posted rates would overesti-mate interest payments on mortgage loans, giventhe prevalence of mortgage rate discounting inCanada. Therefore, estimated discounted rates areapplied to a share of mortgage loans, increasingfrom 25 per cent in 1990 to 80 per cent in 2004.

Information on the term structure of householddebt in the 1980s and 1990s is used to calculate aweighted average of interest rates (Montplaisir1996–97). The key development in this area is theincreasing popularity of variable-rate mortgages,which currently account for about 20 per cent oftotal mortgages.

Estimating the future path of thedebt-service ratio

The debt-service ratio is calculated as the product ofthe debt-to-income ratio and prevailing interestrates. In all scenarios, the debt-to-income ratio isassumed constant at its most recent level of119 per cent.1

Interest payments are the product of debt levels andhousehold borrowing rates, weighted by the share ofeach type of debt in the total debt. Moving averagesof interest rates are applied to fixed-rate debt to cap-ture the lagged effects implied by long-term debtcontracts. For variable-rate debt, the relevant inter-est rate is that prevailing in a given period.

A flattening of the yield curve is assumed to lead to areduction in the share of variable-rate mortgages, asborrowers shift to fixed-rate mortgages. Accordingly,

the share of variable-rate mortgages is adjusteddownwards as interest rates rise.2

Scenario 1: Rates return to more nor-mal levels

We consider the normal range for short-term interestrates in Canada to be 4 to 6 per cent. This is roughlyconsistent with historical experience since 1970and with a 2 per cent inflation rate.

Estimating lending rates that are consistent withnormal short-term rates requires two steps: First,risk-free rates for different maturities are derived byassuming that the historical yield spread is a measureof the normal term premium and using linear inter-polation; second, historical intermediation spreadsare added to the risk-free rates to obtain lending ratesfor different maturities. The assumed rise in interestrates to more normal levels and the associated flat-tening in the yield curve take place at a measured pace.

Scenario 2: Stress testing

In this scenario, interest rates increase in two phases:First, interest rates rise to a normal level of 5 per cent.The assumptions underlying the calculations in thisphase are consistent with Scenario 1. Second, furtherincreases occur for all maturities that are similar inmagnitude to those of the 1994–95 period.3

The second phase is associated with a further flatten-ing of the yield curve, since long-term rates increasemore modestly than shorter-term rates becauseagents perceive the increases as transitory.4 The shareof variable-rate mortgages is not reduced in the sec-ond phase owing to the transitory nature of the rateincreases.

1. Higher interest rates could slow debt growth andlead to a decline in the debt-to-income ratio.Alternatively, if the debt-to-income ratio main-tains its upward trend, then the calculationswould underestimate the impact of rising rates ondebt-service costs.

2. Variable-rate mortgages tend to cap households’debt-service payments, with changes in interestrates affecting mainly the rate at which the princi-pal is repaid. Our calculations do not allow for thisfeature and may therefore overestimate the impactof higher interest rates on household cash flow.

3. This is a commonly used stress-testing scenario inrisk management. Such increases in interest ratescould result from several different circumstances.For instance, rising rates could be related to a gen-eral increase in risk premiums, perhaps due to adecrease in investor tolerance for risk. We assumehere that such circumstances would not affect thedebt-to-income ratio, which is assumed to remainconstant.

4. The overnight rate rises to a peak of 9 per cent inthe stress-testing scenario, before adjusting back toa more normal level of 5 per cent.

Box 1

Simulating the Impact of Rising Interest Rates on the Debt-Service Ratio


Chart 14 Housing Market Indicators

Thousands of units

Source: Bank of Canada, Royal LePage, and StatisticsCanada

5

10

15

20

25

30

95

100

105

110

115

120

1980 1985 1990 1995 2000

Unoccupied dwellings(apt. and row;

single and duplex)(left scale)

Accommodationratio

(rented/owned)(right scale)

Chart 13 International Levels for HousePrices: Real*

1987=100

* Deflated using national consumer price indexes,except for the U.K. where retail prices were used

** Weighted prices for new and existing housesSource: Bank of Canada and Royal LePage (Canada),

Statistics Canada (Canada and U.S.), AustralianBureau of Statistics (Aus.), Halifax plc (U.K.)

50

100

150

200

250

50

100

150

200

250

1980 1985 1990 1995 2000

Canada existing house pricesCanada new house pricesUnited States**United Kingdom**Australia**

but remains below its historical peak of 13 percent, before falling back (Chart 12). Such a rap-id increase in the aggregate debt-service ratiocould adversely affect the quality of householdcredit.

A more likely scenario of cyclically rising inter-est rates should provide households with timeto adjust their spending behaviour. Such an out-come should not significantly affect the abilityof households to service their debts, and there-fore should not significantly affect financial in-stitutions.

Potential challenges: House pricesIn recent years, the prices of new and existinghouses in Canada have risen more graduallythan was the case in the late 1980s, and the in-crease has been more subdued than in other in-dustrialized countries (Chart 13). Nevertheless,it is useful to assess the nature of recent develop-ments in the Canadian housing market andtheir potential impact on the Canadian finan-cial system.

First, rising disposable incomes and historicallylow interest rates in the past several years sup-port rising housing prices (recall Chart 10). Sec-ond, the accommodation ratio, which measuresthe relative cost of renting a dwelling in com-parison to owning, continually increased dur-ing the 1990s (Chart 14); i.e., the cost of rentingincreased compared with the cost of owning.8

This contributed to a rise in the number of po-tential homebuyers and supported housingprices. Although recent house-price increaseshave contributed to a very gradual decrease inthe accommodation ratio since 2000, it remainsnear its historical average. This suggests thathouse prices are not out of line with their fun-damental value. Indeed, various other measuresof housing affordability in Canada indicate thathousing costs remain near or below their averageover the past 20 years.9 Third, recent home pur-chases have been made largely for occupancypurposes, rather than for potentially speculativeinvestment purposes. Indeed, the increase inhousing prices over the past five years has

8. The accommodation ratio is equal to the rented-accommodation component of the CPI divided bythe owned-accommodation component of the CPI.

9. These include the ratio of household income takenup by mortgage costs (principal and interest compo-nents) and the RBC Housing Affordability Index.

11


Chart 16 Bank Capital Ratios

%

Source: OSFI

4

6

8

10

12

14

4

6

8

10

12

14

1980 1985 1990 1995 2000

Tier 1 capital ratio

Total capitalratioOSFI Total benchmark

OSFI Tier1 benchmark

Chart 15 Housing Starts and New MLSListings

Thousands of units Thousands of dwellings

Source: MLS and CMHC

50

100

150

200

250

300

30

40

50

60

70

80

1980 1985 1990 1995 2000

Housing starts, allareas (left scale)

New MLS listings(right scale)

been accompanied by a general downwardtrend in the number of unoccupied (vacant)dwellings across Canada and in major urban areas(Chart 14).10 Finally, recent strong increases inthe supply of housing in response to demandand price pressures suggest that market forcesare working effectively. Indeed, listings of exist-ing houses have risen strongly over the past year,and housing starts have increased significantlysince 2002 (Chart 15).

Taken together, these developments indicatethat risks to the financial system related to de-velopments in the Canadian housing marketare currently limited. Because current housingprices seem to be supported by strong funda-mentals—and should continue to be supportedby positive economic conditions—with veryfew signs of speculative behaviour, the possibil-ity of a significant reversal in house prices inmajor Canadian markets is unlikely.

Although the current assessment of the poten-tial impact of changes to the financial environ-ment on the ability of households to servicetheir debt levels does not suggest near-term dan-gers, other longer-term structural developmentsaffecting households may warrant monitoring.Indeed, it appears that certain risks may be mi-grating from various segments of the financialsystem to the household sector and may thusaffect the future savings behaviour of house-holds.11

Exposure of lending institutions to householdsAs indicated in Chart 8, household credit repre-sents over one-half of the total loan exposureof financial institutions. Much of this exposureis, however, backed by assets. Indeed, mortgagedebt, which accounts for almost 70 per cent oftotal household debt, is supported by the valueof the property, and high-ratio mortgages are

10. This analysis does not exclude the possibility thatmodest imbalances exist in certain regional or spe-cific segments of the housing market (e.g., condo-miniums).

11. For instance, there appears to be a shift away fromdefined-benefit pension plans towards defined-con-tribution plans. Should this trend become more gen-eralized, the retirement income of an increasingnumber of future pensioners will depend on the per-formance of pension plan assets rather than being aguaranteed level. Such structural developments,although diversifying overall financial system risks,suggest that risks borne by households are rising.

12


covered by mortgage insurance.12 Even theremaining portion of household credit, thatof consumer credit, is partially secured byassets such as real estate. Over the past fewyears, it is estimated that approximately30 per cent of consumer credit is secured.13

In addition, the Canadian banking systemcurrently retains capital well in excess of thatrequired (Chart 16). This suggests that theCanadian banking system could withstand adeterioration in the quality of household credit.

Overall, financial system risks relating to house-hold credit quality remain low. A cyclical in-crease in interest rates should not significantlyaffect the credit quality of household debt, andthus should not significantly affect financial in-stitutions. As well, the possibility of a significantreversal in house prices in major Canadian mar-kets is unlikely.

Selected initiatives to enhanceinvestor confidence in Canadiancapital markets

Broader participation in financial markets islikely to result when participants can be confi-dent that markets are free from manipulation andfraud. This provides a clear incentive for stake-holders to foster markets that operate in a way thatmerits confidence and trust. Towards this end,Canadian federal and provincial governments,securities regulators, and market participantshave implemented various initiatives to supportfinancial market integrity.

There is a link between financial system efficien-cy and the stability and resiliency of both the fi-nancial and real sectors of the economy.14 Assuch, initiatives that help improve the efficiencyof the financial system have the potential to

12. The securitization of household debt has also beenanother significant trend in consumer debt. Securiti-zation allows banks to restructure their exposure tothese loans, effectively selling it in the form of bondsto a range of investors.

13. The importance of lines of credit (of which morethan 40 per cent are secured, mainly by residentialproperty) has grown, and these now represent 48 percent of consumer credit. Personal loans (of whichabout 45 per cent are secured) represent 22 per centof consumer credit. See CIBC World Markets (2004)and Clayton Research (2003).

14. See Bauer (2004) on page 37 of this Review for defini-tions of market efficiency.

improve financial stability. One approach towardsreducing financial inefficiencies is to increasethe integrity of financial markets.

There is no standard definition of market integ-rity. The lack of integrity is easier to identify; forexample, instances of price manipulation by asingle firm (or investor), misleading accountinginformation, poor governance practices, or ille-gal insider trading. A lack of market integrity hasclear negative implications for the efficiency ofthe market. If investors do not have confidencein the quality of financial information and inthe regulators’ ability to enforce market princi-ples of fairness, they may require higher riskpremiums or reduce their willingness to partici-pate in financial markets. Both such occurrencescould increase the cost of capital.

Over the last several years, financial markets inseveral countries have experienced a number ofhigh-profile corporate scandals (e.g., Enron,WorldCom, Parmalat, and Bre-X) that have hada negative impact on investor confidence in thefairness and integrity of financial markets.These occurrences, while relatively isolatedevents, highlight the need to continue to worktowards a high level of market integrity. Severalinitiatives have been designed to strengthen in-vestor confidence in Canadian capital markets.15

Market behaviourSome of these initiatives relate to developingprinciples and best practices to guide the con-duct or behaviour of market participants, withthe aim of improving liquidity and efficiency inthe markets.

In 1998, the Bank of Canada and the federal De-partment of Finance worked actively with marketparticipants in developing a series of initiativesto promote and maintain the integrity of Canadianfixed-income markets in general, and the well-functioning of the Government of Canada secu-rities market, in particular. These efforts were di-rected to the primary and secondary markets sothat the possibility and perception of marketmanipulation were minimized.

In the primary market, the Bank and the Depart-ment of Finance revised auction rules for the is-suance of Government of Canada securities and

15. See Armstrong (2003) for an earlier discussion of theinitiatives that aim to enhance domestic governancepractices.

13


enhanced the Bank’s role in monitoring auc-tions to increase confidence in the auction pro-cess for government securities and to encourageparticipation from customers and dealers.

The Bank and the Department of Finance alsoworked actively with the Investment DealersAssociation (IDA) to develop Policy No. 5, acode of conduct for trading in domestic debtmarkets. While Policy No. 5 applies directly toIDA members, its principles are expected to be aguide to all participants in the wholesale debtmarket.

Initiatives to enhance standards in the Canadi-an foreign exchange market and the global mar-ket for the Canadian dollar have also beenundertaken. In 2001, the Canadian Foreign Ex-change Committee (CFEC), together with theCanadian Committee for Professionalism(CCFP) and the Financial Markets Associationof Canada (FMAC), adopted the ACI16 ModelCode as the recommended standard for bestmarket practices in the Canadian foreign ex-change marketplace. The Model Code combinesthe recommendations of six pre-existing codesof conduct (those of New York, London, France,Singapore, Tokyo, and the original ACI Code).As such, it is a comprehensive set of globalguidelines for best market practices and person-al conduct for over-the-counter foreign ex-change markets.

EnforcementThe lack of enforcement of security laws can bea challenge to market integrity. Credible en-forcement by regulators is vital to establishingthe right incentives to deter market participantsfrom committing any fraud or malpractice. Ini-tiatives to bolster enforcement in Canada havebeen implemented by governments, securitiesregulators, and self-regulatory organizations(SROs), especially since 2002. (See Box 2 for adescription of enforcement initiatives.)

Corporate governance and disclosureThere have also been initiatives to encouragefirms to develop practices that are in the best in-terest of all investors. Accounting and disclosurerules have also changed to provide investorswith better quality and availability of financialinformation. Timely, accurate, and truthful fi-nancial information is critical for investors,

16. ACI-The Financial Markets Association.

14

since it allows them to properly evaluate the riskand value of a company in their investment de-cisions.

In June 2002, major Canadian institutional in-vestors created the Canadian Coalition forGood Governance (CCGG), a not-for-profit or-ganization whose mission is to promote “bestcorporate governance practices and to align theinterests of board and management with thoseof the shareholders.” In 2003, the CCGG intro-duced governance guidelines that propose min-imum standards and best practices for Canada’slargest publicly traded companies. Although theCCGG does not have the authority to imposebinding rules on companies, it has reviewed thegovernance practices of over 100 companiesand has recommended improvements. More re-cently, the Toronto Stock Exchange followedCCGG recommendations and introduced newstock symbols that clearly identify shares withnon-conventional voting rights.

In July 2002, federal and provincial regulators,as well as Canada’s chartered accountants, an-nounced the creation of the Canadian PublicAccountability Board (CPAB). CPAB is an inde-pendent public oversight system for accoun-tants and accounting firms that audit “reportingissuers.”17 Its main mission is to strengthen theindependence of auditors and to support inves-tor confidence by promoting consistent, high-quality, independent audits of Canadian publiccompanies. To carry out its mission, CPAB hasdeveloped a comprehensive inspection pro-gram to review the quality-control systems ofparticipating audit firms. With the new rules inplace, accounting firms that audit reporting is-suers must be registered as participating mem-bers in CPAB’s oversight program.

In March 2004, the Canadian Securities Admin-istrators (CSA) adopted a number of regulatoryinstruments governing auditor oversight andthe disclosure requirements applicable to publiccompanies.18 These new rules follow the enact-ment in the United States of the Sarbanes-Oxley

17. CPAB defines reporting issuers as companies thathave raised funds from the Canadian investing publicand who, for that reason, must file financial state-ments with one or more provincial securities com-missions.

18. The CSA is a forum where the 13 securities regulatorsof Canada’s provinces and territories can coordinateand harmonize the regulation of Canadian capitalmarkets.

15


High-profile corporate scandals have refocused at-tention on the enforcement of Canadian securitieslaws. A wide range of coordinated initiatives to bol-ster enforcement have been undertaken by theGovernment of Canada, provincial securities com-missions, and self-regulatory organizations (SROs).These initiatives, described below, include the ded-ication of more resources to investigation and pros-ecution of white-collar crime, the passage of newlaws, adoption of new regulatory instruments,changes to the structure and processes of self-regulatory organizations, and the development ofnew electronic reporting systems.

In 2002, the federal government committed toboosting investor confidence in Canadian financialmarkets. As a result, Integrated Market Enforce-ment Teams (IMETs) were established in 2003.IMETs are multidisciplinary teams jointly managedby the Royal Canadian Mounted Police and otherpartner agencies and departments (e.g., the Depart-ment of Justice). They work closely with securitiesregulators and other federal and provincial author-ities and are aimed at improving the investigationand prosecution of the most serious market-relatedcrimes. IMETS are currently operating in Toronto,Vancouver, Montréal, and Calgary.

Another major federal government initiative wasthe passing of Bill C-13 in March 2004. Bill C-13makes insider trading an offence under the Crimi-nal Code and increases the maximum sentences forexisting fraud offences. It also endows investigatorswith new search powers to obtain data and infor-mation from persons under investigation and pro-vides whistleblower protection to employees whoreport unlawful conduct. These requirements cameinto force in September 2004.

At the provincial level, Ontario brought amend-ments to the Securities and Commodity FuturesActs into force in April 2003. These amendmentsincrease penalties and give the Ontario SecuritiesCommission (OSC) more enforcement power forcertain Securities Act offences. In November 2003,the OSC also launched the Mutual Fund Probe toinvestigate potential late-trading and market-timingabuses by retail mutual funds in Ontario.

Self-regulatory organizations bear certain responsi-bilities for the participants in the Canadian secu-rities industry. Major initiatives at the InvestmentDealers Association’s Enforcement Divisioninclude the implementation of a risk-based

operational approach, development of the CaseTracking System, and the creation of a completeset of policies and procedures for enforcementpersonnel. As of 15 October 2002, all IDA membersare required to use a web-based system (ComSet)to file reports every time the firm receives a publiccomplaint, is subject to a lawsuit relating to securi-ties transactions, or becomes aware of a criminal orregulatory investigation relating to the firm.

The Toronto Stock Exchange’s demutualization ul-timately led to the creation of Market RegulationServices Inc. (RS Inc.)—an independent supervisorof trading in Canada’s stock markets. RS Inc. hasimplemented a common set of trading rules acrossthe country, instituted new early-warning systemsto mark potential trading violations, launched anew national Cease Trade Order database, and par-ticipated in the creation of a national task force toexamine illegal insider trading.

In November 2003, the Insider Trading Task Force,created earlier by regulatory authorities, released itsreport, which recommended practices for prevent-ing, detecting, and deterring illegal insider tradingin Canada. One of the issues identified was tradingthrough nominee and offshore accounts that mayprevent regulators from obtaining sufficient evi-dence in suspected illegal insider trading cases. InMay 2004, the Canadian Securities Administratorsapproved amendments to the IDA’s policy on theownership of corporate accounts, which requiremember firms to identify and verify beneficialowners for every new and existing corporateaccount.

Technological advances have also facilitated com-pliance with the requirements of securities laws. InMay 2003, the System for Electronic Disclosure byInsiders (SEDI) became fully functional. SEDI is aweb-based service that allows insiders to file theirtrades in electronic format and provides informa-tion to the market about the trading activities of di-rectors, senior officers, or significant shareholdersof reporting issuers. The system also serves to deterinsider trading based on confidential information,since insiders know they must publicly disclose allof their trades.

