The Return–Volatility Relation in Commodity Futures Markets
The Volatility Paradox: Tranquil Markets May Harbor … has declined across major asset classes and...
-
Upload
truongtuong -
Category
Documents
-
view
221 -
download
2
Transcript of The Volatility Paradox: Tranquil Markets May Harbor … has declined across major asset classes and...
Second Quarter 2017 (data as of June 30)
A review of financial market themes and developments
This monitor reflects the best interpretation of financial market developments and views of the staff of the Office of Financial Research (OFR). It does not
`n ecessarily reflect a consensus of market participants or official positions or policy of the OFR or the U.S. Department of the Treasury. Contributors: Meraj Allahrakha, Viktoria Baklanova, Danny Barth, Ted Berg, Jill Cetina, Dagmar Chiella, Arthur Fliegelman, Dasol Kim, Francis Martinez, John McDonough, Philip Monin, Mark Paddrik, Eric Parolin, and Daniel Stemp.
The Volatility Paradox: Tranquil Markets May Harbor Hidden Risks Financial markets were exceptionally calm in the second quarter by most measures. Only three times in the past 90 years has volatility been so low: twice during bull markets in the 1960s and 1990s, and once in the lead-up to the financial crisis of 2007-09 (see Figure 1). Is today’s low volatility a sign of calm or a threat to financial stability — or both? This edition of the OFR’s Financial Markets Monitor investigates the volatility paradox: the possibility that low volatility leads investors to behave in ways that make the financial system more fragile and prone to crisis. We analyze three channels through which a prolonged period of low market volatility may introduce financial stability risks: increased leverage, reduced hedging, and institutional investors’ use of risk-management models. We find some supportive evidence of these channels at work, but better data are needed to make definitive conclusions. Volatility alone is not a good indicator of impending financial stress.
Figure 1: S&P 500 Index 90-day Realized Volatility (percent)Volatility in U.S. equity markets is the lowest in decades
Note: Volatility is the standard deviation of daily returns over 90 days expressed as annualized percent change.Source: Bloomberg Finance L.P.
Key findings • Volatility for most asset classes across the world fell below historical averages during the second quarter. In some
cases, volatility is near all-time lows. Drivers of low volatility may include expectations that the long U.S. economic expansion and still-easy funding conditions will persist.
• Some institutional investors have adapted by increasing leverage and the use of yield-enhancing strategies. • Shocks could produce procyclical responses if market participants use measures of realized volatility to manage
the risk of their portfolios.
0
25
50
75
1928 1938 1948 1958 1968 1978 1988 1998 2008 2018
4.01 on April 23, 1964
6.45 on May 2, 1995
6.47 on May 14, 2017
7.12 on Jan. 2, 2007
OFR MARKETS MONITOR Second Quarter 2017 | 2
Volatility alone is a weak risk indicator. Volatility measures for most asset classes across global financial markets fell below their historical averages during the second quarter (see Figure 2). Some measures approached all-time lows (for example, see Figures 1 and 3), which may have been driven by expectations that the long U.S. economic expansion and still-easy funding conditions will persist. There are two types of volatility: realized and implied. Realized volatility reflects the historical price fluctuations of an asset. Implied volatility is forward-looking. It captures the market’s expectation offuture price fluctuations of an asset, derived from the options markets. When implied volatility exceeds realized volatility, the difference reflects the extra return investors demand to hold a security solely because it is volatile. This difference is known as the volatility risk premium. One of the most widely cited measures of implied volatility is the Chicago Board Options Exchange Volatility Index (VIX). The VIX is the 30-day implied volatility of options on the benchmark S&P 500 equity index. A low VIX doesn’t necessarily signal that severe financial stress is unlikely. For instance, the VIX provided no advance warning of extreme volatility in the months leading up to the financial crisis. Realized volatility of the S&P 500 index was often substantially higher than the VIX had predicted 30 days earlier (represented by the blue dots over the 45-degree line in Figure 4). The relationship between realized and implied volatility for other asset classes followed a similar pattern during the crisis. Market risks may seem low when volatility is low. However, low volatility may also serve as a catalyst for market participants to take more risk, thereby making the financial system more fragile. Thisphenomenon is known as the volatility paradox.