Given the relatively recent implementation of allthese initiatives, it will be important to periodicallyassess their contribution to the efficiency of theCanadian financial system.

Box 2

Initiatives Regarding Enforcement of Securities Laws in Canada


Act of 2002. The Canadian rules are designed toreflect Canada’s particular financial and institu-tional setting. The stated purposes of the rulesare (i) to maintain investor confidence and(ii) to harmonize and improve disclosure re-quirements. Among other requirements, publicissuers are now required to disclose off-balance-sheet arrangements that have, or are likely to have,an effect on the results of operations or financialconditions, important contractual obligations,and material forward-looking information in themanagement discussion and analysis that ac-companies their financial statements. To supportthe reliability of financial disclosure, the rulesstipulate that the Chief Executive Officer andChief Financial Officer of a public company willbe required to certify the accuracy of the finan-cial statements and management discussionand analysis. Moreover, reporting issuers musthave an independent and financially literate auditcommittee with prescribed duties.

In May 2004, the federal government proposed anamendment to the Canada Business CorporationsAct (CBCA) to improve the corporate governancestandards of Canadian companies. The CBCAsets the regulatory framework for all federallyincorporated business corporations, excludingfinancial institutions. Some of the issues beingconsidered are the role and composition ofboards, auditor oversight and independence,and financial reporting and offences.

In October 2004, the CSA published a nationalproposal that would require issuers listed on theTSX to disclose their corporate governance prac-tices with reference to specific guidelines. Com-panies listed on the junior TSX Venture Exchangewould, pursuant to this proposal, be required todisclose only their general practices. These newrules are targeted to come into effect in time forthe 2005 proxy season.

Conclusion

These are just a few of the initiatives underwayin Canada designed to bolster trust and confi-dence in financial markets. Given their recentimplementation and the evolving nature of cor-porate governance, it will be important thattheir contribution to the efficiency of the Cana-dian financial system be assessed periodically.

16


Chart 17 Oil Prices

US$ per barrel, quarterly averages


0

5

10

15

20

25

30

35

40

45

0

5

10

15

20

25

30

35

40

45

1972 1977 1982 1987 1992 1997 2002

Nominal

Real (constant1982–90 dollars)

Chart 18 Evolution of Consensus Estimatesfor Annual Growth of IndustrializedEconomies*

%

* North America, Western Europe, and JapanSource: Consensus Economics Inc. and Bank of

Canada calculations

2.4

2.6

2.8

3.0

3.2

3.4

3.6

2.4

2.6

2.8

3.0

3.2

3.4

3.6

2003 2004

2005

2004

Chart 19 Default Rates on Speculative-GradeBonds

%

* 12 months to OctoberSource: Standard & Poor’s

0

2

4

6

8

10

12

14

0

2

4

6

8

10

12

14

2002 2003 2004*

The MacrofinancialEnvironment

Financial vulnerabilities in Canada and else-where have continued to subside since the JuneReview, aided by the solid pace of the globaleconomic expansion. Nevertheless, high levelsof debt for many borrowers worldwide couldpose some challenges in an environment ofrising interest rates. At the same time, financialinstitutions in many countries, including Canada,are in a sound position, while others are work-ing to improve their balance sheets. These ef-forts will further enhance the financial system’sability to withstand major adverse shocks.

Global environment

In response to higher oil prices (Chart 17), ex-pectations for economic growth in the industri-alized countries in 2004 have stabilized in recentmonths, whereas the outlook for growth in 2005has been revised down slightly (Chart 18). Nev-ertheless, expected growth remains robust, andproduction levels are approaching capacity lim-its in some countries. As a result, some centralbanks have started to reduce the amount ofmonetary stimulus, and further increases in pol-icy interest rates are anticipated. For instance, fi-nancial market participants expect the U.S. federalfunds rate to be raised by a further 25 basispoints this year and by roughly 100 basis pointsnext year (from the current level of 2 per cent).

Improved corporate profitability and ongoingfavourable financing conditions have contributedto a further decrease in various indicators offinancial distress. For instance, according toStandard & Poor’s, the global corporate defaultrate for speculative issuers, based on a 12-monthrolling average, fell to 1.9 per cent in Octoberfrom 4.7 per cent at the end of 2003 (Chart 19).19

Globally, financial market conditions have re-mained favourable since the completion of theJune Review, and volatility has been low. Never-theless, the surge in oil prices since early Augusthas created some unease.

19. The average rate over the 1981–2003 period was5.3 per cent.

GlobalUnited States

European Union

17

18


The increased use of exceptional access to IMFlending, resulting from capital account crises inemerging-market economies, has led to a greaterconcentration of Fund credit. In 2002, in recogni-tion of the increased risk associated with the IMF’sexposures to a few, large borrowers, the ExecutiveDirectors approved a doubling of precautionarybalances to about SDR10 billion. The ExecutiveDirectors have also asked the staff to review possi-ble analytical approaches to assessing the adequacyof precautionary balances (IMF 2004).

While concentration risk in the IMF’s portfolio isnot new, what has changed is the composition ofIMF borrowers. Under the Bretton Woods systemof fixed exchange rates, all member countries couldexpect to find themselves on both the creditor anddebtor side of the balance sheet. But by the 1980s,the industrial countries that were once on bothsides of the balance sheet emerged to becomepermanent creditors. Consequently, the IMF hasgradually evolved to become a lender to higher-riskemerging-market countries. Moreover, the term ofFund programs has lengthened, and in some cases,programs have been followed by back-to-backsuccessor arrangements. Larger loans for longerperiods to a relatively small group of emergingeconomies has arguably increased the risks beingborne by the Fund, all other things being held equal.

To assess the degree of credit risk in the IMF’sportfolio, a measure of expected credit loss (ECL)was constructed using sovereign credit ratings.When credit ratings were not available, ratingswere estimated using macroeconomic and insti-tutional information. Expected credit loss wasthen constructed as follows: For every year, theIMF’s exposure under the General Resource Ac-count for each country is multiplied by the re-spective country’s default rate and then summedacross all borrowing countries:

where ECLt is the expected credit loss at time t,EXPit is the IMF’s exposure to country i, DEFit is therespective default rate associated with the country’sactual or estimated sovereign credit rating, andLGD is the percentage loss, given default assumedconstant across countries and time. We find thatECL had increased over time. While ECL averagedSDR550 million between 1990 and 1998, by 2003it had increased to over SDR2.9 billion. As a per-centage of total loans, ECL in 2003 was nearly4.6 per cent—twice the average of the 1990s.

The calculation of a simple expected credit-lossmetric confirms priors that credit risk has increasedat the Fund in the last decade. Critics of this credit-risk approach would argue that any measure of ECLis inappropriate, given the unique nature of theFund’s balance sheet and the lack of defaults overits 60-year history. Moreover, critics would under-score that the existence of the Fund’s preferred-creditor status and economic policy conditionalityimply that the IMF faces lower credit risk than sug-gested by any conventional credit-risk model.While true, in lieu of more standard risk-manage-ment techniques, the IMF may have to rely moreheavily on its preferred-creditor status, and the ef-fectiveness of conditionality to protect the integrityof its balance sheet, given the structural change inits lending behaviour.

ECLt EXPit DEFit LGD×( ),×i 1=

n

∑=

Box 3

Assessing Credit Risk in IMF Lending


Chart 20 Emerging-Market Bond Index(EMBI+) Spread over U.S.Treasuries

Percentage points

Source: J.P. Morgan Chase & Co.

3

4

5

6

7

8

9

10

11

3

4

5

6

7

8

9

10

11

2002 2003 2004

Chart 21 Corporate Debt-to-Equity Ratios

Note: Country data are not necessarilycomparable.

Source: U.S. Federal Reserve and Thomson FinancialDatastream

0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

1990 1994 1998 2002

Japan

UnitedKingdom

United States

Vulnerabilities to rising interest rates in theexternal economyMost financial system participants appear wellpositioned to face the prospect of rising interestrates. Higher interest rates would likely occur inthe context of solid economic activity, sustainedemployment gains, and robust growth in in-come. As a result, higher borrowing costs wouldnot likely lead to adverse effects on the financialsystem, although some highly indebted individ-ual borrowers or countries could experiencefinancial strain.

Spreads on emerging-market assets have de-creased significantly in recent months, afterhaving increased sharply between the end ofApril and mid-May 2004, following the pros-pect of an earlier-than-expected tightening bythe U.S. Federal Reserve (Chart 20). As dis-cussed in the June Review, there remains a risk ofa possible sell-off of emerging-market assetsshould interest rates in industrial countries risemore quickly and/or higher than currently ex-pected. According to Standard & Poor’s, amoderate rise in global interest rates would notseverely affect the credit quality of emerging-market bonds, but a larger increase could putadded pressure on the credit rating of some ofthese countries.20

In the event of a marked deterioration in thecredit quality of emerging-market issuers, credi-tors, both private and official, could be adverse-ly affected. Questions could even be raisedregarding the balance sheet of the InternationalMonetary Fund (Box 3). But there is every rea-son to believe that the IMF’s financial positionwill continue to remain strong because of thecombination of its existing frameworks forlending and dealing with arrears, its increaseduse of debt-sustainability analysis in guidinglending decisions, and the maintenance of itspreferred-creditor status.

In many industrial countries, corporations haveworked hard to repair their balance sheets

20. Standard & Poor’s (2004). Two scenarios are consid-ered. In the first, yields on U.S. 10-year Treasuries risesteadily to 5.8 per cent in 2008 from an average of4.3 per cent in 2004. The second scenario assumes anincrease to 7.3 per cent over the same period. Thereport analyzes the effect of such scenarios on the fis-cal position of a few sovereign issuers rated invest-ment grade as well as a few rated below investmentgrade.

19


Chart 22 Corporate Profits

Percentage of GDP

* Includes depreciation, unlike profits of othercountries depicted

Note: Profits are defined differently in the countriesdepicted, so comparisons of profitability shouldnot be made.

Source: U.S. Bureau of Economic Analysis, ThomsonFinancial Datastream, and Bank for InternationalSettlements (BIS)

0

5

10

15

20

25

0

5

10

15

20

25

1990 1994 1998 2002

Japan

UnitedKingdom*

UnitedStates

Chart 24 Ratio of Equity Capital to TotalAssets: Aggregate Banking Sector

* Monetary and financial institutions for the euro zoneNote: Country data are not necessarily comparable.Source: Federal Deposit Insurance Corporation and

Thomson Financial Datastream

4

6

8

10

4

6

8

10

1998 2000 2002 2004

United Kingdom

Euro zone*

United States

Chart 23 Household Debt

Percentage of disposable income

Note: Country data are not necessarily comparable.Source: U.S. Federal Reserve, Thomson Financial

Datastream, and BIS

50

100

150

50

100

150

1987 1992 1997 2002

Australia

UnitedKingdom

United States

since 2002. They have reduced their indebted-ness (Chart 21), improved their liquidity posi-tions by extending the maturity profile of theirdebts, and improved their profitability (Chart22). Although the improvement has not beenuniform among countries, corporations inmany industrial countries appear well posi-tioned to cope with higher interest rates.21

In contrast to the situation of corporations,households in most industrialized countrieshave continued to increase their debt load(Chart 23). However, a large share of householddebt is mortgage related and is as such support-ed by the value of the property financed. As a re-sult, household indebtedness is not viewed asposing financial system risks in most countries.

Financial institutions in most industrial coun-tries seem well placed to cope with the risks dis-cussed above in the event that they materialize.They have continued to report robust profitsand maintain a substantial amount of capital(Chart 24). In Japan, corrective measures,which have been adopted to address the vulner-abilities of the banking sector, are having visiblepositive effects.22

The discussion on “Global imbalances” on page12 of the June Review also remains topical, as ex-ternal imbalances in the United States, whichmirror external imbalances elsewhere, continueto be a source of uncertainty for the globalfinancial system.Various policy or structuralchanges are required to facilitate the reductionof these imbalances. One such adjustment istaking place. Since the June Review, there hasbeen a general depreciation of the U.S. dollar ona trade-weighted basis.

21. In the June 2004 issue of the Financial Stability Review,however, the Bank of England expressed the view thatU.K. corporate indebtedness was above its long-rundesired level.

22. Owing to ongoing restructuring efforts and improvedeconomic conditions, three of the four major Japa-nese banks marked a return to profitability, withdeclining loan losses (as a share of total loans) and arising capital-adequacy ratio for the year. This strongperformance is reflected in the Moody’s July 2004upgrade of the credit ratings of four Japanese banks.Nevertheless, many individual banks are still weak,especially regional banks. In addition, although capi-tal-adequacy ratios appear satisfactory, a large shareof capital, about 30 per cent for the sector, still con-sists of deferred tax assets.

20


Chart 25 Real GDP Growth: Canada

%


Year-over-yearpercentage change

Quarterly growthat annual rates

-2

0

2

4

6

8

-2

0

2

4

6

8

1999 2000 2001 2002 2003 2004

Chart 26 Financial Position of the CanadianNon-Financial Corporate Sector

%

Source: Statistics Canada and Bank of Canada

-2

0

2

4

6

8

10

12

14

16

0.95

1.00

1.05

1.10

1.15

1.20

1.25

1.30

1.35

1.40

1988 1993 1998 2003

Return on equity(left scale)

Debt-to-equityratio

(right scale)

Chart 27 Canadian Business Confidence

2000Q4 = 100

Source: The Conference Board of Canada andCFIB Research

60

70

80

90

100

110

120

60

70

80

90

100

110

120

1997 1999 2001 2003

Large firms

Small firms

Canadian Developments

Canadian economyCanada experienced substantial economicgrowth in the first half of 2004 (Chart 25). Inparticular, increases in both final domestic de-mand and exports were quite strong. Continuedstrong growth of final domestic demand is ex-pected to contribute importantly to solid eco-nomic expansion through to 2006. Whilefurther global economic expansion should im-ply ongoing gains for Canadian exports, this islikely to be more than offset by faster growth inimports. Indeed, the appreciation of the Cana-dian dollar since early 2003 is expected todampen the growth of Canada’s exports andboost that of imports through 2005.

Corporate sectorThe financial situation of the non-financial cor-porate sector improved still further in the firstthree quarters of 2004. Profitability strength-ened between the end of 2003 and the thirdquarter of 2004, and leverage continued to bevery low (Chart 26). In particular, profitabilityimproved considerably in most industries witha high exposure to international trade, givenstronger output growth and substantial increas-es in selling prices, notwithstanding the appre-ciation of the Canadian dollar. Profit levels alsoremained quite high in most sectors with a lowexposure to international trade.

The surge in the prices of both energy and non-energy commodities so far in 2004 contributedto the increases in selling prices and thus hadimportant effects on the profitability of certainsectors.23 Profits in commodity-producing sec-tors rose considerably in the first three quartersof this year. On the other hand, profitability insome sectors that use commodities intensively(e.g., air transportation) has been adversely af-fected, especially in cases where firms had limit-ed ability to pass on their higher costs in theform of increased selling prices. At the sametime, other industries with relatively high mate-rial costs (e.g., chemical manufacturing) still ex-perienced marked gains in profits because of

23. The general increase in commodity prices chieflyreflects the strong global economic recovery. Theeffects on profits have been limited in the short run,by hedging (by both commodity producers and con-sumers) and by the presence of long-term pricingcontracts.

21


Chart 28 Return on Equity: AutomotiveManufacturing

%

* EstimateSource: Statistics Canada

-10

0

10

20

30

-10

0

10

20

30

1999 2001 2003

1988–2003average

*

Chart 29 Return on Equity: Wood and PaperManufacturing

%


-5

0

5

10

15

-5

0

5

10

15

1999 2001 2003

1988–2003average

Chart 30 Return on Equity: Electronics andComputer Manufacturing

%


-20

-15

-10

-5

0

5

10

15

-20

-15

-10

-5

0

5

10

15

1999 2001 2003

1988–2003 average

substantial increases in demand and, hence,marked rises in output and product prices.

Given the overall improvement in profitability,the confidence of large firms has remained ro-bust (Chart 27). On the other hand, the confi-dence of small firms has eased slightly since theend of 2003, reflecting the adverse impact ofmuch higher insurance and energy costs.24

IndustryA limited set of industries have continued to beunder some degree of financial stress so far in2004. These industries account for only about13 per cent of the capital of the non-financialbusiness sector, and so the risks they pose to thefinancial system are limited.

The profitability of Canada’s auto manufactur-ing industry recovered markedly in 2003 andthe first three quarters of 2004 (Chart 28). Ratesof return for auto parts producers are likely toease considerably over the coming months be-cause of much higher raw materials costs (aslong-term contracts are renegotiated) and thethreat of overseas competition.

Profitability in the wood and paper industry fellconsiderably in late 2003 and early 2004, chief-ly as a result of the substantial appreciation ofthe Canadian dollar and much higher energycosts (Chart 29). Stronger product prices (espe-cially for building products) boosted profits inthe second and third quarters of 2004. But withthe substantial drop in these prices since mid-September and the subsequent appreciation ofthe Canadian dollar, profits are likely to easemarkedly towards year-end.

Profitability in the electronic and computermanufacturing industry remained low in thefirst three quarters of 2004, despite a substantialrecovery in production since late 2002 (Chart30). Ongoing intense competitive pressures,particularly in the market for telecom equip-ment, are likely to result in continued low ratesof return over the near term.

The financial position of the Canadian air trans-port industry improved so far this year, reflect-ing both the impact of restructuring measuresand recent gains in air traffic. But the sharp rise

24. See the Highlighted Issue on page 5 of this Review fora discussion of the potential impact of expectedincreases in interest rates on Canadian corporate bal-ance sheets.

22


Chart 31 Canadian Dollar Exchange Rate

1992=100 U.S. cents

* Euro, yen, U.K. pound, Swedish krona, and Swiss francSource: Bank of Canada

82

84

86

88

90

92

94

96

98

70

72

74

76

78

80

82

84

86

2004

Spot exchange ratevis-à-vis U.S. dollar

(right scale)

Index vis-à-visC-5 currencies*

(left scale)

Chart 32 Volatility of Yields on 2-Year and10-Year U.S. Notes

10-day annualized historic volatility


0

20

40

60

80

100

0

20

40

60

80

100

2004

10-year notes

2-year notes

5-year average

5-year average

in fuel costs has been an important offsettingfactor. As well, some Canadian aircraft suppliershave come under intensified financial pressureas a result of the increased financial stress ofsome of their major U.S. customers.

The financial situation in Canada’s cattle indus-try remains weak because of the continuingban on exports of live cattle, imposed after thediscovery of BSE in Canada. With insufficientCanadian slaughtering capacity, there is animportant risk that Canadian cattle prices willremain low for some time to come. However,federal and provincial governments have prom-ised further financial aid to the industry.

Over the near term, the financial positions ofmany industries with a strong net export orien-tation—and no offsets such as a further rise inworld demand or in commodity prices—arelikely to be adversely affected by the furtherstrengthening of the Canadian dollar in the pastfew months (Chart 31). The adverse impact onprofitability among industries already experi-encing financial stress would likely be especiallypronounced for wood and paper products, autoparts manufacturing, and electronic and com-puter manufacturing. However, it is unlikely thatfinancial institutions with well-diversified port-folios would be strongly affected overall by theworsening financial situation in such industries.

The Financial System

Financial markets

Recent developments in financial markets havebeen driven by factors relating to global eco-nomic growth and the removal of monetarystimulus. Thus far, market reaction to the re-moval of monetary stimulus has been orderly.Indeed, volatility has remained subdued, asboth bonds and equities continue to trade inrelatively narrow ranges (Chart 32). Sources ofrisk to global financial markets remain, howev-er. These include the recent surge in oil prices,and the possibility that interest rates may rise bymore and faster than expected.

Adapting to higher policy interest ratesAs noted earlier, several central banks have al-ready started to reduce monetary stimulus. Thetiming and magnitude of the tightening to datehas, however, been widely anticipated by market

23


Chart 33 Yield on U.S. and Canadian10-Year Notes/Bonds

%


3.6

3.8

4.0

4.2

4.4

4.6

4.8

5.0

5.2

3.6

3.8

4.0

4.2

4.4

4.6

4.8

5.0

5.2

2004

Canada

United States

Chart 34 North American Stock MarketIndexes


8000

8200

8400

8600

8800

9000

1000

1050

1100

1150

1200

1250

2004

TSX(left scale)

S&P 500(right scale)

participants and, thus, has not led to an increasein market volatility.

Overall, the impact of higher policy rates onlonger-term yields in Canada and in the UnitedStates has been relatively muted (Chart 33).Their effect on 10-year yields in both countrieshas been largely counterbalanced by concernsfrom market participants about the sustainabilityof the U.S. recovery (and its impact on Canada).

To date, higher short-term interest rates havenot had a material impact on other asset classes.Equity markets have generally traded within anarrow range in 2004 (Chart 34). While expect-ed earnings respond negatively to higher inter-est rates, they have, nonetheless, remainedstrong, owing to robust global economic activity.The risk that the pace and magnitude of interestrate increases prove to be greater than currentlyexpected, and the resultant impact on asset prices,including equities, remains.25

Oil pricesCurrently, the main risk to global financial mar-kets relates to the surge in oil prices to recordnominal highs, and its impact on longer-termglobal growth prospects and financial marketvolatility (Chart 17). If high energy pricespersist, they will affect financial markets throughvarious channels, including a dampening effecton earnings and thus on equity prices of non-energy companies, as well as through changesin nominal exchange rates. The net impact ofhigh oil prices on fixed-income markets is likelyto be more ambiguous because such prices couldpotentially temper global economic growth,and thus put downward pressure on yields.They could also result in higher inflation, andthus put upward pressure on yields. Given theunderlying low-inflation environment andwell-anchored inflation expectations, however,market participants expect high oil prices tohave their major impact on economic growth.

Government-sponsored enterprisesThe Office of Federal Housing EnterpriseOversight (OFHEO) in the United States, whichregulates government-sponsored enterprises

25. See the June 2004 Review, pages 4-8 for a detailed dis-cussion of factors affecting the prices of financialassets, including prospects for higher interest ratesand the risks to asset valuations. That discussionremains current.

24


Chart 35 Total Assets: Fannie Mae andFreddie Mac

US$ billions

Source: Fannie Mae and Freddie Mac

Fannie Mae

Freddie Mac

0

200

400

600

800

1000

1200

0

200

400

600

800

1000

1200

1996 1997 1998 1999 2000 2001 2002 2003

Chart 36 Bank Profits

% Can$ millions

Source: Banks’ quarterly financial statements

0

5

10

15

20

25

0

1000

2000

3000

4000

5000

1996 1998 2000 2002 2004

Net income(right scale)

Return on equity(left scale)

(GSEs),26 recently released a report critical ofFannie Mae’s accounting policies. Specifically, theOFHEOfoundpervasivemanipulationofearnings.The manipulation asserted in the OFHEO reportwas allegedly designed to smooth earnings vola-tility and help meet earnings expectations. Fur-thermore, the report suggested that Fannie Maemay have violated accounting rules in the waythat it accounted for derivative products used tohedge its underlying mortgage positions.

Although the effect of these findings on financialmarkets has so far been modest and relativelyisolated, a risk exists that further negative devel-opments concerning U.S. GSEs more generally,in light of their scale and level of participationin financial markets, could have important neg-ative effects on financial market stability. Whilethis impact would be largely in U.S.-dollar mar-kets, given the tight linkages between U.S. andCanadian fixed-income and derivatives markets,there is the potential for spillover effects intoCanadian markets.

Fannie Mae is America’s third-largest financialcorporation in terms of assets and plays a keyrole in financing home mortgages in the UnitedStates (Chart 35). It is also the second-largestissuer, behind the U.S. Treasury, of U.S.-dollarsecurities and a significant participant in theinterest rate derivatives market. As such, anymaterial change in the firm’s use of derivativesas hedging instruments has the potential tosignificantly affect U.S.-dollar interest rate swapand options markets.

Since the release of the OFHEO report, FannieMae has agreed to hold 30 per cent more capitalagainst loans than previously required by theregulator. This decision on capital holdingsis identical to that agreed to by Freddie Mac lastyear, following findings of similar accountingirregularities. As well, Standard & Poor’s hasplaced the agency’s subordinate debt and pre-ferred stock on CreditWatch with negativeimplications.

Although much attention has focused on thereporting problems of Fannie Mae and FreddieMac, their senior debt continues to carry the

26. U.S. housing-related GSEs, such as Fannie Mae andFreddie Mac, are congressionally chartered compa-nies owned by private shareholders. They acquirehome mortgages from the lenders that initiallyextended credit and aim to help low- and moderate-income Americans purchase homes.

25

26


On 23 November, the Office of the Superintendentof Financial Institutions (OSFI) published a revi-sion to its Guideline B-5 for asset securitizationtransactions. Many of the changes aim to alignCanada’s regulatory treatment of securitization withthat of other countries. The revision also clarifies anumber of old B-5 provisions.

The revisions that may have the most significantimpact on the Canadian financial system are thosepertaining to facilities for enhancing the liquidityof asset-backed commercial paper (ABCP). They re-move a major impediment to the potential growthof the Canadian ABCP market. This should impartgreater efficiency to Canadian-dollar capital mar-kets.

Because the assets that comprise the collateral aretypically of longer maturity than the ABCP financ-ing them, some sort of liquidity buffer is needed toprotect against rollover risk and timing mismatch-es. Hence, ABCP issuance programs purchase li-quidity protection. At a minimum, in the old B-5such protection should have safeguarded againstwhat it called a “general market disruption”(GMD), which was never defined but was interpret-ed by market participants to mean a situation inwhich “not a single dollar of corporate or asset-backed commercial paper can be placed in the mar-ket at any price.”

Because Standard & Poor’s and Moody’s viewedCanadian liquidity enhancements as too restrictive,they have been reluctant to give their highest in-vestment-grade ratings to Canadian ABCP. But if aCanadian bank provided less-restrictive liquidity, itwould, at a minimum, incur increased regulatorycapital charges that would make the ABCP lesseconomical.

The new B-5 defines a GMD as a “disruption in theCanadian commercial paper market resulting inthe inability of Canadian paper issuers, includingthe SPE, to issue any commercial paper, and wherethe inability does not result from a diminution inthe creditworthiness of the SPE or any originator orfrom a deterioration in the performance of the as-sets of the SPE.”1 This would allow the liquidity fa-cility to be tapped in the event of any non-credit

disruptions, satisfying Moody’s and Standard &Poor’s standards to provide their highest invest-ment-grade ratings to Canadian ABCP, while stillretaining a zero capital charge.

Furthermore, the new B-5 acknowledges that the li-quidity facility might not even need to includeGMD as a restriction, leading to the type of liquid-ity support common in other markets. However,the use of this alternative would also come at thecost of capital charges, as outlined in the Basel IIFramework. Moody’s and Standard & Poor’s haveboth suggested that this approach might also meettheir highest short-term rating standards.

Hence, with the new B-5 it will be possible for Ca-nadian banks to offer expanded liquidity protectionto Canadian ABCP programs on a cost-effectivebasis. Although it is premature to speculate which,if any, of the new liquidity options the banks willadopt, and therefore whether Moody’s or Standard& Poor’s will be able to give Canadian ABCP theirhighest short-term ratings, it appears that the Cana-dian market will at least have that option.

Although Canada’s Dominion Bond Rating Service(DBRS) already gives Canadian ABCP their highestshort-term rating, many institutional investors re-quire two ratings for investments to be acceptable.Hence, the new B-5 could significantly broadendemand for Canadian ABCP and give corporateborrowers expanded and lower-cost access tofinancing.

Box 4

OSFI’s New Guideline B-5 and the Canadian ABCP Market

1. An ABCP program bundles together numerousassets into a “special-purpose entity” (SPE), whichin turn, issues marketable securities.


highest (AAA) rating assigned by the three ma-jor global rating agencies on the strength of theimplicit government guarantee. Although theirsubordinate debt is rated at AA- and is on Cred-itWatch, the default of Fannie Mae or FreddieMac is deemed highly unlikely.

Regulatory and other developmentsOn 26 June 2004, central bank governors andthe heads of bank supervisory authorities in theGroup of Ten (G-10) countries endorsed thepublication of the “International Convergenceof Capital Measurement and Capital Standards:a Revised Framework,” the new capital-adequa-cy framework commonly known as Basel II.

This framework should improve the stability ofthe financial system as a whole. It sets out thedetails for the adoption of more risk-sensitiveminimum capital requirements by banking or-ganizations.27 The new framework reinforcesthese risk-sensitive requirements by laying outprinciples for banks to use in assessing the ade-quacy of their capital and for supervisors to usewhen reviewing such assessments so that bankshave adequate capital to support their overalllevel of risk. Basel II also seeks to strengthenmarket discipline by enhancing transparency inbanks’ financial reporting. The implementationof the new framework is expected to commencein member jurisdictions as of year-end 2007.

In Canada, in November 2004, the Office of theSuperintendent of Financial Institutions (OSFI)published revisions to its Guideline B-5 for assettransactions. Many of the revisions aim to alignCanada’s regulatory treatment of securitizationwith those of other countries. Box 4 outlineshow these revisions may encourage growth inthe Canadian market for asset-backed commer-cial paper.

The securities industries of Canada and theUnited States plan to move to straight-throughprocessing (STP) to clear and settle securitiestransactions.28 STP should increase the

27. The article “Basel II and Required Bank Capital” onpage 61 of this Review examines some potentialimplications of these changes.

28. STP is defined as the seamless passing of informationelectronically—on a timely, accurate, system-to-sys-tem basis—to all parties in the end-to-end chain of asecurities transaction without manual handling orredundant processing. See the Box on page 25 of theJune 2004 Review.

operational efficiency of the securities-settlementprocess. The Bank of Canada supports the effortsof the Canadian Capital Markets Association(CCMA) and all Canadian industry and regulatoryparticipants to take the necessary steps to achieveSTP capability.29 To move towards this objective,the CCMA has, since the June 2004 Review,decided to realign its key priorities. It will nowfocus its efforts on achieving institutional trade-date matching (agreeing to the details of a tradeon the day the trade was executed), the area con-sidered to be the largest hurdle for the Canadianmarketplace to overcome in achieving STP.

Financial institutions

The major Canadian banks reported continuedstrong profitability through their first three fis-cal quarters of 2004. Average return on equity inthe third quarter was 19.3 per cent, comparedwith 18.7 per cent in the second quarter and16.5 per cent for 2003 as a whole (Chart 36).

To date in 2004, the diversified business strate-gy of the major Canadian banks has been sup-portive of continued gains in profitability.Although there were some notable differencesamong the banks, in aggregate, the major seg-ments of their diversified business strategy per-formed well. Credit performance continued tostrengthen, as provisions for loan losses as ashare of average assets declined to 0.04 per centin the third quarter of 2004 (Chart 37). Resultsfrom foreign operations remained mixed, how-ever.

The major Canadian banks continue to reporthigh capital levels (Chart 16), well above mini-mum requirements. From a financial stabilityperspective, the strong capital position of theCanadian banking system may provide a bufferto absorb unexpected shocks that could nega-tively affect banks. These high levels of capitalprovide banks with the reserves from whichthey may choose to carry out future acquisitions,as well as continuing to raise dividends and/orconduct common share repurchase programs.Indeed, TD Bank Financial Group announcedthe purchase of a $5 billion controlling interestin Banknorth Group of Maine. The deal, whichis subject to approval by Banknorth’s shareholders

29. Created in August 2000, the CCMA serves as a forumfor industry experts to provide leadership and coordi-nate the industry-wide implementation of STP.

27


Chart 37 Bank Loan-Loss Provisions

% Can$ millions

Source: Banks’ quarterly financial statements

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0

500

1000

1500

2000

2500

3000

3500

1996 1998 2000 2002 2004

Percentage ofaverage assets

(left scale)

Level(right scale)

Chart 38 Return on Equity of InsuranceIndustry

4-quarter moving average

%

Source: OSFI and Bank of Canada calculations

0

2

4

6

8

10

12

14

16

18

0

2

4

6

8

10

12

14

16

18

1996 1998 2000 2002 2004

Life and health

Property andcasualty

Chart 39 Average Nominal PremiumsWritten: Private PassengerVehicles*

Annual index (1992=100)

* Figures represent the average premium across sixprovinces: ON, ALTA, NS, NB, NFLD, and PEI

Source: Insurance Bureau of Canada

100

120

140

160

180

100

120

140

160

180

1993 1995 1997 1999 2001 2003

and by U.S. and Canadian regulatory authori-ties, is expected to close in February 2005.30

The life insurance industry continued to reportsolid financial performance in the first threequarters of 2004, with the three largest Canadi-an insurers reporting a return on equity of about15 per cent (Chart 38). Strong balance sheets,firm sales in most product lines, and good geo-graphical diversification helped insurers sustainprofitability.

The profitability of the property and casualty in-surance industry, which experienced a signifi-cant gain in 2003, stabilized in the first threequarters of 2004. Rising premiums and im-proved investment income allowed the industryto report a return on equity of about 18 per centso far in 2004, following 11.3 per cent in 2003,and only 1.7 per cent in 2002. There is still un-certainty, however, surrounding the prospectsfor the performance of this industry in light ofincreasing competition. The automobile mar-ket, which accounts for more than one-half ofall premiums collected, faces particular chal-lenges. Some provincial governments imple-mented premium rollbacks or rate freezes in2003–04 in exchange for cost savings for theauto insurance industry (in the form of legislat-ed reforms expected to reduce adverse claims).The effect of the reduced premiums on industryrates of return are, however, expected to begreater in 2005 as premium reductions are fullypassed on to holders of auto insurance (Chart 39).

In late October, the New York Attorney Generalsued Marsh & McLennan, the world’s largestinsurance broker, on the basis that it receivedspecial payments (known as contingent com-missions) from insurance companies that werecharging above-normal commissions. Althoughthere has been a significant impact on the mar-ket value of the equity of most insurance firms,including Canadian firms, it is not clear to whatextent the investigation in the United States willaffect the Canadian insurance industry.31

Clearing and settlement systems

Clearing and settlement systems are a key com-ponent of the financial system, allowing pay-ments and other financial obligations to clear

30. Other, smaller foreign acquisitions by major Cana-dian banks have also taken place.

31. In 2004, Canadian property and casualty insurerspaid out special commissions to insurance brokers.

28


Chart 40 Value of Payments Processed by theLVTS

Average daily amount per month

$ billions

Source: Canadian Payments Association

90

100

110

120

130

140

150

90

100

110

120

130

140

150

2001 2002 2003 2004

Chart 41 Value of Payments Processed by theACSS


$ billions

Source: Canadian Payments Association

14

16

18

20

22

24

14

16

18

20

22

24

2001 2002 2003 2004

and settle. The Bank of Canada supplies servicesto a number of these systems, including theLarge Value Transfer System (LVTS), for settlinglarge-value or time-sensitive payments; the Au-tomated Clearing Settlement System (ACSS),used mostly for smaller-value retail and someelectronic payments; the CDSX, Canada’s secu-rities settlement system; and the ContinuousLinked Settlement Bank (CLS), an internationalsystem for the settlement of foreign exchangetransactions. Because of their systemic impor-tance, the Bank has formal oversight responsi-bilities for the LVTS and the CDSX, and sharesoversight responsibilities with other centralbanks whose currencies are included in the CLSBank.32

Recent developmentsPayment flows in the LVTS grew modestly inthe first three quarters of 2004 compared withthe same period in 2003, increasing by 3 per centfrom year-earlier levels to an average of$129 billion per day (Chart 40). On 2 July, thefirst day following the 1 July holiday—a daygenerally associated with large payment flows—a record number of transactions (about 29,000with a value of $163 billion) settled through theLVTS.

A continuing factor in the growth of LVTS pay-ment flows has been the migration of large-value payments from the ACSS. The average dailyvalue of payments sent through the ACSS fell byabout $0.7 billion compared with the sameperiod in 2003, to average about $16 billion inthe first three quarters of 2004 (Chart 41). ACSSpayment flows, which are not as well protectedagainst risk as those in the LVTS, have been ona downward trend in the last four years and arenow about $5 billion below the daily averagein 2000.

On 18 October, a new participant, State StreetBank and Trust Company, began operating inthe LVTS, bringing the number of LVTS partici-pants to 15 (including the Bank of Canada).This is the first new participant to join the LVTSsince it began operations in February 1999.

October 2004 signalled the successful comple-tion of a 12-month transition period of the riskmodel for CDSX, Canada’s system for settlingvirtually all securities denominated in Canadian

32. The Federal Reserve is the lead overseer for the CLSBank.

29


Chart 42 Canadian-Dollar Foreign ExchangeTrades Settled by the CLS Bank


$ billions %

Source: The CLS Bank

0

5

10

15

20

25

30

35

40

5

10

15

20

25

30

35

40

45

2002 2003 2004

Liquidity ratio(right scale)

Daily averagevalue of payins

(left scale)

Daily average value oftrades

(left scale)

dollars. In March 2003, CDSX became opera-tional for the settlement of debt. Equities wereadded in July 2003. To complete the transitionperiod, The Canadian Depository for SecuritiesLtd. (which owns and operates CDSX), partici-pants, and regulators worked on several issues,proposing solutions that ultimately led to ruleamendments and technical adjustments to im-plement the new features of the system.

Foreign exchange trades settled by the CLS Bankcontinue to grow. For the 11 eligible CLS cur-rencies combined, the volume of sides settled inthe first three quarters of 2004 averaged126,000 per day, up 70 per cent from a year ear-lier, for an average daily value of US$1.4 tril-lion.33 The value of sides settled reached arecord US$2.5 trillion on 15 September.

For transactions involving the Canadian dollar,the value of trades settled by the CLS averagedCan$25 billion per day in the first three quartersof 2004, an increase of about 58 per cent fromyear-earlier levels (Chart 42). The value of fundsrequired to settle these transactions, relative tothe value of the transactions themselves, hastrended down to average about 7.5 per cent inthe third quarter of 2004. This ratio represents ameasure of the liquidity savings provided byCLS settlement and compares with a liquidityratio for trades settled by the CLS in all curren-cies combined of just over 2 per cent. In June2004, an additional Canadian bank began set-tling trades through the CLS. It is still the case,however, that the majority of large Canadianbanks do not yet settle the bulk of their foreignexchange trades through the CLS.