Low volatility directly incentivizes risk-taking. Lower volatility may contribute to greater leveragingand risk-taking through at least three channels. Thefirst channel is through changing asset-returncorrelations, which tend to increase when markets are volatile. Low correlations could entice investors to
-3
0
3
6
9
2007 2010 2013 2016
Figure 2: Realized Volatility by Asset Class (z-score)Volatility has declined across major asset classes and markets
Note: Realized volatility is the standard deviation of daily returns over 30 days, expressed as annualized percent change. U.S. equities are represented by the S&P 500 index. U.S. interest rates are the weighted average of the Treasury yield curve. Global currencies are based on weights from JPMVXY index. Global equities are MSCI All Countries World Excluding U.S. Index. Standardization uses data since Jan. 1, 1993.Sources: Bloomberg Finance L.P., OFR analysis
U.S. interest ratesU.S. equities
Global currenciesGlobal equities (ex-U.S.)
0
20
40
60
80
100
1990 1995 2000 2005 2010 2015
Figure 3: Chicago Board Options Exchange Volatility Index (VIX) (percent) Implied volatility on equities has fallen to near all-time lows
Note: Implied volatility is derived from options markets and is the expected standard deviation of daily returns over the next 30 days, expressed as annualized percent change.Source: Bloomberg Finance L.P.
9.31 on Dec. 22, 1993 9.75 on June 2, 2017
0
25
50
75
100
0 25 50 75 100
Figure 4: VIX and Realized Volatility of S&P 500 Index (percent)The VIX did not predict the global financial crisis
Note: Realized volatility is the standard deviation of daily returns over 30 days, expressed as annualized percent change. Sources: Bloomberg Finance L.P., OFR analysis
Global financial crisis(8/1/2008 to 11/1/2008)
VIX
Other dates(3/1/1993 to 5/31/2017)
Realizedvolatility
OFR MARKETS MONITOR Second Quarter 2017 | 3
accumulate risky exposures, believing they arediversified. Prolonged periods of low volatility may further decrease correlations, encouraging further risk-taking. This procyclical behavior increasesinvestors’ risk of loss from a systematic shock, when volatility spikes and asset-return correlations revert to historical levels. Some evidence exists that this channel may be at work in equity markets. Sector correlations have declined significantly during the past two years, while volatility has remained low (see Figure 5). Second, low volatility could encourage the use of other yield-enhancing strategies, such as selling deep out-of-the-money put options (those with a strike price substantially below current prices). Investors collect a premium from selling these options, but can be obligated to purchase the underlying assets if the price drops below the strike price. Investors who accumulate these risky exposures could be more likely to experience financial stress if prices sharply decline. Available data on investor portfolios are not sufficient to assess this channel adequately. Third, low volatility can directly incentivizeleveraging by lulling investors into underestimating the odds of a volatility spike. One measure of marketwide leverage is the ratio of margin debt to market capitalization. This measure is imperfect because it doesn’t account for other positions on investor balance sheets, including derivatives positions. Figure 6 uses margin debt balances and market capitalization data from the New York Stock Exchange. The ratio increased from 2002 to 2007 amid low volatility, declined after the crisis, and has been climbing since as volatility again reached long-term lows. Evidence also exists that some large investors are highly leveraged and, for that reason, may be susceptible to volatility events. For example, the top decile of macro and relative-value hedge funds has been leveraged about 15 times in recent quarters. These funds combined account for more than $800 billion in gross assets, about one-sixth of all hedge fund assets.
Low volatility could also disincentivize investor hedging.