Preparations are underway to include four addi-tional currencies in CLS settlement. These arethe Hong-Kong dollar, the Korean won, theNew Zealand dollar, and the South Africanrand. Settlement will begin once all technical re-quirements are met and assuming all regulatoryand CLS approvals are received.

On 31 May 2004, a major Canadian bank expe-rienced a computer problem whose effects werenot fully resolved until 8 June. The most severeconsequence of this problem was at the retail

33. The 11 currencies currently eligible for settlement bythe CLS Bank are the Australian dollar, the Canadiandollar, the Danish krone, the euro, the U.K. pound,the Japanese yen, the Norwegian krona, the Sin-gapore dollar, the Swedish krona, the Swiss franc,and the U.S. dollar.

30


level, because the problem affected the ability ofthis bank to update its client accounts (whichhad widespread repercussions across Canada).These events did not, however, cause significantdisruptions to the operation of systemicallyimportant clearing and settlement systems inCanada. More recently, two other banks experi-enced technical problems affecting their retailactivity.

References

Anderson, S., R. Parker, and A. Spence. 2003.“Development of the Canadian CorporateDebt Market: Some Stylized Facts andIssues.” Bank of Canada Financial SystemReview (December): 35–41.

Armstrong, J. 2003. “Restoring Investor Confi-dence: Background on Recent Develop-ments in Canada.” Bank of CanadaFinancial System Review (December): 53–58.

———.2004. “What Is the Funding Status of Cor-porate Defined-Benefit Pension Plans inCanada?” Bank of Canada Financial SystemReview (June): 45–52.

Bank of England. 2004. Financial StabilityReview. June.

Bauer, G. 2004. “A Taxonomy of Market Effi-ciency.” Bank of Canada Financial SystemReview (this issue).

CIBC World Markets. 2004. “Household CreditAnalysis.” (November).

Clayton Research. 2003. FIRM Household Bor-rowing Survey.

Cooley, T., R. Marimon, and V. Quadrini. 2004.“Aggregate Consequences of Limited Con-tract Enforceability.” Journal of PoliticalEconomy 112: 817–47.

Dolar, V. and C. Meh. 2002. “Financial Struc-ture and Economic Growth: A Non-Technical Survey.” Bank of Canada WorkingPaper No. 2002-24.

Haldane, A., V. Saporta, S. Hall, and M. Tanaka.2004. “Financial Stability and Macroeco-nomic Models.” Financial Stability Review.Bank of England (June): 80–88.

IMF. 2004. Global Financial Stability Report,Chapter 2. (September).

La Porta, R., F. Lopez-De-Silanes, A. Shleifer,and R.W. Vishny. 1997. “Legal Determi-nants of External Finance.” The Journal ofFinance 52: 1131–50.

Montplaisir, M.-C. 1996–97. “The MaturityStructure of Household Financial Assetsand Liabilities.” Bank of Canada Review.(Winter): 33–46.

Moody’s Investor Services. 2004. “Moody’sUpgrades Deposit and Debt Ratings ofFour Japanese Banks of Three MajorBanking Groups.” (July).

Palombi, R. 2003. “Canadian Defaults andRating Transitions 2003.” Standard &Poor’s.

Standard & Poor’s. 2004. “Interest Rate HikeWon’t Spike Sovereign Ratings.” (Septem-ber).

31

Reports


Introduction

eports address specific issues of relevance tothe financial system (whether institutions,markets, or clearing and settlement sys-tems) in greater depth.

The ability of households and firms to confidentlyhold and transfer financial assets is one of thefundamental building blocks of the Canadianeconomy, as is the manner in which savings aredirected to investments. Frictions in financialmarkets can affect the matching of savers andborrowers, impeding the effective allocation offinancial resources. To understand these fric-tions, the “efficiency” of financial markets mustalso be understood. The report, A Taxonomy ofMarket Efficiency, describes the three main def-initions of market efficiency: informational, op-erational, and allocative. The author points outthat these three definitions are linked, with thedegrees of informational and operational effi-ciency helping to determine the degree of al-locative efficiency.

Over the last few years, both institutional andretail investors in Canada have started to placeassets in hedge funds. Previously, these wereavailable only to wealthy investors. The increaseddemand has led to the establishment of a numberof hedge funds and “funds of funds” in Canada.To better understand the implications of thesedevelopments for the Canadian financial system,the Bank of Canada organized a workshop inJune 2004. Portrait of the Canadian Hedge Fund In-dustry describes the nature of these investmentvehicles, their weight in the Canadian financialsystem, and discusses questions raised by thedevelopment of this industry in Canada, parti-cularly the factors affecting its growth, regulation,and potential impact on the Canadian financialsystem.

R

35


A Taxonomy of Market EfficiencyGregory H. Bauer

he Bank of Canada has a long-standinginterest in the stability and efficiency ofCanadian financial markets. In terms ofefficiency, the Bank is concerned with

how well the financial system allocates capitalbetween savers and investors.

This article describes the three main definitionsof market efficiency: informational, operation-al, and allocative. These concepts are describedas they are used in finance theory.1 One impor-tant point raised is that these three componentsof efficiency are linked via a hierarchy: the de-grees of informational and operational efficien-cy help to determine the degree of allocativeefficiency.2 Some important policy implicationsarising from existing research are also examined.3

Informational Efficiency

An asset market is informationally efficientwhen the price of the asset incorporates all theinformation about its “fundamental value.”4

The definition is further refined depending onthe information available to market partici-pants. A market is “weak form” efficient if onlythe information in past prices is contained inthe current price. This rules out using technicaltrading rules to make excess (i.e., risk-adjusted)

1. The definitions provided here were used by DeputyGovernor Sheryl Kennedy (2004).

2. For a summary of the evidence regarding Canadiancapital market efficiency, see Hendry and King(2004).

3. Although the Bank of Canada does not have legisla-tive authority to design and implement policy inmost areas directly affecting informational and oper-ational efficiency, the linkages between these andallocative efficiency motivate the Bank’s involvement.

4. The fundamental value of an asset is the discountedsum of expected future cash flows from the asset,where the discount rate is the risk-free rate plus theexpected risk premium on the asset.

T
returns.5 A market is “semi-strong form” effi-cient if all public information is reflected in theasset price. This rules out trading on public in-formation, such as dividend yields or interestrates, to make excess returns. A market is “strongform” efficient if prices contain all private andpublic information.6 This rules out makingexcess returns via insider trading, because theprices already reflect that information.7
It is important to note that there is no suchthing as a perfectly informationally efficientmarket (the Grossman-Stiglitz paradox). Thiscan be demonstrated by examining what a per-fectly efficient market would entail. In a marketwhere the asset’s price contained all private andpublic information, no one would have an in-centive to do any research on the asset becauseno gains could be made from obtaining superi-or information. The lack of research implies thatthere would be no way for information to be in-corporated into the asset price in the first place.Thus, the price of an asset could not contain allprivate and public information.

The best way to describe the informational effi-ciency of a market is by its degree of relative

5. Trading rules are “technical” when they are basedonly on movements in past prices and volumes.

6. Private (asymmetric) information is informationknown by sophisticated investors in the market butnot known by ordinary investors. This could be(i) insider information about a particular firm;(ii) better forecasts of public information that hasnot yet been released; or (iii) a clearer understandingof information that is in the public domain. Informa-tion in the last two categories can affect either indi-vidual firms or groups of firms. In the financeliterature, the role of private information on assetprices is examined by studying investors’ order flow.

7. Note that trading by insiders may be either legal orillegal, depending on the context. See King andPadalko (2004) for further details.

37

Reports

efficiency. The amount of information in theasset’s price is such that the marginal cost ofproducing the information is equal to the marginalbenefit from trading on the information. Atany given time, an asset’s price does not reflectall available information, however defined. Theinteresting questions are: (i) how long does ittake for information to be incorporated intoprices, and (ii) how does the information getinto the price? The first question is importantbecause savers will want to know that the priceof the asset they are investing in is “fair;” i.e.,that they will not be negatively affected bypreviously known bad news after they invest.The second question relates to market integrity.If insiders have superior information that theasset is overvalued, how do ordinary investorsget that information? Do the ordinary investorsreceive the information after the insiders have(illegally) traded the stock or in a public newsrelease?

Informational efficiency is often confused withthe idea of “random walks” in stock prices.8,9 Itis important to note that the two concepts areseparate. If the risk premium on stocks is mov-ing over time, then stock prices will change inresponse to current market conditions. Thus,stock prices will not be a random walk. However,if the market is semi-strong form efficient, noone will be able to make excess returns by tradingon public information.

Policy implications

• Most research shows that markets react veryquickly to public news announcements(e.g., interest rate shocks). However, suchnews appears to play a very small role in thedynamics of asset prices. Rather, the bulk ofreturns and volatility in stock, bond, andforeign exchange markets comes from therevelation of private information. It is there-fore important for policy-makers in general

8. Stock prices follow a “random walk” if the change ina stock’s price cannot be forecast based on any avail-able information.

9. Loosely speaking, an asset’s price will follow a mathe-matical process called a “random walk” if all marketparticipants are risk neutral, something not observedin everyday life. The “random walk” is a statisticalmodel of prices that does not fit many real-worldprices.

38

to understand why some agents appear tohave information superior to that of othersand how this private information is releasedto the market. The Bank’s research on mar-ket transparency is related to these issues.

• Lessons from previous work on small,open economies carries over to this line ofresearch. Private and public informationgenerated in the U.S. equity and moneymarkets has an impact on Canadian equityprices. (See Albuquerque, Bauer, and Schnei-der 2004.) Importantly, a portion of thisprivate information is related to the beliefsof sophisticated U.S. investors about thepath of future U.S. interest rates (Bauer andVega 2004).

• In general, smaller firms or markets willlikely be less informationally efficientbecause fewer resources will be devoted toproducing market research. This could beworrying for small firms in Canada or forthe Canadian corporate bond market inaggregate. In addition, markets in the earlystages of development (e.g., the Canadiancredit-risk transfer market) are likely to beless informationally efficient and to containmore profit incentives for investors who doresearch.

• Small amounts of informational inefficiencycan significantly affect the price of an asset.Suppose that the price of the asset equalsits fundamental value, as described above.Under this definition, future cash flows arediscounted by a rate composed of a risk-freerate plus an expected risk premium. Empiri-cal work has shown that the expected riskpremium is very “persistent” (i.e., the levelof the risk premium next month is closelyrelated to its value this month). If the cur-rent expected risk premium is “wrong”because of some inefficiency, the error willcarry through to many future periods. Thus,the future cash flows from the asset will bediscounted for some time by an expectedreturn that is incorrect. This would signifi-cantly affect the current price. Thus, smallchanges in policies related to improving


informational efficiency could have a majorimpact.10

• Tests of informational efficiency are compli-cated since they must be performed jointlywith a test of the predictions of an asset-pricing model. For example, researchers can-not say, that the Government of Canadabond market is (relatively) informationally“efficient” without stating which asset-pric-ing model is used to evaluate the prices inthe market. The problem for policy-makersis that there is no consensus as to the “right”asset-pricing model, suggesting thatresearchers have to temper their conclusionsabout informational efficiency. To under-stand the efficiency of a market, policy-makers must understand how prices are setin that market.

Operational Efficiency

Operational (or transactional) efficiency is ameasure of the cost of transferring funds fromsavers to borrowers. It is thus concerned withtransactions costs. In a perfect world, the trans-actions costs present in a market should (withcompetition) reflect the marginal costs of pro-viding the services to the market participants.11

Work on operational efficiency is often con-cerned with the “liquidity” of a particular mar-ket: can investors trade in “reasonable” sizewithout paying large transactions costs? (See forexample, D’Souza 2002.) Finance theory showsthat sophisticated investors (those with privateinformation) trade in markets where there aremany liquidity-based (i.e., non-informed) in-vestors so that they can hide their trades. Thus,the degree of informational efficiency (largeramount of information in prices) is linked to

10. An example using the standard Gordon growthmodel of stock prices illustrates this point. Suppose astock has a dividend of $1 per year that is expected togrow by 3 per cent per year. Also suppose that therequired rate of return on the stock is 5 per cent peryear. Under these assumptions, the price of the stockwould be $50. Now suppose that a market friction iseliminated, causing the required rate of return of thestock to decline by 25 basis points (to 4.75 per centper year). In this case, the price of the stock wouldincrease to $57.14.

11. For a good overview of the operational efficiency ofthe clearing and settlements system, see McPhail(2003).

the degree of operational efficiency (largeramount of liquidity in the market).

Policy implications

• The link between the first two types of effi-ciency raises concern about attempts toimpose transparency on markets (Zorn2004). Sophisticated investors produceprivate information on an asset in order totrade on it and make a profit. This informa-tion is revealed to the market through thetrades and quotes of the investors. Thishelps make the market more information-ally efficient as defined above. Suppose pol-icy-makers cause an operational change byforcing investors to reveal price quotes ortrades that they wish to keep private. Theinvestors will then have less incentive toproduce that private information. Thismeans that the informational efficiency ofthe market will decline. This, in turn, meansa decline in the market’s liquidity, whichwould hurt non-informed (small) liquidity-based traders.

• There are global implications to thisresearch as well. Barriers to transferring capi-tal across borders can exist because of eitherformal capital controls or microstructureissues, such as lack of available liquidity,concerns about asymmetric information,etc. Differences in operational and informa-tional efficiency may also cause traders tochoose alternative markets in differentcountries in which to conduct the sametrade.

Allocative Efficiency

A market is allocatively efficient when the mar-ginal rate of return (adjusted for risk) is equalfor all borrowers and savers. This implies thatinvestors provide funds for projects that havethe highest net present value and that no“good” investment projects go unfunded.12

The concept of allocative efficiency is related tothe large body of literature on the investmentchoices of firms. It is also related to the con-sumption/saving decisions of consumers. Ingeneral, to evaluate whether a market is allocatively

12. This definition is known to most economists as“Pareto optimality.”

39

Reports

efficient requires a very sophisticated model of theeconomy.

The finance literature is, in general, concernedwith a different set of questions. However, animportant and very recent strand in the litera-ture looks at the role played by informationaland operational efficiency in allocative efficien-cy. For example, some papers look at how theamount of private information in a marketaffects the equilibrium required rate of return inthe market (Easley, Hvidkjaer, and O’Hara 2002).If investors fear that certain more sophisticatedinvestors possess information or superiorknowledge about the asset (and that this infor-mation is not currently priced in), then they willdemand a higher rate of return on the asset.Another part of the literature looks at the role ofliquidity in equilibrium rates of return (Pástorand Stambaugh 2003). It is safe to say that theliterature has not sorted out the separate rolesplayed by information and liquidity in assetprices. It is clear, however, that these micro-structure phenomena have an effect on equilib-rium rates of return. Hence, it is safe to say thatmicrostructure finance no longer provides only“small answers to small questions,” which wasa common perception of the early literature.

Thus, the amount of allocative efficiency in themarket can be viewed as depending on the de-gree of informational and operational efficien-cy.13 Prices will allocate resources in an optimalmanner to the degree that they correctly incor-porate information about an asset’s fundamen-tal value.

Conclusions

Research at the Bank has so far focused on theinformational and operational aspects of effi-ciency in various Canadian capital markets. Asnoted above, improving informational and op-erational efficiency can significantly affect assetprices. Thus, changing these aspects via an exog-enous policy shock could lead to significanteffects on the required rates of return for Cana-dian corporations and, in turn, change the wayfunds are allocated in the market. Small policychanges imposed on financial market structurecould thus potentially have large effects on real

13. Indeed, there are different definitions of allocativeefficiency, depending on the information set used tomeasure the equilibrium outcomes.

40

activity. Such policy directives therefore requirea great deal of analysis before implementation.

References

Albuquerque, R., G.H. Bauer, and M. Schneider.2004. “Characterizing Asymmetric Infor-mation in International Equity Markets.”Working Paper, William E. Simon GraduateSchool of Business Administration,University of Rochester.

Bauer, G.H. and C. Vega. 2004. “MonetaryPolicy, Private Information, and Interna-tional Stock Markets.” Bank of CanadaFinancial System Review (this issue).

D’Souza, C. 2002. “Canadian Foreign ExchangeMarket Liquidity and Exchange RateDynamics.” Bank of Canada FinancialSystem Review (December): 59–61.

Easley, D., S. Hvidkjaer, and M. O’Hara. 2002.“Is Information Risk a Determinant ofAsset Returns?” The Journal of Finance47: 2185–2221.

Hendry, S. and M. King. 2004. “The Efficiencyof Canadian Capital Markets: Some Bankof Canada Research.” Bank of CanadaReview (Summer): 5–17.

Kennedy, S. 2004. “Canada’s Capital Markets:How do they Measure Up?” Bank ofCanada Review (Summer): 33–40.

King, M. R. and M. Padalko. 2004. “Pre-BidRun-Ups Ahead of Canadian Takeovers:How Big Is the Problem?” Bank of CanadaFinancial System Review (this issue).

McPhail, K. 2003. “Managing Operational Riskin Clearing and Settlement Systems.” Bankof Canada Financial System Review (June):79–81.

Pástor, L. and R.F. Stambaugh. 2003. “LiquidityRisk and Expected Stock Returns.” Journalof Political Economy 111: 642–85.

Zorn, L. 2004. “Bank of Canada Workshop onRegulation, Transparency, and the Qualityof Fixed-Income Markets.” Bank of CanadaFinancial System Review (June): 39–44.


Portrait of the Canadian Hedge FundIndustryMiville Tremblay

n the early 1980s, a few wealthy Canadianfamilies were investing in large U.S. hedgefunds. Ten years later, a handful of institu-tional investors had joined them. The

practice has spread, and today even small inves-tors have access to this method of managingsecurities through guaranteed-principal notesand closed-end mutual funds. This growingdemand stimulated the emergence of Canadianhedge funds and funds of funds. To better under-stand the rise of this new industry, the Bank ofCanada hosted a workshop last June, and thisarticle presents the highlights of those discus-sions.1

The development of hedge funds in Canada ischaracteristic of an activity that is integrated in-ternationally. Some elements can be identified,but not isolated, as specifically Canadian. Thislevel of integration complicates measurementefforts all the more because data on the globalindustry are themselves imprecise. There are,in fact, no reliable and complete data on theCanadian component of this industry, and wemust, at this point, rely on the judgment of par-ticipants to obtain even estimates of its size. Theinformation presented here is from several formaland informal sources that use different methodsof compilation. (See the box on page 43 for adescription of hedge funds.)

By the end of 2003, there were about 7,000hedge funds around the world, with total assetsof approximately US$800 billion (Hedge FundResearch, Inc. 2004).2 According to various in-formal sources, Canadian capital invested in

1. Most of the speakers at the workshop were represen-tatives of various segments of the Canadian hedgefund industry. We wish to thank them for their valu-able contributions.

2. Others have put the number of funds at over 8,000and have assessed their total assets at nearlyUS$1 trillion.

I
these financial instruments may total up toCan$23 billion.3 However, a large share of thesesavings were managed by funds establishedabroad, mainly in the United States. Funds es-tablished in Canada administer over $5 billionin assets, but much of this is from foreign inves-tors. Funds of funds— specialized organizationsthat manage portfolios invested in several hedgefunds—handled $3.7 billion of Canadian capi-tal in 2003, according to Investor Economics(2003).4 This amount does not include Canadiancapital entrusted to funds of funds that are notregistered in Canada. Finally, retail sales ofproducts linked to hedge funds have reached$7 billion. This simple statistical overviewunderscores the breadth of the hedge fundphenomenon in Canada.
Strong Demand

Canadian demand was at first exclusively, andremains largely, satisfied by foreign hedgefunds. These funds initially targeted large pri-vate fortunes, but it is the growing interest ofinstitutional investors that explains their highgrowth rate in recent years. In keeping withworldwide trends, Canadian pension funds alsoseek to diversify their portfolios with new assets,the prices of which are weakly correlated withthe prices of stocks and bonds and that generatean absolute return, such as real estate, venturecapital, and hedge funds.

While there is a high level of interest in hedgefunds, relatively few pension funds have, as yet,made the move into them. Institutions with con-siderable resources, such as the Ontario Teachers’

3. From here on, all amounts are in Canadian dollars.For purposes of comparison, the market for mutualfunds is $475 billion.

4. This includes only funds of funds registered with pro-vincial regulators.

41

Reports

Pension Plan, the Caisse de dépôt et placementdu Québec, and OMERS, have established theirown portfolios of hedge funds. Some have alsoset up internal teams that use hedge fund strat-egies to generate an absolute return. Note thatCanadian banks have been using these samestrategies for some time in their proprietarytrading operations.

Smaller pension funds have tended to play thismarket using foreign funds of funds, althoughsome pioneers began with direct investments,yielding mixed results. Although they extracthigher management fees, funds of funds offerinstantaneous diversification, as well as exper-tise in the selection and monitoring of invest-ments. Pension funds involved in theseinvestments generally allocate 3 per cent of theirportfolio, although their ultimate target is 5 to10 per cent. Overall, Canadian pension fundshave placed approximately $10 billion withhedge funds.

The retail market is developing rapidly, espe-cially in Quebec, where the Desjardins Groupand the National Bank are selling term depositsand structured notes5 in which the capital isgenerally guaranteed and the yield is that of afund of funds. Such structures have also beenset up by independent funds of funds, whichuse notes issued by Crown corporations for thatpurpose.6 The minimum investment may be aslittle as $500. Northwater, the largest Canadianfund of funds, opted to enter the retail marketwith closed-end mutual funds listed on theToronto Stock Exchange.7 This small-investormarket also exists in Europe but not in the UnitedStates, where only wealthy individuals andinstitutions have access to hedge funds.

A Limited Canadian Supply

On the supply side, approximately 150 hedgefunds established in Canada can be identified,8

but most of them have less than $25 million

5. A structured note is a negotiable debt security withspecial features. In this case, interest is replaced bythe return on a fund of funds.

6. The advantage for the Crown corporation is a lowerborrowing cost.

7. A Canadian insurance company also sells on theretail market a segregated fund linked to a fund offunds.

8. There are fewer managers, since some firms haveseveral funds.

42

under management. The bulk of the capitalmanaged in Canada—an increasing proportionof which is from foreign funds of funds—isin the hands of only six firms: Sprott AssetManagement, Salida Capital, Polar Capital,Leeward Capital, Mapleridge Capital, andJ.C. Clark (Cohen 2004).9 To date, few Canadianinstitutional investors, including funds offunds, have made direct investments in Canadianhedge funds. Medium-sized money managers,for their part, attract private Canadian investorswho are guided by financial advisers. By far themost popular strategy is long/short equities.

To penetrate this lucrative and rapidly growingmarket, traditional fund managers, such as TDAsset Management, Natcan, and Fiera Capital,recently created portfolios that are handled inthe same fashion as hedge funds. These remainsmall operations, however, when comparedwith the volume of capital under traditionalmanagement.

In Canada, the funds of funds sector is relativelymore developed and growing more rapidly thanthe hedge fund sector. Aside from Northwater,the largest independent firms include Norshield,Maple Partners, Tricicle, HR Strategies, andArrow Hedge. The products of several foreignfunds of funds are also sold in Canada, eitherdirectly or through a domestic partner. Exceptfor the National Bank and Desjardins, the majorCanadian deposit-taking institutions are stilllargely absent from this market. In total, about60 funds of funds are active in Canada.

On the other hand, all the major banks have es-tablished prime brokerage services that, in addi-tion to offering execution of trades, providespecialized services to hedge funds, such as fi-nancing, custody of securities, transactions set-tlement, securities lending for short sales, risk-management systems, and even promotionalsupport among investors. Globally, this activityis dominated by three large investment banks.10

In Canada, the main players to date are theRoyal Bank and the Bank of Montreal.

9. Most Canadian hedge funds are based in Toronto,although several Canadian funds of funds are managedfrom Montréal. Quebec’s pension funds are generallymore inclined to invest in this type of asset thanthose of the other provinces.

10. These are: Morgan Stanley, Bear Stearns, and Gold-man Sachs.

43


The term “hedge fund” covers a very diverse field of or-ganizations and behaviour that defies any simple defini-tion. The analysis in this article is presented from twocomplementary angles, which are useful to distinguish,since the typical organization called a “hedge fund” isno longer the only one to exhibit these traits. A small,but growing, number of traditional financial institu-tions, seeking to generate an absolute return, are man-aging their capital in the manner of hedge funds.

Organization

The typical hedge fund is a private investment pool,limited to a few wealthy or institutional clients,1 each ofwhom commits a large amount of money. The organi-zation is usually small and is centred around the exper-tise of its portfolio managers. These, in turn, are oftenveterans of traditional money-management firms orformer proprietary traders at large banks. Various sec-ondary functions are usually subcontracted to special-ized administrators and to prime brokers.Hedge funds cherish a culture of discretion, even secrecy.Regulation bars them from soliciting business from thegeneral public through advertising. But they seek main-ly to protect their market positions, distinctive strate-gies, and the intellectual property of their quantitativemodels. The flip side of the coin is that the investormust usually accept a low degree of transparency in themanagers’ activities and positions.Legally, these funds are constituted as limited partner-ships, are frequently registered offshore, and are subjectto light regulation. The general partners invest theirown capital alongside that of the limited partners, en-suring the alignment of their financial interests. Thegeneral partners receive high management fees, on theorder of 1 to 2 per cent of the assets plus 15 to 25 per centof returns.Since hedge funds often invest in illiquid markets or posi-tions, the limited partners can withdraw their moneyonly after giving advance notice, once per quarter or peryear.Finally, hedge fund managers have realized that theirreturns decline when the assets under management ex-ceed a certain threshold relative to the opportunitiesidentified. Possibilities for arbitrage tend to dissipatewhen too much capital seeks to take advantage of them.Thus, the best-performing funds refuse to accept newinvestors when approaching what they deem to be theoptimal size.

Behaviour

The conduct that includes, but goes beyond, that ofhedge funds is described as absolute-return manage-ment. The central motivation of hedge funds and relat-ed management methods is, indeed, the search for anabsolute return. The goal is expressed as a fixed percent-age (e.g., 15 per cent) or as a markup on a short-term in-terest rate.Traditional management of institutional investmentsfocuses on a relative return—outperforming some

market index. This distinction is blurring, since an in-creasing number of institutional investors now pursuean absolute return.Hedge funds are sometimes called speculative funds,and some of them clearly are. But in general, seeking anabsolute return requires a rigorous and selective manage-ment of risk in which it is more important to minimizelosses than to maximize returns. In fact, the target re-turn is paired with an acceptable level of volatility. Tra-ditional management, on the other hand, seeks tominimize the negative difference with the index, wheth-er the index is rising or falling.Absolute-return managers seek out a specific risk, aboutwhich they have a strong opinion, and neutralize allother risks arising in the investment. Most of the time,they seek to eliminate market risks (and returns), thebeta, and maximize the value added by their talent, thealpha. They usually accomplish this by pairing short po-sitions with long positions. For example, we can imaginetwo equal positions taken on two pulp and paper com-panies. This combination will generate neither profitnor loss in response to broad fluctuations in the stockmarket or in the paper and forest products sub-index. Itwill, however, generate a profit if there is a change in therelative value of the two companies, provided the onesold short declines relative to the other.Thus, while traditional managers can make profits onlyon rising stock prices, absolute-return managers canalso earn money on falling prices. Their scope for prof-itable investments is therefore much broader.Short positions provide liquidities that can be partiallyreinvested in long positions. This hedging naturallygenerates leverage, which may be enhanced by borrow-ing or using derivatives. The extent of the leverage varieswidely according to management strategies and styles.It is estimated at between two and five times, althoughit can be completely absent.Note that some funds of funds also use leverage, whichincreases their returns but also their losses. Given thevarious sources of leverage, it is difficult for the investorto measure its total magnitude.It should also be pointed out that the word “hedge” in“hedge funds” can be misleading in some cases, sincesome of the management styles employed by thesefunds do not seek to hedge against market risks, butrather to speculate on market direction.Firms that compile return indexes for the various typesof hedge funds have established very elaborate classi-fications. They categorize funds according to decisionprocesses, instruments used, and geographical markets.Global Macro funds, which opportunistically bet onsignificant movements in currencies or interest rates,are only one type among many.In practice, freedom of choice in terms of markets andstrategies is constrained by the particular style and spe-cific expertise of the manager, although it is still great-er than that of the traditional portfolio manager.Finally, absolute-return management is characterizedby returns that are weakly correlated with traditional as-set classes, such as stocks and bonds, theoretically al-lowing the creation of portfolios that are less volatile fora given return. There is also a weak correlation amongthe returns from various management styles.

1. This includes funds of funds; i.e., organizations thatactively manage a portfolio of hedge funds.

Box 1

The ABCs of Hedge Funds

Reports

The Issues

Some workshop participants expressed disap-pointment over the fact that relatively fewhedge funds have as yet set up shop in Canada.The size of the industry in various countries canbe measured in relation to that country’s stockexchange. On this scale, despite an annualgrowth rate of 20 per cent, the industry is onlyhalf as developed in Canada as it is in Europe,and only one-sixth of that in the United States.In contrast, the activities of funds of funds andsales of retail products appear to be more devel-oped here.

According to Greenwich Associates (2003), thepercentage of Canadian institutional investorsin the Canadian hedge fund market is lowerthan that observed in the other major industrialcountries, except the United Kingdom. Thus,the main challenge facing Canadian managersis to raise the necessary capital, especially duringthe start-up phase. Some attribute this to thefact that there are fewer large foundations andfamily estates here and that pension funds aresmaller and more conservative. Others maintainthat reduced access to the required technical ex-pertise and the relative smallness of Canadianfinancial markets impede the implementationof certain strategies; for example, merger andacquisition arbitrage.

Workshop participants agreed that current regu-lation is not an obstacle to the development ofthe industry. After a heated debate, the Securi-ties Exchange Commission recently decided toregister hedge fund managers, as is already thecase in several countries, including Canada.Here, as elsewhere, the small investor cannotinvest directly in hedge funds. However, provin-cial regulatory bodies allow retail sales visas fora restricted class of closed-end mutual fundsand for guaranteed capital products, which, inturn, invest in hedge funds. Moreover, one sem-inar participant, who lost a large amount ofmoney in a fraudulent U.S. fund, suggested thatregulation can deter scam artists. Another, how-ever, maintained that registration generates afalse sense of security among investors.

Several international bodies—notably the Fi-nancial Stability Forum, the International Mon-etary Fund, and the Bank for InternationalSettlements—have examined the impact ofhedge funds on the stability of the global finan-cial system. Among the issues that are still on

44

their agenda is the management of counterpartyrisk by the prime brokers of high-leverage hedgefunds. The opacity of these high-leverage fundsis a further issue, as is the protection of smallinvestors who purchase the industry’s retailproducts. Finally, some emerging-market coun-tries remain concerned about the deleteriouseffect that the rapid capital inflows and outflowsassociated with hedge funds may have on thestability of their nascent financial systems.

Nevertheless, some participants emphasizedthat hedge funds typically buy when traditionalinvestors sell, and vice versa, thus bolstering theliquidity of markets and, consequently, theirstability and efficiency. This observation appliesparticularly to arbitrage strategies, which arebased on an expected return to fundamentalvalue, but not to directional strategies, whichbet on existing trends.

Conclusion

The Canadian hedge fund industry is growingrapidly in several market niches. However, themarketing side appears somewhat more devel-oped than the production side. This industrydoes not currently appear to be raising any con-cerns in matters of financial stability, especiallysince it is still small. Nonetheless, its activitiesare largely integrated into the global hedge fundindustry. Several international bodies continueto examine the potential benefits and risks asso-ciated with that industry.

References

Cohen, B. 2004. “Canada Breaks Out of ItsNorthern Stronghold.” Absolute Return 2(May): 30–33.

Greenwich Associates. 2003. The AlternativeBalancing Act. <http://www.green-which.com>.

Hedge Fund Research, Inc. 2004. Hedge FundIndustry Report for Second Quarter 2004.

International Monetary Fund. 2004. GlobalFinancial Stability Report. Market Develop-ments and Surveys. September.

Investor Economics. 2003. Hedge Funds Report.

Policy and

Infrastructure

Developments


Introduction

he financial system and all of its variouscomponents (institutions, markets, andclearing and settlement systems) are sup-ported by a set of arrangements, including

government policies, that influence its structure andfacilitate its operation. Taken together, these arrange-ments form the financial system’s infrastructure.Experience has demonstrated that a key determinantof a robust financial system is the extent to which itis underpinned by a solid, well-developed infrastruc-ture. This section of the Review highlights work inthis area, including that related to relevant policydevelopments.

In common with central banks around theworld, one of the Bank of Canada’s functions isto act as a “lender of last resort.” In this role, theBank is the ultimate source of liquid funds tothe financial system. This is one way in whichthe Bank of Canada contributes to the longer-term robustness of the Canadian financial system.

The article Bank of Canada Lender-of-Last-ResortPolicies sets out the current policies governingthese activities, which have recently been re-viewed by the Bank. The Bank of Canada hasdistinct roles as a lender of last resort, and thearticle outlines how and under what circum-stances it can routinely provide liquidity tofacilitate payment settlement, as well as thevarious ways it can respond in more exceptionalsituations.

T

47


Bank of Canada Lender-of-Last-ResortPolicies

In common with central banks around the world, one of the functions of the Bank ofCanada is to act as a “lender of last resort.” The Bank has recently reviewed its policiesin this area. This article sets out the policies governing these activities.1

he Bank of Canada is the ultimate sourceof liquid funds to the financial system.As such, it routinely provides liquidityto facilitate payments settlement and re-

sponds in various ways to exceptional or emer-gency situations. The Bank of Canada has threedistinct roles as a lender of last resort (LLR).

• The Bank facilitates the settlement of pay-ments systems by routinely extending over-night credit to participants in the LargeValue Transfer System (LVTS) through theStanding Liquidity Facility (SLF), to covertemporary end-of-day shortfalls in settle-ment balances that can arise in the dailysettlement of payments.

• For solvent financial institutions requiringmore substantial and prolonged credit, theBank can provide Emergency Lending Assis-tance (ELA). ELA is intended to overcome amarket failure associated with financialinstitutions that have a significant share oftheir liabilities as “deposits” (fixed-valuepromises to pay, redeemable at very shortnotice) and whose assets are generallyhighly illiquid.

The Bank of Canada Act requires that such lendingbe secured by collateral pledged by the borrowinginstitution. It is the policy of the Bank to lendonly to institutions that are judged to be solventin order to mitigate moral hazard that can arisefrom such potential intervention, and to avoiddamaging the interests of unsecured creditors.

• In conditions of severe and unusual stresson the financial system more generally, theBank has authority to provide liquidity

1. The Bank last presented its views on its lender-of-lastresort policies in its submission to the Estey Commis-sion in 1986.

T
through outright purchases of a wide varietyof securities issued by any Canadian or for-eign entities, including non-financial firms.
Standing Liquidity Facility

The purpose of the Standing Liquidity Facility isto support settlement in the payments systemby providing collateralized, overnight loans todirect participants in the payments system whoare experiencing temporary shortfalls in theirsettlement balances.2

Terms of the SLF

Provision of credit through the SLF is a routineactivity, given under the following terms.

• The Bank provides overnight loans at theBank Rate, an interest rate currently set at25 basis points above the target overnight rate.

• The Bank is required by the Bank of CanadaAct to secure all lending with collateral. Thecollateral eligible to secure credit from theSLF is the same as that eligible for intradaycredit in the Large Value Transfer System.3

• Collateral is valued at market value less adiscount. Discounts are applied mainly toprotect the Bank from market risk (declinesin the value of its security caused by changes in

2. For a discussion of the LVTS and the AutomatedClearing Settlement System (ACSS), see Dingle(1998) and Northcott (2002).

3. Eligible collateral includes securities issued or guar-anteed by the Government of Canada, securitiesissued or guaranteed by a provincial government,Special Deposit Accounts held at the Bank, bankers’acceptances and promissory notes, commercial paperand short-term municipal paper, and corporate andmunicipal bonds. (The last three categories are sub-ject to minimum credit ratings.)

49

Policy and Infrastructure Developments

market conditions), but these “haircuts” alsoreflect the credit risk of the issuer of the secu-rities. Haircuts are set for broad classes of issu-ers and are larger for less-creditworthy issuersand for instruments with longer maturities.4

Access to Bank of Canadasettlement accounts and the SLF

Direct participants in the LVTS are requiredunder Canadian Payments Association (CPA)bylaws to have Bank of Canada settlementaccounts and access to the SLF. Since November2003, the net settlement obligations in the Au-tomated Clearing Settlement System (ACSS)have settled through LVTS payments (on a next-day basis). As a result, all routine credit from theSLF is provided only in connection with the LVTS.(SLF credit would be provided directly for ACSSaccounts only in the event of an LVTS outage.)

The Bank, therefore, provides a settlement andloan facility to any institution in the CPA aslong as it

• participates directly in the LVTS or the ACSS,

• in the case of ACSS direct clearers, settles allnet ACSS positions with LVTS paymentscredited to its ACSS settlement account atthe Bank of Canada, and

• is able to provide the Bank with valid andenforceable first-priority security in collat-eral of a type that is acceptable to the Bank.

The Bank has additional requirements for accessto its lending facility. These are motivated large-ly by the need for the Bank to have a legallywell-founded security interest in the collateralpledged by an institution to support the SLF.

In addition, the various classes of financial in-stitutions eligible for CPA membership, andtherefore able to hold settlement accounts at theBank, are subject to different bankruptcy lawsand regulatory regimes.5 Accordingly, for some

4. For a list of the relevant haircuts, see the payments sec-tion of the Bank of Canada’s website, http://www.bankof-canada.ca/en/payments/rules.htm#collateral.

5. In 2001, eligibility for membership in the CPA wasbroadened beyond deposit-taking institutions toinclude life insurance companies, securities dealersthat are members of the Investment Dealers Associa-tion or the Bourse de Montréal, and money-marketmutual funds that meet certain requirements regard-ing the investment of their holdings and have accessto an immediate and reliable source of liquidity.

50

classes of institution, the Bank probably wouldnot be able to recover funds from any unsecuredportion of a loan. The Bank, therefore, may allowhaircuts on collateral that vary for differentclasses of borrowing institution, or may set dif-ferent restrictions on the quantities of corporatesecurities that can be pledged by different classesof institutions.