0
25
50
75
Jan2005
Jan2007
Jan2009
Jan2011
Jan2013
Jan2015
Jan2017
0.00
0.25
0.50
0.75
1.00
Figure 5: 3-Month Moving Average of S&P 500 Sector Correlations, VIX Index (correlation, percent) Correlations between sectors have fallen amid low volatility
Note: S&P 500 sector pairwise monthly correlation; 3-month moving average.Sources: Bloomberg L.P., OFR analysis
Sector correlation (left)VIX index (right)
0
25
50
75
100
Jan2007
Jan2009
Jan2011
Jan2013
Jan2015
Jan2017
1
2
3
Jan2005
Figure 6: Margin Debt Balance over Market Capitalization and S&P 500 Index 30-day Realized Volatility (percent) Realized volatility has fallen as investors increased margin debt
Note: Values are the New York Stock Exchange (NYSE) market capitalization and margin debt balances of its members. Dealer margin debt balances may reflect positions on securities not listed on the NYSE. Realized volatility is the standard deviation of daily returns over 30 days expressed as annualized percent change.Sources: Haver Analytics, OFR analysis
Margin debt / market capitalization (left)S&P 500 index realized volatility (right)
OFR MARKETS MONITOR Second Quarter 2017 | 4
Another way investors may adapt to low volatility is by reducing their hedging of risky positions. This behavior was particularly relevant in recent years, when historically low interest rates pressured investors to reach for yield by holding more lower-rated fixed-income securities and more equities (see the OFR’s 2016 Financial Stability Report). OFR analysis of options trading suggests that investors have reduced their hedging of market exposure. Investor hedging activity is difficult to measure, although it can be captured to some extent using contracts outstanding in current-month SPY options. SPY is an exchange-traded fund that mirrors the benchmark S&P 500 equity index. Traders commonly sell SPY options to hedge equity market exposure. Options give investors the right, but not the obligation, to buy or sell a specific security at a specific strike price and time. A call option is a right to buy; a put is a right to sell. Options with a strike price near the current price of SPY are said to be “at the money.” Contracts with a strike price far from the current price are “away from the money.” These options are less likely to be held for hedging purposes and instead may represent yield-enhancing strategies. Investor hedging activity is captured through a hedging rate, calculated as the proportion of contracts on SPY options that is “at the money” versus “away from the money.” Hedging rates are currently lower on average than in the years immediately preceding the financial crisis (see Figure 7), suggesting a structural change in hedging activities after the crisis. However, the evidence is somewhat mixed. Considerable variation has occurred since 2010, and current levels appear to be higher relative to 2014 for both call and put hedging ratios. The absence of sharper measures of aggregate hedging activities makes drawing definitive conclusions difficult, though these hedging ratios at least suggest significant differences before and after the crisis. The Commodity Futures Trading Commission (CFTC) collects data on an alternative measure of hedging activity using positions of futures traders. CFTC data categorize hedge funds and other investors as “non-commercial,” or speculative, traders. As of May 2017, the net short position on VIX futures of non-commercial traders sat at levels larger than even before the crisis (see Figure 8). Common volatility strategies involve taking short positions in longer-dated contracts and long positions in shorter-dated contracts. Reduced
60
70
80
90
100
2005 2007 2009 2011 2013 2015 2017
Figure 7: SPY Options Held for Hedging Purposes (percent)Investors are less hedged compared to the pre-crisis period
Call hedging ratePut hedging rate
Sources: OptionsMetrics, OFR analysis
-150,000
-100,000
-50,000
0
50,000
2004 2006 2008 2010 2012 2014 2016
Figure 8: VIX Futures Noncommercial Net Total (contracts) Speculators increased short bets on VIX to the most since 2004
Sources: Bloomberg Finance L.P., Commodity Futures Trading Commission
Net position
-143,845 on June 20, 2017
OFR MARKETS MONITOR Second Quarter 2017 | 5
hedging in these strategies would imply shorting in the aggregate, consistent with Figure 8. However, establishing a direct link without more granular data is difficult. Together, these data suggest that some investors may have adapted to the low-volatility environment by reducing risk hedges and increasing speculative bets. Data limitations temper the findings to some extent, and leave opportunities for further analysis. With less hedging, these investors’ balance sheets may be less resilient to large volatility shocks when volatility returns to financial markets. Value-at-Risk models may give faulty signals in low-volatility markets. Low realized volatility can affect the behavior of banks, hedge funds, and other asset managers that use a risk management framework based on realized volatility, including some Value-at-Risk (VaR) measures. About 40 percent of large hedge funds, representing about 62 percent of gross hedge fund assets, regularly calculate VaR statistics for their funds, according to Form PF data collected by the Securities and Exchange Commission (SEC). VaR measures the risk of investments. It captures how much value investments might lose over a set time. Although VaR can be a valuable risk-management tool, overreliance on VaR when volatility is low could result in procyclical behavior that makes investors more vulnerable to volatility shocks if market conditions change abruptly. A decline in realized volatility can reduce a portfolio’s VaR, allowing market participants to increase position sizes without exceeding predefined VaR risk limits. The reverse is true when volatility rises. In that case, VaR-sensitive investors may be forced to simultaneously sell assets to get their portfolios below risk limits. A selloff induced by a VaR shock can become self-reinforcing as liquidity dries up and as deleveraging occurs. Some market observers believe VaR shocks contributed to selloffs in the Japanese government bond market in 2003 and in the U.S. Treasury market during the 2013 taper tantrum (see Figure 9). Long-term investors that are not sensitive to VaR, such as pension funds and insurance companies, may not step in and provide liquidity unless prices fall sharply.
0
25
50
75
100
125Day 0 Day
10Day20
Day30
Day40
Day50
Day60
Day70
Day80
Day90
Figure 9: Cumulative Yield Change in 10-year Government Bonds (basis points)VaR shocks may have deepened past selloffs in bond markets
Note: The vertical axis is inverted to reflect lower bond prices as yields increase. Horizontal axis is the number of days since the beginning of the sell-off period.Sources: Bloomberg Finance L.P., OFR analysis
Japanese Government Bond market (6/12/2003 to 9/10/2003)
U.S. Treasuries taper tantrum (5/2/2013 to 7/31/2013)
OFR MARKETS MONITOR Second Quarter 2017 | 6
Most large U.S. banks report data on the VaR of their trading books in quarterly 10-Q filings to the SEC. These data show a dramatic decline since 2010 in the VaR of banks’ trading books, without acommensurate decrease in the fair value of those trading books (see Figure 10). All else being equal, this change suggests that the reduction in VaR may reflect falling realized volatility rather than a decline in the size of banks’ trading books during the period. If volatility rises and banks aim to keep their VaR stable, the banks would need to shrink their trading books. Another possibility is that the declining VaR is evidence that banks have reduced the overallmarket risk in their portfolios, in part responding to additional regulatory oversight. A definitiveconclusion is difficult without detailed data on dealepositions. Targeting a specific level of volatility has recentlybecome an investment strategy. Many institutionainvestors now are holding so-called “volatility control funds” in their portfolios. Assets under management in variable annuity volatility control funds rose to $325 billion at the end of 2016 (see Figure 11). These funds make asset allocation decisions aimed atmaintaining a stable level of volatility for their whole portfolios. If volatility were to rise suddenly in a previously stable asset class, these funds may beforced to rebalance and sell assets. These investors’ activities could have a procyclical effect on asset prices and exaggerate volatility.
Note: G-SIB = Global systemically important bank
r Sources: Bank 10-Q forms filed with Securities and Exchange Commission, OFR analysis Figure 11: Variable Annuity Volatility Control Funds ($ billions,
count)
l Variable annuity volatility control funds have more than doubled
Conclusion Prolonged low market volatility may introduce financial stability risks through at least three channels. First, investors could respond by directly taking on more leverage and risk. Second, investors could reduce hedging activities. Third, institutional investors’ use of VaR or other risk-management models that have realized volatility as a key input could lead them to take more risk. A spike in volatility can result in outsized investor losses from sharp asset price changes. Data limitations hinder the ability to make definitive conclusions regarding the extent to which these channels are at work. However, the evidence is consistent with these channels operating and suggests the need for further analysis.