As a result of these considerations, the Bank re-quires that an institution wishing to establishsettlement and loan arrangements under the SLF

• provides acceptable legal documentation tosupport the Bank’s security interest in pledgedcollateral, and

• accepts the collateral terms and conditionsthat may be set by the Bank, which take intoaccount varying exposures to credit riskacross different types of institutions.

The required legal documentation includes

• signed Bank of Canada account agreementsand loan and security agreements,

• favourable legal opinions regarding theparticipant’s ability to meet the terms andconditions of these agreements, and

• favourable legal opinions from foreignbranches regarding the applicability of theirhome country’s laws to these agreements.

In addition, upon application for a settlementfacility, the Bank would notify the institution’sregulator that the institution intends to open asettlement account. For a federally regulated fi-nancial institution, it is expected that such noti-fication would be provided as a matter of coursethrough the Financial Institutions SupervisoryCommittee (FISC).6

6. The FISC is the primary interagency committee usedto address issues of financial stability in Canada. Itwas established pursuant to the Office of the Superin-tendent of Financial Institutions Act for the purposeof facilitating consultations and the exchange ofinformation among its members on all matters relatingdirectly to the supervision of financial institutions.Its membership consists of the Super-intendent ofFinancial Institutions (who acts as chair), the DeputyMinister of Finance, the Chairperson of the CanadianDeposit Insurance Corporation, the Governor of theBank of Canada, and the Commissioner of the Finan-cial Consumer Agency of Canada.


Emergency LendingAssistance

While provision of credit through the SLF is aroutine activity that facilitates the settlement of thepayments system, Emergency Lending Assistanceis extraordinary and provides credit to institutionsjudged to be solvent, but that are, nevertheless,facing serious and persistent liquidity problems.

More specifically, ELA is designed to address aparticular kind of market failure associated witha financial institution that issues deposits (fixed-value promises to pay, redeemable at short notice)and that holds a portfolio of non-marketable assetsthat dominates its operations. A large and suddenincrease in the redemption of deposits at such aninstitution could lead to its insolvency, eventhough it is otherwise sound, because its assetscan be liquidated only with difficulty and are sub-ject to discounts. As a practical matter, whetheran institution is subject to this kind of marketfailure is a matter of judgment, and is increasinglyunlikely, given financial developments in Canada,including changes in the regulatory environment.

Terms and conditions of ELA

Under the Bank of Canada Act, the Bank canprovide ELA to a member of the CPA for a maxi-mum term to maturity of six months. The loanscan be renewed for periods up to six months asmany times as the Bank wishes. The minimumrate that the Bank can charge on ELA loans is theBank Rate. While the Bank has discretion tocharge a higher interest rate if it sees fit, in itslimited experience with ELA situations, the Bankhas charged the Bank Rate.

As noted, the Bank is required under the Bank ofCanada Act to secure all lending with collateral.For ELA, the Bank is willing to accept a broaderrange of collateral than that approved for creditunder the SLF. In practice, this would typicallymean taking a security interest in an institution’sCanadian-dollar non-mortgage loan portfolioto support ELA, and the Bank would lendagainst this collateral.7

7. Under the law, mortgages are considered to be a con-veyance of “real property,” which the Bank cannottake as collateral. In cases where the primary assetsavailable to an institution to secure Bank lending aremortgages, the security interest would have to bestructured as an assignment of the mortgage receiv-ables only, and not as an assignment of the mort-gages themselves.

Taking such collateral would require that theBank search security registers for prior securityinterests in the assets to be pledged, deal withany prior secured creditors, and complete speciallegal documentation and agreements with theinstitution—a process that could take two daysto a week or more, depending on the complica-tions that arise. This means that advance legalpreparation is desirable in probable ELA cases,but this is at the discretion of the relevant finan-cial institution.8

Eligibility Criteria for ELA

While the provision of ELA is extremely rare, therisk to the Bank is greater under ELA than underSLF.9 Under SLF, there is no presumption of aprotracted liquidity problem or solvency risk. Incontrast, under ELA, there is clearly a significantliquidity problem affecting the institution, anda prima facie reason to question the solvency ofthe borrower prior to making an ELA loan. Aswell, under the SLF, only high-quality market-able securities are accepted as collateral, whileunder ELA, collateral that is subject to greaterliquidity and credit risk is likely to be taken.

As a result of the significant inherent risk in ELAsituations, the Bank takes more stringent mea-sures with regard to ELA.

• ELA addresses a particular type of marketfailure (discussed above), and the Bank pro-vides ELA only to classes of institutions thatare vulnerable to this type of failure.

• To minimize moral hazard and to avoidimpairing the interests of unsecured credi-tors of the institution, the Bank providesELA only to institutions judged to be solvent.Therefore, a fundamental and critical con-sideration is whether the Bank can receivetimely and accurate judgments on solvency—this is essential to the Bank’s due diligence.

• Since the Bank relies primarily on prudentialsupervisors for this information, a soundsupervisory framework is critical for ELAdecisions and ELA management. Such aframework would include a clear supervi-sory mandate, adequate authority, a programof early intervention, and information-

8. In such a case, the Bank would register in advance itssecurity in the public, personal property security reg-istry of the institution’s home province.

9. The last instance of such lending was in 1986, to theContinental Bank.

51


sharing protocols with the Bank. It wouldalso provide a means to jointly establishremedial measures and to implement work-out strategies. A strong framework mitigatesincentives for supervisors to delay dealingwith a problem institution; such forbearancecould shift risks to the Bank.

• As with lending under SLF, it is importantthat the Bank have a valid first-priority secu-rity interest in any collateral pledged to sup-port ELA.

Implications Regarding Eligibilityfor ELA

These considerations have the following impli-cations for the eligibility of various classes ofinstitutions for ELA.

• Federally incorporated banks (includingforeign bank subsidiaries) and federallyincorporated trust and loan corporationswould be eligible for ELA.10 These firms aresusceptible to the relevant market failure(referred to above). The Bank can be confi-dent of receiving timely and accurate infor-mation regarding the solvency of theseinstitutions. And the federal supervisoryregime provides a reliable means to establishremedial measures and to implement work-out strategies. In addition, the CanadianDeposit Insurance Corporation can act as alimited provider of liquidity to its memberinstitutions (both federal and provincial)through purchases of assets, and loans oradvances (with or without security).

• Insurance companies, mutual funds, andinvestment dealers would not be eligible forELA, since they do not issue deposits or holda significant share of their assets in illiquid,hard-to-value claims.11

• Credit union locals and caisses populaireswould not generally be eligible for ELA. Inmost cases, these institutions have access toprovincial centrals, the Corporation de

10. In the case of trust companies, the “in-trust” nature ofthe assets held by such a firm means that ELA couldbe provided only through a loan secured by companyassets, or through an outright purchase of assets,associated with provisions to sell the assets back tothe trust company at predetermined prices.

11. However, see the section on “Systemic Risk and Bankof Canada Intervention, p. 54.”

52

Fonds de Sécurité de la Confédération Des-jardins (CFSCD), or the Credit UnionCentral of Canada (CUCC), for liquidityassistance.12

• In the case of an extraordinary, widespreadevent that would have significant, adverseconsequences for a provincial credit union/caisse populaire system, the Bank wouldconsider providing ELA through the CUCC,a provincial central, the Caisse centrale Des-jardins, or the Fédération des caisses Desjar-dins, as appropriate, provided that legalarrangements satisfactory to the Bank wereestablished by these entities.13

• With regard to foreign bank branches, in aprospective ELA situation, it could be difficultto receive timely and accurate informationon solvency from foreign supervisors, and tosuccessfully manage the conflicts in incen-tives faced by the relevant supervisors wheninteracting with the Bank in such cases.There can also be legal complications andrisks with regard to establishing a securityinterest for the Bank in some of the assets ofthese institutions in an ELA situation. Accor-dingly, foreign bank branches would notnormally be eligible for ELA. Nevertheless,in very exceptional circumstances where thehome central bank was unable to lend for aday or two (for operational reasons), theBank of Canada could provide interim lendingfor a very brief period, typically against col-lateral that would be eligible for credit throughthe SLF.

Managing ELA

The management of ELA with respect to financialinstitutions subject to federal regulation wouldbe in close collaboration with the Financial Ins-titutions Supervisory Committee, which serves

12. As well, very few credit union locals or caisses popu-laires are members of the CPA.

13. Such lending could require the establishment of par-ticular legal mechanisms to allow the Bank to take asecurity interest in the assets of a credit union orcaisse populaire. (See, for example, footnote 7 above.)It could also require a process of rehypothecation ofthe collateral to the provincial central, the CUCC, orCaisse centrale Desjardins. These arrangements canbe complex and costly to set up. The Bank is preparedto work with relevant institutions to prepare the legalgroundwork for such arrangements.


as a forum to exchange information relevant forsupervision and to coordinate the strategies ofits member agencies when dealing with trou-bled institutions subject to federal regulation.

• The FISC—through the Office of the Super-intendent of Financial Institutions (OSFI)—would normally be aware of prospective ELAsituations. In this regard, the Bank wouldkeep the FISC informed regarding such pos-sibilities, and vice versa.

• The Bank would notify the FISC immedi-ately in the event that the Bank providedELA to an institution.

• The Bank would use the FISC as the primaryforum for the exchange of informationregarding an institution receiving ELA, andthe FISC or a relevant subcommittee wouldmeet at least weekly to consider the situation.

• The borrowing institution would be requiredto provide a business plan to OSFI that out-lined remedial measures to rectify its liquidityproblems, and to provide increased reporting(data and other information) on its evolvingsituation.

• Contingency planning would also be con-ducted at the FISC. Such planning couldinclude possible private sector solutions, aswell as alternative work-out arrangements.

While the repayment of SLF loans is routine, ter-minating ELA is likely to be more complicated.If all goes well, the management of ELA wouldfocus on normalizing the institution’s positionin the market, or facilitating a merger of the in-stitution, such that ELA could be expeditiouslywithdrawn.

Following are the main features of the Bank’sELA management procedures.

• The Bank’s Financial System Committee14

would meet immediately and then at leastweekly to review any ongoing ELA, formallyreconsider the borrowing institution’s sol-vency and the appropriateness of continuingto provide ELA, as well as the limits on lendingto the institution.

14. The Financial System Committee comprises the sixmembers of the Bank’s Governing Council, the Gen-eral Counsel/Corporate Secretary, the Regulatory Pol-icy Adviser, and the Chief of the CommunicationsDepartment.

• If, at any time, the Bank wanted additionalinformation concerning the financial condi-tion of the borrower, the Bank could hire athird-party agent to perform an examinationof the institution.

• The ELA loan agreements between the Bankand the borrowing institution would createa one-day, revolving facility in which theBank would have discretion to decline tomake any further one-day loans. This wouldallow the Bank to readily cease ELA if itjudged that the borrowing institution wasinsolvent, or that the available collateral tosupport ELA was at a higher risk of beinginadequate.

• The Bank would cease ELA when this wasjudged by the Bank to be appropriate, mostnotably, when the institution was judged bythe Bank to be insolvent, on the basis ofinformation received from OSFI and possiblythird-party agents, or when available colla-teral was inadequate to support further ELA.

• If the Bank became aware of a borrowinginstitution’s insolvency or pending insol-vency, it would refrain from taking any newcollateral as security for outstanding advancesmade when the institution was still solvent.At the same time, the FISC would be work-ing to implement an orderly work-out.

Foreign currency ELA

Liquidity support in a foreign currency is an im-portant consideration for Canadian financialinstitutions, given the significance of foreigncurrency activities (mainly U.S. dollar) for manyof these institutions. However, providing liqui-dity support in a foreign currency is considerablymore difficult than providing Canadian-dollarELA: while the Bank can create liquidity inCanadian dollars, it cannot do so in foreign cur-rencies.

• Financial institutions are responsible forensuring that they have reliable arrangementsfor private sector liquidity support in foreigncurrencies important to their business.

• Canadian financial institutions shouldarrange access through foreign central banksto liquidity facilities in those currenciesimportant to their business.

• Provided that the institution qualified forELA, the Bank could lend Canadian dollars

53


on a collateralized basis to the illiquid insti-tution which, in turn, could purchase theneeded foreign currency in the market withthose Canadian dollars.

The Relationship between theSLF and ELA

As noted above, direct participation in the LVTSrequires (under CPA bylaws) access to settlementaccounts at the Bank of Canada and access to theSLF. The Bank provides loans through the SLF tofacilitate the efficient operation of the paymentssystem, provided that the Bank’s requirementsfor SLF (described above) are satisfied.

As discussed, lending under the SLF is routineand low risk: in SLF lending, there are no concernsabout the solvency of the borrowing institution;SLF lending is collateralized by high-quality,discounted securities; and, for any given finan-cial institution, SLF lending is transitory (over-night).

In contrast, ELA is, by its very nature, a high-riskundertaking: ELA arises when there are concernsabout the solvency of an institution; ELA wouldprobably be secured by collateral that is subjectto greater risks; and the potential engagementby the Bank in ELA is indefinite.

It is possible that an institution’s borrowing re-lationship with the Bank might evolve from SLFto ELA under some circumstances. This wouldhave implications for the Bank’s managementof that lending and for the Bank’s relationshipwith that institution. Accordingly, the Bankmonitors the use of the SLF to identify whethera financial institution is using the SLF for ELA-type borrowing. In such a case, the followingwould apply.

• If the institution were considered to be eligiblefor ELA, the Bank would initiate internaland FISC-related processes for managingELA activity, and would require the institu-tion to sign additional ELA legal documen-tation.

• For other LVTS participants that are not con-sidered to be eligible for ELA, upon identifyingELA-type borrowing, the Bank would indicateto the financial institution that additionalborrowing based on a broader range of col-lateral would not be granted, and the Bankwould contact the institution’s regulator.The Bank would deny access to additional

54

liquidity once the institution had exhaustedits SLF-eligible collateral.

Systemic Risk and Bank ofCanada Intervention

Under extreme conditions, the Bank can pro-vide liquidity to any firm. The Bank of CanadaAct, paragraph 18 (g.1), gives the Bank the au-thority, under conditions of “severe or unusualstress on a financial market or financial system”to provide liquidity via outright purchases of awide variety of claims issued by any Canadianor foreign entities, for the purpose of promotingthe stability of the financial system.15

In other words, the Bank has the authority toprovide liquidity to a broad range of financialand non-financial institutions when the Governorof the Bank judges that such transactions arejustified to safeguard the safety and soundnessof Canada’s financial system. All such trans-actions would be fully disclosed and justified inthe Bank’s public statements, including theAnnual Report. The Bank would also need topublish in the Canada Gazette notice that itbelieves that there is a situation of severe andunusual stress on the financial system.

More specifically, Section 19 of the Bank ofCanada Act states that if the Bank takes any ac-tion under paragraph 18 (g.1) it must publish anotice in the Canada Gazette that “the Governorhas formed an opinion that there is a severe andunusual stress on a financial market or financialsystem.” The notice is to be published as soon asthe Governor is of the opinion that its publica-tion will not materially contribute to the stressto which the notice relates.

If problems in a financial institution not eligiblefor ELA under the above policy (but a CPAmember) could, in the Bank’s judgment, lead tosevere or unusual stress on a financial market orfinancial system, then the Bank may choose tomake a liquidity loan instead of making purchasesor undertaking repos under paragraph 18 (g.1).

15. This does not include more general liquidity providedthrough monetary policy actions. The policiesexplained here are over and above the liquidity pro-vided in response to shocks to the financial system,such as the stock market crash of 1987 or the terroristattacks of 11 September 2001.


Forced LVTS Loans

A final category of Bank lending can occur in thecontext of a default in the LVTS. In the eventthat an LVTS participant defaults, the Bank ofCanada could be obliged (under LVTS bylaws)to knowingly lend to an insolvent institution,on the basis of collateral pledged earlier.16 Morespecifically, the Bank would be obliged to lendto the defaulting institution on the day of failureagainst previously pledged collateral to settlethat member’s obligations to other participantsin the LVTS, and so protect against systemic risk.

In the extremely unlikely event of the failure ofmore than one LVTS participant on the sameday during LVTS operating hours, where thesum of the exposures of the failed participantsexceeds the value of all the collateral pledged inthe system, the Bank of Canada guarantees set-tlement of the LVTS.17 In this event, the Bankcould be obliged to lend to a failed institution,on a partially unsecured basis, to ensure settle-ment of the LVTS and so protect against systemicrisk.

As noted, the likelihood of this scenario is ex-tremely remote, and the fact that participantspledge collateral sufficient to cover the singlelargest possible default provides a large elementof co-insurance (a deductible) that providesstrong incentives for LVTS participants to managetheir risks prudently in the system.

16. To secure potential payment obligations, LVTS parti-cipants pledge in advance sufficient collateral to coverthe largest single possible settlement obligation.

17. The Bank provides such a guarantee to ensure cer-tainty of settlement of the LVTS in all possible cir-cumstances. For more on these points, see Goodlet(1997).

References

Dingle, J. 1998. “The LVTS—Canada’s large-value transfer system.” Bank of CanadaReview (Autumn): 39–55.

Goodlet, C. 1997. “Clearing and settlementsystems and the Bank of Canada.” Bank ofCanada Review (Autumn): 49–64.

Northcott, C.A. 2002. “Systemic Risk, Designa-tion, and the ACSS.” Bank of CanadaFinancial System Review (December):29–35.

55

Research

Summaries


Introduction

ank of Canada staff undertake researchdesigned to improve overall knowledge andunderstanding of the Canadian and inter-national financial systems. This work is

often pursued from a broad, system-wide perspectivethat emphasizes linkages across the different parts ofthe financial system (institutions, markets, andclearing and settlement systems). Other linkagesof importance may include those between theCanadian financial system and the rest of the eco-nomy, as well as those with the international environ-ment, including the international financial system.This section summarizes some of the Bank’s recentwork.

Within the next several years, the implementa-tion of an updated global bank capital accord(Basel II) developed by the Basel Committeeon Banking Supervision will begin. It shouldstrengthen the stability of the global financialsystem as a whole. One way in which Basel IIis designed to achieve this goal is by aligningcapital more closely with credit risk in order toensure that appropriate levels of capital are heldby the banking system. But because credit risk isstrongly related to the business cycle, some ob-servers have wondered whether the new ruleswould exacerbate cyclicality in the banking in-dustry and, in particular, the behaviour of bankcapital and lending. In Basel II and RequiredBank Capital, the potential for cyclicality in bankcapital requirements is explored by applyingBasel II rules to Canadian bank data on corpo-rate and sovereign exposures under various sce-narios. One finding is that required capital forcorporate exposures could be more volatile thanunder Basel I, in particular, the greater the use ofmarket-based measures of credit risk relative to“through-the-cycle” measures.