0.00
0.25
0.50
0.75
1.00
Mar2010
Mar2011
Mar2012
Mar2013
Mar2014
Mar2015
Mar2016
Mar2017
0
500
1,000
1,500
2,000
Figure 10: U.S. G-SIBs' Combined Trading Books ($ billions) Big banks' VaR has collapsed but portfolio size is little changed
Fair value of trading book (left)Trading book VaR (right)
0
100
200
300
400
Dec2010
Dec2011
Dec2012
Dec2013
Dec2014
Dec2015
Dec2016
0
100
200
300
in size
Source: Milliman Financial Risk Management LLC
Number of funds (left)Assets under management (right)
OFR MARKETS MONITOR Second Quarter 2017 | 7
Selected Global Asset Price Developments
1
Select U.S. Interest Rates
U.S. T reasury yields and yield curve
S ource: Bloomberg Finance L.P .
percent basis po ints2-year (Left Axis)10-year (Left Axis)10-year - 2-year spread (Right Axis)
Jul2016
O ct2016
Ja n2017
Ap r2017
Jul2017
0
1
2
3
75
100
125
150
175
200
225
250
U.S. T reasury t erm premium (basis point s)
Note : Adrian, Crump, & Moe nch mode lS ource: Bloomberg Finance L.P .
10-year 2-year
2010 2011 2012 2013 2014 2015 2016 2017-100
-50
0
50
100
150
200
250
300
Professional vs. market - implied U.S. inflat ion expect at ions(percent )
S ource: Bloomberg Finance L.P .
Survey o f Pro fessional Forecasters 10-year annual average5y5y forward breakeven rate CPI current
Sep2015
No v2015
Ja n2016
Ma r2016
Ma y2016
Jul2016
Sep2016
No v2016
Ja n2017
Ma r2017
Ma y2017
Jul2017
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Short -t erm market rat es (percent )
S ource: Bloomberg Finance L.P .
1-month Treasury bill 3-month LIBORGCF Treasury Repo
Jul2016
Sep2016
No v2016
Ja n2017
Ma r2017
Ma y2017
Jul2017
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
T hree-mont h Eurodollar fut ures (percent )
Note s: T he high and low points of the De c FOMC proje ctions are the maximumand minimum fore casts. T he re ctangle re pre se nts the me dian.S ource: Bloomberg Finance L.P .
Jun-30 May-17 Apr-17FOMC pro jections from June 2017
Jul2017
Ja n2018
Jul2018
Ja n2019
Jul2019
Ja n2020
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
Money market and policy int erest rat es (percent )
S ource: Bloomberg Finance L.P .
GCF Treasury repoFed funds effectiveInterest on excess reservesReverse repo facility
2011 2012 2013 2014 2015 2016 20170.0
0.25
0.5
0.75
1.0
1.25
1.5
2
U.S. Corporate Debt Markets
U.S. corporat e bond opt ion-adjust ed spreads(basis point s)
S ource: Haver Analytics
Investment grade (Left Axis) High yield (Right Axis)
Jul2016
O ct2016
Ja n2017
Ap r2017
Jul2017
100
120
140
160
180
200
220
240
200
400
600
800
1000
U.S. corporat e CDS indexes (basis point s)
Note : Five -ye ar maturity CDS Inde xS ource: Bloomberg Finance L.P .