Models of information asymmetry suggest thatif investors believe that insiders systematicallytrade corporate shares on material, non-publicinformation, this will ultimately raise the cost

B
of capital for firms. In Pre-Bid Run-Ups Ahead ofCanadian Takeovers: How Big Is the Problem?Bank of Canada staff investigate whether thereis evidence of illegal insider trading in Canadaahead of a specific type of corporate event;namely, a takeover bid. This line of researchaddresses an important dimension of capitalmarket efficiency—the degree of informationasymmetry among market participants.
What drives the common variation in the move-ments of international stock markets? Knowledgeof how asset prices in an open economy likeCanada’s are affected by foreign information isof direct importance to understanding how theBank of Canada could reduce the likelihoodthat problems in another financial system wouldtrigger instability in Canada. In particular, anassessment of the stability of financial marketsrequires an understanding of how and what in-formation is incorporated into asset prices. Thearticle Monetary Policy, Private Information, andInternational Stock Markets summarizes a paperin which researchers assess whether sophisticatedinvestors possessing superior (private) informa-tion about future U.S. interest rates and stockmarket movements affect international stockmarket comovements.

59


Basel II and Required Bank CapitalMark Illing and Graydon Paulin*

Within the next several years, implementation of an updated global bank capital accord(Basel II) will begin in a number of countries. The new framework is designed to alignbank capital more closely with risk, thereby ensuring that appropriate levels of capitalare held by the banking system. In particular, capital requirements for credit risk will bemodified along the lines of how the most sophisticated banks currently calculate eco-nomic capital for their loan books. Since credit risk is strongly related to the businesscycle, however, it is useful to examine the degree to which required bank capital is likelyto be cyclical.

asel II is composed of three “pillars.”The first is an enhanced set of rules forcalculating minimum capital require-ments, embodying advances in risk

measurement since the first capital accord(Basel I). The second pillar addresses the super-visory review of bank capital adequacy, whilethe third addresses disclosure rules to facilitatethe public assessment of banks.1 The three pil-lars together will determine the actual level ofcapital held by banks, but this article focuses onthe minimum capital requirements arising fromthe first pillar.

The central objective of Basel II’s first pillar is toincrease the sensitivity of bank capital to therisks associated with specific classes of financialassets (particularly credit risk). To this end,Basel II offers banks two potential approachesfor calculating required capital: the Standardapproach and the Internal Ratings Based (IRB)approach. The latter is divided into the Founda-tion and Advanced approaches. The majorCanadian banks, provided they meet regulatoryrequirements, are most likely to adopt theAdvanced IRB approach.

With respect to credit risk, a key aspect of BaselII’s Advanced IRB approach is its use of a creditvalue-at-riskmodel(VaR)2 todetermineminimum

1. A full description of the pillars can be found in BIS(2004).

2. A value-at-risk model generates a statistical distribu-tion of the potential loss associated with holding aspecific financial portfolio over a given period (oneyear in the case of Basel II).

* This is a summary of a recently published workingpaper (Illing and Paulin 2004).

B
levels of regulatory bank capital and lossprovisions.3 Banks that meet rigorous standardswill be allowed to use their own parameterestimates in this model.
If the credit risk faced by a bank is cyclical, it isconceivable that the output of this VaR willyield cyclical minimum capital requirements.Credit risk in Canada does indeed contain astrong cyclical component. Together with theobservation that over 90 per cent of the creditlosses of Canadian banks in the past two de-cades have occurred on their corporate and sov-ereign exposures,4 this raises the issue of howBasel II might affect the cyclicality of requiredbank capital held against their corporate andsovereign portfolios.

To address this question, we applied Basel IIrules to two decades of Canadian bank data oncorporate and sovereign exposures and exam-ined the results under various scenarios.5 An

3. Bank capital will continue to be defined according tothe rules outlined in Basel I and its subsequentrevisions. Loss provisions, alternatively known asreserves, are an amount set aside by banks to coveranticipated losses on assets, potential litigation costs, andother costs not usually defined as operating expenses.

4. Exposures include loans, securities, and other claims.The corporate sector includes interbank exposures.Corporate and sovereign exposures currently repre-sent approximately 28 per cent and 7 per cent ofoverall assets in the Canadian banking system,respectively.

5. Other types of bank exposures were not examined(e.g., residential mortgages or asset-backed commer-cial paper), since in aggregate they do not generatesignificant losses, and are thus expected to have rela-tively stable capital requirements.

61

Research Summaries

Chart 1 Requirements for the CorporateExposures of Canadian Banks

Basel I versus simulated Basel II(credit ratings)

Per cent of corporate portfolio

Basel I ratio includes capital plus general and specificprovisions.

Basel II ratio includes minimum required capital forunexpected loss, required specific provisions forexpected loss, and a charge for operational risk.

Source: Authors’ calculations based on BIS (2002).

Basel I

Basel IIBB-median portfolio

Basel IIBBB-median portfolio

Basel IIA-median portfolio

0

2

4

6

8

10

12

0

2

4

6

8

10

12

1984 1989 1994 1999

Table 1

Volatility of Basel II Requirements

For the Canadian Banking System, 1984–2003

Note: The comparison is based on Basel I capital requirements plus actualprovisions. Basel I requirements are estimated prior to 1988. BaselII requirements include capital for unexpected loss, provisions forexpected loss, and an operational charge as per BIS (2002).

Portfolioquality

Standard deviation in percentage points

Basel Iincludingprovisions

Basel IIusing credit

ratings

Basel IIusing credit

spreads

Corporateexposures ofCanadianbanks

A-median 0.39 0.44 1.49

BBB-median 0.60 0.65 1.80

BB-median 0.86 0.96 2.71

Memorandum item:Observed standard deviation of Canadian banks’ total actual eligiblecapital plus allowances for losses was 0.90.

important caveat is that these simulations cannotcapture behavioural responses that might beinduced by the new rules. In addition, we had toestimate a significant amount of the data—mostcritically, the credit-quality distribution of cor-porate exposures. Therefore, we report results fora range of scenarios that cover different portfoliodistributions and assumptions. These scenariosprovide a sense of how significant the behaviouralresponses might be and how sensitive the Basel IIrequirements are to various assumptions. Finally,our base-case simulations use what we considerto be the most plausible and realistic assumptionsfor the Canadian banking system.

The simulations use detailed data on actualbanking system exposures to corporations (byindustry) and to sovereigns (by country). How-ever, since the precise credit-quality distributionof the corporate exposures is unknown, we pro-vide results for high-, medium-, and low-qualityportfolios (indicated by their median credit rat-ings).The distribution for sovereign exposures isknown precisely, so estimation is not necessary.

We use two methods to track the evolution ofthe corporate distributions over the period1984–2003. First, we use credit-rating-transitionmatrices based on the actual evolution of Cana-dian corporate credit ratings (from ratings agen-cies) over this period. Credit ratings providerelatively stable estimates of credit risk but aretypically slow to respond to a rapid change in creditquality. Second, we track the change in creditquality with credit spreads on corporate bonds.These spreads tend to respond quickly to chang-es in credit quality but are more volatile thancredit ratings.

Both methods are based on data that pertain toonly large Canadian corporations.6 However, thesetwo methods are simplified characterizations ofcommon techniques that banks use to measurecredit risk. These assumptions and data are fedinto Basel II’s Advanced IRB model to generateour simulated results.

Note that the Basel II model distinguishesbetween expected (average) loss and unexpected(upper-bound) loss. Banks must make provi-sions against expected loss (or hold capitalagainst the shortfall), and they must holdcapital against unexpected loss. We present results

6. Thus, we assume that the credit-quality distributionof small corporations is the same as that of largecorporations.

62


Chart 2 Alternative Simulated Basel IIRequirements for CorporateExposures of Canadian Banks

Per cent of corporate portfolio

0

2

4

6

8

10

12

0

2

4

6

8

10

12

1984 1989 1994 1999

Both ratios include minimum required capital forunexpected loss, required specific provisions forexpected loss, and a charge for operational risk. Amedian portfolio rating of BBB+ and an LGD of45 per cent are assumed.

Source: Authors’ calculations based on BIS (2002)

Using creditspreads

Using creditratings

Chart 3 Requirements for the SovereignExposures of Canadian Banks

Basel I versus simulated Basel II

Per cent of exposures

The Basel II ratio includes minimum required capitalfor unexpected loss, required specific provisions forexpected loss, and a charge for operational risk.

Sources: Moody’s (2004), S&P (2004), and authors’calculations based on BIS (2002)

Basel I

Basel II(credit ratings)

1984 1989 1994 1999

0

5

10

15

20

25

30

0

5

10

15

20

25

30

Basel II(credit spreads)

Country-specificprovisions

for the total requirements (minimum requiredcapital and provisions combined). Althoughthe tax implications vary between the two, bothaffect earnings. Canadian banks raise most newcapital through retained earnings, while provi-sions are a deduction from earnings.

Corporate Exposures

When credit ratings are used to measure corpo-rate credit risk, the simulated minimum capitaland provisions for corporate exposures requiredunder Basel II over the period 1984–2003 fell byabout one-third relative to the equivalent Basel Imeasure (Chart 1).7 And there was only a modestincrease in the volatility of requirements(Table 1, middle column).8

In contrast, when yield spreads on bonds areused to measure credit risk, simulated minimumrequirements for corporate exposures were clearlymore volatile under Basel II than under Basel I(Table 1, last column). For example, using thismeasure, required capital and provisions for aBBB-median-rated corporate portfolio doubledbetween 1997 and 2002 (Chart 2), a periodwhere there was substantial cyclical deteriorationin credit quality.

Sovereign Exposures

Next, we measured sovereign credit risk withboth country-specific credit ratings and yieldspreads on sovereign bonds. In both cases, thesimulated Basel II capital and provisions forsovereign exposures were about two times high-er than the capital requirements under Basel I(Chart 3). However, during the late 1980s sev-eral Canadian banks made large country-specif-ic provisions for the debts of less-developedcountries (LDC), most of which occurred beforethe Basel I rules actually took effect.9 If we add

7. We compare Basel II total requirements to Basel Icapital requirements plus actual provisions adjustedfor the credit-quality distribution of the portfolio.

8. Although the decline in capital sounds dramatic, theresults are consistent with those obtained elsewhere.See, for example, Kiesel, Perraudin, and Taylor (2003)and French (2004) for U.S. banks.

9. In addition to the implementation of Basel I in 1988,the supervisory regime in Canada was being reorga-nized, and the large provisions partly reflected therecognition of losses that had, in fact, occurred earlierin the decade during the previous supervisory regime.

63

Research Summaries

these provisions to the Basel I ratio (the red barin Chart 3), then the Basel II requirements ap-pear to be less volatile. This is because Basel IIrules require banks to either provision againstor capitalize probable losses as they are identified.

Conclusions

Our simulations illustrate the change in mini-mum required bank capital in response to his-torical changes in the level and credit distribu-tion of bank assets. By definition, however, theydo not capture the behavioural changes thatwould be induced by the different incentivesunder Basel II. To some extent, however, the dif-ferent scenarios provide an indication of thepotential behavioural impact.

We find that minimum required capital for cor-porate exposures could be more volatile thanunder Basel I. The increase in volatility is greaterthe lower the quality of the portfolio and thegreater the use of market-based measures ofcredit risk (such as yield spreads) relative to“through-the-cycle” measures (such as creditratings). In contrast, we find that for sovereignexposures the new rules could produce higher,but less volatile, minimum capital requirements.

If the increased risk sensitivity in Basel II con-tributes to changes in overall required capitalthat are unacceptable to the banks, they may tryto mitigate this effect by adjusting their lending(reducing it during periods of deterioratingcredit quality) or by adjusting the quality distri-bution of their portfolios (shifting towardshigher-quality assets). Thus, the actual observedvolatility in capital may not change significantlyonce Basel II is implemented, but perhaps onlybecause banks are adjusting their loan portfoli-os accordingly. This is precisely the cyclical be-haviour that has raised some concern.

Several factors may mitigate the potential impactof Basel II on the cyclical behaviour of capital,however. Cyclicality is already present in thebanking system. Indeed, the volatility of actualbank capital over the 1984–2003 period wasalready comparatively high relative to our base-case scenario and most of the alternatives exam-ined, suggesting that non-regulatory phenomenaare also important factors influencing volatilityin bank capital.

64

Our analysis shows that an important consider-ation is precisely how banks choose to calculatetheir capital requirements, which will also beinfluenced by accounting and tax regimes thatvary across countries. Our expectation is thatthey would tend towards smoother measures ofcredit risk (such as credit ratings), althoughthese effectively reduce the short-term sensitivi-ty to changes in risk. Canadian banks are alsowell capitalized, and they may use this high levelof capital to create an effective buffer to absorbvolatility in required capital.

Eligible banks might be expected to opt for theIRB approach if it provides them with potentialefficiency gains (i.e., owing to lower requiredlevels of capital than under the Standardapproach). As suggested above, to offset theincreased volatility of minimum capital require-ments that arises from the IRB rules, banks maytend to maintain buffer stocks of capital, inwhich case, there may be little induced cyclicalityin lending via this channel. They may followthis strategy if the resulting level of capital,including the buffer, would be lower than underBasel I.

The analysis in this article focused on the impli-cations of Basel II’s first pillar, and implies thatbanks need to carefully assess which methodthey will use to calculate required capital in theIRB approach, as well as the implications for thedesired level of buffer capital. In practice, thelevel of capital actually held by banks will alsobe influenced by Basel II’s second and third pil-lars. This analysis emphasized the banking sys-tem’s corporate and sovereign portfolios, whichmake up about 35 per cent of total bank assetsand which have the greatest potential for cycli-cality in capital requirements. One would ex-pect the results to be less pronounced for thebanking system as a whole, because the capitalrequirements for the remaining 65 per cent ofbank assets are expected to be relatively stable.


References

Bank of Canada. 2004. Bank of Canada Bankingand Financial Statistics. Tables C1-11, K2.

Bank for International Settlements (BIS). 2002.Quantitative Impact Study 3, TechnicalGuidance. Basel Committee on BankingSupervision. Available at <http://www.bis.org/bcbs/qis/qis3tech.pdf>.

———. 2004. International Convergence of Capi-tal Mea-surement and Capital Standards: ARevised Framework. Basel Committee Pub-lications No. 107 (June). Available at<http://www.bis.org/publ/bcbs107.htm>.

French, G. 2004. “Estimating the CapitalImpact of Basel II in the United States.”U.S. Federal Deposit Insurance Corpora-tion.

Illing, M. and G. Paulin. 2004. “The New BaselAccord and the Cyclical Behaviour ofBank Capital.” Bank of Canada WorkingPaper No. 2004-30.

Kiesel, R., W. Perraudin, and A. Taylor. 2003.“The Structure of Credit Risk: SpreadVolatility and Ratings Transitions.”Bank of England Working Paper No. 131.Available at <http://www.defaultrisk.com/pp_other_23.htm>.

Moody’s. 2004. “Credit Trends.” Data subscrip-tions available at <http://www.moodys.com/credittrends>.

Standard & Poor’s (S&P). 2004. Data subscrip-tions available at <http://www.stand-ardandpoors.com>.

65


Pre-Bid Run-Ups Ahead of CanadianTakeovers: How Big Is the Problem?Michael R. King and Maksym Padalko*

his study addresses an important di-mension of capital market efficiency;namely, the degree of informationasymmetry among market participants

(Bauer 2004). Insider trading is defined as trad-ing by managers and board members in thestock of their own firms. Regulators in manycountries have adopted securities laws restrictingwhen and how these corporate insiders maytrade in these shares. These laws make it illegalfor insiders to trade while in possession of ma-terial, non-public information, or to share thisinformation selectively with other investors.Instead, companies are required to disclose ma-terial information through a press release sothat all investors have an equal opportunity totrade on this information. Despite argumentsthat suggest illegal insider trading is a victimlesscrime that promotes market efficiency and pro-vides an efficient means of compensatingmanagers, regulators have taken the view that itis harmful to public welfare (Bainbridge 2000).Models of information asymmetry suggest thatif investors believe that insiders systematicallytrade on material, non-public information, thiswill increase the rate of return demanded byless-informed investors, widen the bid-askspreads set by market makers, and reduce liqui-dity in secondary markets. These effects wouldraise the cost of capital for firms and ultimatelyhurt public welfare by reducing economicgrowth.

Scope

This paper investigates whether there is evidenceof illegal insider trading in Canada ahead of aspecific type of corporate event; namely, a take-over bid. We examine 420 takeover bids of

T

* This article summarizes a forthcoming Bank of Can-ada working paper.

publicly listed Canadian firms from 1985 to 2002.We determine whether there are any systematicprice and volume increases in the target firm’sshares ahead of the first public announcement(a pre-bid run-up). We document the pattern ofthese pre-bid run-ups and compare them withthe results from similar studies of U.S. takeovers.We propose a test to differentiate between com-peting explanations of run-ups based on the co-incidence of abnormal price movements andabnormal volume, and the timing of the pre-bidrun-up in relation to the first public announce-ment.

Methodology

Pre-bid run-ups ahead of a takeover announce-ment may be caused by information leakage asa result of insider trading, market anticipationby investors who correctly identify a potentialtakeover target prior to the announcement, orsome combination of both. We begin with theassumption that capital markets exhibit infor-mational efficiency; namely, that stock pricesincorporate all public and private informationabout a firm. As our null hypothesis, we proposethat pre-bid price run-ups reflect the market’santicipation of a takeover announcement. In-vestors anticipate that a given firm will be sub-ject to a takeover based on rumours in the press,an analysis of industry trends, or factors specificto a company, such as financial distress. Thismarket anticipation—whether accurate or not—becomes incorporated into prices through trades,leading to a run-up ahead of the first publicannouncement.

The alternative hypothesis is that pre-bid run-upsare caused by information leakage associatedwith insider trading. In this scenario, the increasein the stock price ahead of the announcement ofa takeover bid is caused by insiders who aretrading illegally to profit from the price jump

67

Research Summaries

when the takeover is announced. Studies ofactual cases of illegal insider trading supportthis view. These studies document that illegalinsider trades are accompanied by both abnor-mal price movements and abnormal tradingvolume in a stock (Cornell and Sirri 1992;Meulbroek 1992). Illegal insider trading typicallytakes place far ahead of the announcement,since insiders seek to avoid the period shortlybefore the announcement when regulatoryscrutiny is highest. We use these stylized facts toidentify illegal insider trading, consistent withthe detection algorithms used by regulatorswhen reviewing trading patterns after majorcorporate events. This approach cannot be usedto prove illegal insider trading, but it can beused to detect its presence or to suggest its ab-sence. The key point to note is that abnormalprice movements that are not accompanied byabnormal volume changes (or vice versa) wouldconstitute a rejection of this alternative hypo-thesis. Likewise, abnormal price movements orvolumes that occur shortly before the announce-ment are more likely to be caused by marketanticipation.

We conduct a standard event study to examineabnormal price movements and trading volumes(MacKinlay 1997). This approach involveschoosing an event—such as a takeover announ-cement—and looking at the behaviour of thestock before and after the event. The aim is todetermine how the event affected the stock bycomparing actual movements in stock priceswith changes that might have been expected ifthe event had not taken place. For each takeoverin our sample, we set the date of the announce-ment as day 0, and we calculate daily abnormalprice movements over the prior three months.We then calculate the average abnormal pricemovements across the 420 transactions for eachday in our event window, and we accumulatethese daily abnormal price movements oversome time horizon. Given that we expect noabnormal price movements in the absence of atakeover announcement, we test to see whetherthese average and cumulative abnormal pricemovements are statistically different from zerousing both a standard parametric z-test and anon-parametric, signed-rank test. We conduct asimilar analysis of trading volume using averageabnormal volume and cumulative average ab-normal volume for each of the 420 takeoverannouncements.

68

Summary of Findings

We find that both average and cumulative ab-normal price movements become positive andstatistically significant only shortly before thefirst public announcement (Chart 1). Across oursample, the average abnormal price movementon day 0 is 9.8 per cent, which captures the in-crease in the stock price on the day when thetakeover is announced. The magnitude and tim-ing of pre-bid run-ups for the Canadian sam-ple are very similar in magnitude to run-upsdocumented for U.S. takeovers, suggesting thatstock prices react in the same manner in bothcountries.

We divide our sample into various sub-samplesto investigate the impact on the run-ups of in-dustry membership and the time period whenthe takeover bid occurred. Previous studies sug-gest that a clustering of takeovers in one sectoror during one time period increases the abilityof the market to anticipate future potential take-overs. Our sample exhibits a high number oftakeovers in the natural resource sector, and aclustering of bids over a few key years. We hy-pothesize that the cumulative abnormal pricemovements for takeovers of natural resourcefirms should be higher than for non-resourcefirms that are more heterogeneous. Contrary toour expectations, the run-up for natural resourcefirms is almost half the comparable run-up fornon-resource firms. Additional analysis is neededto explain this result.

We also consider the impact of institutionalchanges on pre-bid run-ups. If illegal insidertrading is the source of pre-bid run-ups, increasedsupervision and enforcement, as well as advancesin technology should discourage this behaviourby making it easier to detect ex post. The re-sources devoted to monitoring and enforce-ment increased significantly in 1998, after theOntario Securities Commission became self-funded. At the same time, the TSX closed itstrading floor and moved all stocks to an elec-tronic trading system. Both changes lead us toexpect that pre-bid run-ups may be smallerpost-1997 than during the earlier period. In-stead, we find that both the pre-bid run-ups andthe price jump over the event window werelarger for takeovers announced after 1997. Thisfinding, together with the finding that moremedia rumours are observed over this period,suggests that market anticipation has increased,


Chart 1 Cumulative Abnormal Return andVolume for 420 Takeovers, 1985to 2002

% %

Cumulative averageabnormal volume

(left scale)

Cumulative averageabnormal return

(right scale)

Trading day relative to takeover

-60 -50 -40 -30 -20 -10 0 10 20

-5

0

5

10

15

20

25

30

35

40

-4

0

4

8

12

16

20

24

28

32

possibly because of improvements in markettransparency. This hypothesis will be tested infuture research.

To test whether pre-bid price run-ups are explainedby information leakage or market anticipation,we examine whether abnormal price movementsduring the pre-event window are accompaniedby abnormal trading volumes. A naïve compar-ison of the abnormal price movements duringthe pre-event window with the abnormal vol-umes on the same day suggests that there are al-most no cases, on average, when both wereobserved on the same day (Chart 1). A more for-mal test of the relationship is provided by run-ning panel regressions of abnormal volumes onabnormal price movements. Abnormal pricemovements are statistically associated with ab-normal volumes at the 99 per cent level, al-though the small size of the coefficient suggeststhat the relationship is not economically impor-tant. From these panel regressions, we concludethat abnormal price movements during our pre-event window are not importantly associatedwith abnormal volumes. We fail to reject thenull hypothesis, and conclude that pre-bid run-ups are caused by market anticipation, not by in-formation leakage as a result of illegal insidertrading.

Conclusion

We find evidence of pre-bid run-ups in a sampleof 420 Canadian takeovers, consistent with simi-lar studies of U.S. takeovers. In our study, pre-bidrun-ups occurred shortly before the first publicannouncement and were of comparable magni-tude to the run-ups ahead of U.S. takeovers. Thesize of price run-ups increased in our sample fordeals announced after 1997, during a periodwhen regulators devoted greater resources to themonitoring of markets and the enforcement ofinsider-trading regulations. Contrary to our ex-pectations, run-ups were lower for firms in thenatural resource sector, despite the clustering ofdeals in this sector.

Based on the pattern of run-ups, the absence ofabnormal trading volumes on days with abnor-mal price movements, and the timing of therun-up shortly before the announcement date,we conclude that pre-bid run-ups are consistentwith market anticipation and reject an explana-tion based on information leakage from illegalinsider trading. This study suggests that Canadian

69

Research Summaries

equity markets are efficient, and does not sup-port the view that Canada has a greater problemwith insider trading than the United States.

While this conclusion applies to the averagetakeover announcement in our sample of 420,we cannot dismiss the possibility of illegal in-sider trading in any of the individual takeoversin our sample. Likewise, this article has notexamined insider trading ahead of other impor-tant corporate events, such as earnings announ-cements, dividend changes, and bankruptcyannouncements. We leave these topics to futureresearch.

References

Bainbridge, S. 2000. “Insider Trading.” In Ency-clopedia of Law and Economics Vol. III.The Regulation of Contracts, 772–812.B. Bouckaert and G. DeGeest eds.(Cheltenham: Edward Elgar).

Bauer, G. 2004. “A Taxonomy of Capital MarketEfficiency.” Bank of Canada FinancialSystem Review (this issue).

Cornell, B. and E.R. Sirri. 1992. “The Reactionof Investors and Stock Prices to InsiderTrading.” The Journal of Finance 47: 1031–59.

MacKinlay, A.C. 1997. “Event Studies in Econo-mics and Finance.” Journal of EconomicLiterature 35: 13–39.

Meulbroek, L.K. 1992. “An Empirical Analysisof Illegal Insider Trading.” The Journal ofFinance 47: 1661–99.

70


Monetary Policy, Private Information, andInternational Stock Marketsby Gregory H. Bauer and Clara Vega*

Financial economists are uncertain as to the cause of common movements in interna-tional stock markets. Previous studies have found that only a small fraction of the move-ments can be attributed to public news about macroeconomic variables. This has ledsome to suggest that investor irrationality is responsible. In this article, we assesswhether investors possessing superior (private) information about future U.S. interestrates and stock market movements affect international stock market comovements.

hat drives the common variationsin international stock market move-ments? This is an important ques-tion for a number of reasons. First,

researchers have identified a large degree of riskreduction in equity portfolios from diversifyingfunds into international stocks. The more thatstock markets around the world move together,the smaller are the gains from diversification.Second, it is important to know whether the co-movements are driven by rational responses tonew information or by the over-reaction of onemarket to movements in another. Third, it is im-portant for the Bank of Canada to understandhow asset prices in a small, open economy areaffected by foreign information. In particular,an assessment of the behaviour of financialmarkets requires an understanding of how andwhat information is incorporated into prices.Finally, it is important for central bankers toknow how much of the movement is caused bychanges in monetary policy. This will, in turn,help explain how the cost of capital for domes-tic corporations is determined in global markets.

Background

Financial research provides a compelling answerto this question. Asset-pricing models show thatexpected stock returns vary in response to changesin risk-free interest rates, changes in expected fu-ture cash flows, and/or changes in the equity riskpremium.1 In a rational asset-pricing framework,

1. The equity risk premium is the extra return requiredon (risky) stocks above the return required on (lessrisky) bonds.

* This article summarizes a recently published Bank ofCanada working paper (Bauer and Vega 2004).

W
with integrated international stock markets,comovements in international stock returnswould be driven by news about macroeconomicvariables that affect cash flows, risk-free rates, orrisk premiums in many countries.2
But existing studies, using daily or monthly data,show that public news about macroeconomicvariables has a limited impact on internationalequity returns. For example, King, Sentana, andWadhwani (1994) construct a factor model ofmonthly returns in 16 national stock marketsand examine the influence of 10 key macro-economic variables.3 They conclude that thepublic news components of these observablevariables contribute little to variations in worldstock markets. Rather, there is a dominant un-observable (i.e., non-public) factor driving in-ternational returns. They interpret the commonfactor as an index of “investor sentiment,” sug-gesting a degree of market irrationality. Otherstudies also find that public news about macro-economic variables has little cross-market im-pact; e.g., Karolyi and Stulz (1996), Connollyand Wang (2003).

If public news about macroeconomic variablesis not responsible for the comovements, couldsome “market friction” be responsible? Onepotential friction is asymmetric information.Information is asymmetric when some investorshave superior (private) information about the

2. In a “rational” market, prices fully and accuratelyreflect all available information. Markets are “inte-grated” when there are no barriers to trade in finan-cial assets between countries.

3. In a factor model, the expected returns on a largenumber of stocks are explained by a much smallernumber of variables.

71

Research Summaries

returns on assets in their own country. The stan-dard view is that these “sophisticated” agentsare either insiders in a company or obtain theprivate information about the firm by analyzingpublic information in a superior manner. Whensophisticated agents trade, their private infor-mation is (partially) revealed to the market,causing revisions in asset prices. Trading basedon private information could thus be a poten-tial cause of the comovements in internationalstock returns if the agents had superior knowl-edge about the common macroeconomic fac-tors that price equities in many countries. Butthe economic origins of such private informa-tion remain unexplored. Indeed, Goodhart andO’Hara (1997) wonder, “in the internationalcontext, how could private information beexpected to have a global impact?”

One possible answer to this question is that so-phisticated investors could have superior infor-mation about future macroeconomic factorsthat will affect both U.S. equity prices and inter-est rates. This private information about U.S.macroeconomic factors would likely be usefulin making trades in stocks around the world.The large size of the U.S. economy and the linksbetween U.S. and foreign firms suggest that U.S.macroeconomic conditions are likely to have aglobal impact. Indeed, if international equitymarkets are integrated, then private informa-tion about U.S. factors will give informed agentssuperior knowledge about the global factorsthat price stocks in many countries (Albuquer-que, Bauer, and Schneider 2003). Thus, it islikely that the private information of sophisti-cated investors trading in (liquid) U.S. marketswill help explain the cross-section of interna-tional equity returns.4 This article therefore ad-dresses the question of whether both public andprivate news about U.S. macroeconomic factorscould help explain comovements in interna-tional stock markets.

Methodology

In our paper, we test the potential cross-marketeffect of trading on private information by lookingat information revealed in two U.S. markets.

4. Another source of such global private information isthe trading floors of large banks. Traders see customerorder flows in many markets around the world. Thissuperior knowledge of global demand conditionswould be useful for the firm’s proprietary traders.

72

The first is the Eurodollar futures market thattrades on the Chicago Mercantile Exchange. TheEurodollar futures contract is considered to bethe most liquid exchange-traded money-marketinstrument in the world. Traders use the marketto hedge against, and speculate on, futuremovements in the Eurodollar interest rate (therate on U.S.-dollar deposits in banks outside theUnited States). We look at holding-period re-turns on and trades in the six-month Eurodollarfutures contract.

The second market is the S&P 500 ExchangeTraded Fund (ETF), known by its acronym asthe SPDR (Spider), which began trading on theAMEX in 1993.5 This fund is designed to trackthe performance of the S&P 500 Index, abroad index of U.S. stocks. The SPDR is a veryliquid security; in mid-2003, the fund had overUS$37 billion in assets under management,with an average daily trading volume totallingUS$4 billion.

We examine the impact on foreign stock mar-kets of public and private news originating inthese two U.S. markets. We wish to observe theprices of foreign stocks at the same time that weobserve prices in the two U.S. markets. To dothis, we use foreign ETFs. Barclays Global FundAdvisors manage “iShares” that trade on theAMEX. They are baskets of foreign stocks puttogether to track the performance of foreignmarket indexes compiled by Morgan StanleyCapital International. We also use the ETFs of12 other countries (Germany, Japan, the UnitedKingdom, Switzerland, Canada, France, Nether-lands, Hong Kong, Spain, Sweden, Australia,and Italy). These countries were selected be-cause they have developed equity markets andETF data that are available over the sampleperiod.

5. ETFs are shares of a portfolio of stocks that trade con-tinuously on an exchange and are designed to closelytrack the performance of a specific index. Managers ofETFs may buy either all the stocks in the index or asample of stocks to track the index. Market partici-pants are able to create and redeem shares in an ETFwhen its market price differs from the value of itsunderlying index. This ability to “open” the funds atany time ensures that ETFs trade near their net assetvalue. Elton et al. (2002) conclude that the SPDRclosely tracks the S&P 500 Index, since the differencebetween the two is less than 1.8 basis points perannum.


Our analysis proceeds in two steps. In the first,we obtain public and private informationshocks in the two U.S. markets by adaptingtechniques from the microstructure literature(primarily Hasbrouck 1991). Our analysis relieson the use of high-frequency data to obtain bet-ter estimates of the impact of news on prices.We use regression analysis to remove the impactof short-run microstructure effects from thehigh-frequency price and trade data. The residu-als from these regressions represent (noisy mea-sures of) public and private information shocksoccurring in U.S. markets.

In the second step, we use a factor model of in-ternational equity returns to evaluate the effectsof the U.S. public and private informationshocks on foreign equity markets.6 We examinehow well these factors are able to explain co-movements in international stock markets overholding periods ranging from one-half hour outto one week.7 The effects of private informationwill be revealed if unanticipated trades in U.S.markets are significant for longer holding peri-ods (one day to one week). This is because un-anticipated trades are a combination of randomliquidity shocks plus private information. Li-quidity shocks have only a short-run effect onprices, while the effect of private information ispermanent.

Results

The analysis yields a number of interesting find-ings. The first stage of the analysis reveals thatsome agents have superior knowledge about fu-ture U.S. interest rates and aggregate equity mar-ket returns. This suggests that the old way ofviewing private information as a “firm-specific”phenomenon is not correct; sophisticated inves-tors can have private information about entiremarkets. This comes from their superior inter-pretation of public news.

In the second stage of the analysis, there isstrong evidence of information spillover acrossmarkets. Both private and public information

6. The factors are linear combinations of the public andprivate information shocks from U.S. money andequity markets.

7. Foreign stocks will also respond to news released intheir home markets. Thus, the approach does notmeasure the effects on asset prices of all trades basedon private information, but only a subset of thesetrades.

shocks revealed in U.S. markets are componentsof the factors that model the cross-section of in-ternational equity returns. Contrary to the earli-er literature, public information shocks do havean effect, because they are more precisely mea-sured in the microstructure data than they werein the daily or monthly data. Private informa-tion shocks are also a statistically significantpart of the factor. Sophisticated investors havean impact on global markets when their superi-or information is incorporated into internation-al equity returns. This trading based on privateinformation is partly responsible for the com-mon variation in the movements of interna-tional stock markets.

An interesting finding concerning monetarypolicy is that unanticipated interest rate changesmade by the U.S. Federal Reserve Board influ-ence foreign stock markets. A policy-drivenincrease in U.S. interest rates lowers foreignstock returns. Changes in U.S. interest rates thatare not associated with monetary policy arenot statistically significant. Thus, changes inU.S. interest rates affect the international costof equity capital only when they are associatedwith changes in monetary policy.

These shocks are quantitatively important. Forexample, a shock of one standard deviation toprivate information about future U.S. interestrates is equivalent to almost 25 per cent of thestandard deviation of the total factor drivingweekly returns. A similar shock to private infor-mation about U.S. equity markets is equivalentto 17 per cent of the standard deviation of thefactor.

Private information can originate in two ways.Sophisticated agents, such as hedge fundmanagers, can conduct “top-down” analyses,where they generate private information aboutmacroeconomic fundamentals from a superiorinterpretation of public information.8 The funda-mentals could be related to either the U.S. eco-nomy or foreign economies. In either case, withintegrated international markets, such informa-tion would be useful for capturing return varia-tions in many countries.

Alternatively, order flow in U.S. markets couldbe acting as a “bottom-up” aggregator of diffuse

8. A “top-down” fund manager is an individual whohas a well-developed view of the macroeconomy anduses this view to invest in many different sectors.

73

Research Summaries

private information. Evans and Lyons (2004)present a model of the foreign exchange marketwhere order flow aggregates the dispersed pri-vate information about productivity shocks intwo countries. They note that while productivityshocks would occur at the level of the firm, ag-gregate trades by agents in the country wouldgive a more precise estimate of the country’sproductivity shock for that period. They alsonote that agents’ trades could be aggregatinginformation about other variables at the microlevel, such as money demand. Our U.S. shockscan be interpreted as money-demand shocksand real shocks arising from firm-level informa-tion. Financial firms in the United States thatobserve a large cross-section of customer orderflows could extract such information and use itfor proprietary trading. Again, with integratedmarkets, such U.S. information shocks wouldhave an international effect.

Conclusion

The goal of this research is to deepen our under-standing of the links between movements in theprices of foreign assets and news (public andprivate) originating in U.S. money and equitymarkets. Our first contribution is to show thatsome agents have private information aboutfuture U.S. interest rates and about aggregatereturns in equity markets. Our second contributionis to show that this superior knowledge affectsequity markets abroad. This finding gets to thecore of Goodhart and O’Hara’s (1997) questionof how private information can have a globalimpact. Not only do we show that public andprivate information about U.S. interest ratesand aggregate equity markets predicts futuremovements in foreign equity markets, but wealso show that these are components of factorsthat are priced in the cross-section of interna-tional equities.

The analysis raises a number of additional ques-tions. Are other sources of private informationavailable to sophisticated investors? Whilemonetary shocks are important, there may be“real” shocks related to technology or produc-tivity that sophisticated investors observe. In ad-dition, who are the investors who obtain thisinformation? Is it solely American investorswho have superior knowledge about Americanmarkets? It is likely that sophisticated foreigninvestors—such as offshore hedge funds—could also obtain this information. Finally,

74

does this private information affect other assets,such as foreign exchange and fixed-income mar-kets? Answering all these questions requires fur-ther analysis.

References

Albuquerque, R., G.H. Bauer, and M. Schneider.2003. “Characterizing Asymmetric Infor-mation in International Equity Markets.”Mimeo, University of Rochester.

Bauer, G. and C. Vega. 2004. “Monetary Policy,Private Information, and InternationalStock Markets.” University of RochesterWorking Paper.

Connolly, R. and A.F. Wang. 2003. “Interna-tional Equity Market Comovements:Economic Fundamentals or Contagion?”Pacific-Basin Finance Journal 11: 23–43.

Elton, E. J., M.J. Gruber, G. Comer, and K. Li.2002. “Spiders: Where Are the Bugs?”Journal of Business 75: 453–72.

Evans, M. and R. Lyons. 2004. “A New MicroModel of Exchange Rate Dynamics.”National Bureau of Economic ResearchWorking Paper No. 10379.

Goodhart, A.E.C. and M. O’Hara. 1997. “HighFrequency Data in Financial Markets:Issues and Applications.” Journal ofEmpirical Finance 4: 73–114.

Hasbrouck, J. 1991. “Measuring the Informa-tion Content of Stock Trades.” The Journalof Finance 46: 179–207.

Karolyi, G.A. and R.M. Stulz. 1996. “Why DoMarkets Move Together? An Investigationof U.S.-Japan Stock Return Comovements.”The Journal of Finance 51: 951–86.

King, M., E. Sentana, and S. Wadhwani. 1994.“Volatility and Links between NationalStock Markets.” Econometrica 62: 901–33.

BANK OF CANADA Financial...Reducing financial inefficiencies can, in principle, lead to a better...

Documents

Transcript of BANK OF CANADA Financial...Reducing financial inefficiencies can, in principle, lead to a better...