Investment grade (Left Axis) High yield (Right Axis)
Jul2016
O ct2016
Ja n2017
Ap r2017
Jul2017
0
30
60
90
120
150
200
300
400
500
600
U.S. non-financial credit gross issuance ($ billions)
S ources: Dealogic, S tandard & Poor's Leveraged Commentary & Data
Investment Grade High Yield Leveraged Loans
Aug2015
O ct2015
D ec2015
Feb2016
Ap r2016
Jun2016
Aug2016
O ct2016
D ec2016
Feb2017
Ap r2017
Jun2017
0
50
100
150
U.S. corporat e credit fund flows ($ billions)
Note : Flows data are re le ase d with one -month lag.S ource: Haver Analytics
High yield Leveraged loans
Jul2016
O ct2016
Ja n2017
Ap r2017
Jul2017
-15
-10
-5
0
5
10
Leveraged loan issuance by use of proceeds (percent )
Note : Data for 2017 are ye ar-to-date as of January.S ources: S tandard & Poor's Leveraged Commentary & Data, OFR analysis
M&A/LBO Dividend/Buyback RefinancingOther
2000 2002 2004 2006 2008 2010 2012 2014 20160
10
20
30
40
50
60
70
80
90
100
Leveraged loan price act ivit y
Note s: S&P Le ve rage d Loan Inde x. Inde x 100=January 01, 2012.S ource: Bloomberg Finance L.P .
Ja n2012
Jul2012
Ja n2013
Jul2013
Ja n2014
Jul2014
Ja n2015
Jul2015
Ja n2016
Jul2016
Ja n2017
Jul2017
95
100
105
110
3
Primary and Secondary Mortgage Markets
Primary mort gage rat es (percent )
S ource: Bloomberg Finance L.P .
5-year/1-year adjustable rate 30-year fixed
Jul2016
O ct2016
Ja n2017
Ap r2017
Jul2017
2
3
4
5
MBS yield and opt ion-adjust ed spread t o U.S. T reasurysecurit ies
S ource: Bloomberg Finance L.P .
percent basis po ints
Current coupon (Left Axis) Spread (Right Axis)
Jul2016
O ct2016
Ja n2017
Ap r2017
Jul2017
1
2
3
4
40
60
80
100
120
30-year home mort gage fixed and jumbo rat es and spread
S ource: Bloomberg Finance L.P .
percent basis po ints30-year fixed (Left Axis)30-year jumbo (Left Axis)30-year jumbo-conforming spread (Right Axis)
Jul2016
O ct2016
Ja n2017
Ap r2017
Jul2017
3.0
3.5
4.0
4.5
5.0
5.5
0
20
40
60
80
100
120
Refinance and purchase loan applicat ions
Note : Inde x 100 = July 01, 2016.S ource: Bloomberg Finance L.P .
percent
Purchase Index (Left Axis) Gov Refi Index (Left Axis)Conv Refi Index (Left Axis)Refi % of to tal apps (Right Axis)
Jul2016
O ct2016
Ja n2017
Ap r2017
Jul2017
40
90
140
190
240
40
45
50
55
60
65
70
Convent ional mort gage severe delinquencies(percent , 90+ days lat e, seasonally adjust ed)
S ource Haver Analytics
Prime Subprime
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 20170
4
8
12
16
4
Equity Markets
Global equit y indices
Note : Inde x = July 01, 2016.S ource: Bloomberg Finance L.P .
S&P 500 MSCI EM Nikkei 225Shanghai Euro Stoxx 50
Jul2016
O ct2016
Ja n2017
Ap r2017
Jul2017
85
95
105
115
125
135
U.S. equit y indexes
Note : Inde x 100 = Jan 01, 2000.S ource: Bloomberg Finance L.P .
S&P 500 NASDAQ Russell 3000
2000 2003 2006 2009 2012 201520
60
100
140
180
S&P 500 sect or performance
Note : Inde x 100 = July 01, 2016.S ource: Bloomberg Finance L.P .
S&P 500 Financials Consumer staplesEnergy
Jul2016
O ct2016
Ja n2017
Ap r2017
Jul2017
80
90
100
110
120
130
140
S&P 500 price-t o-earnings and price-t o-book rat ios(mult iple)
S ource: Bloomberg Finance L.P .
Price-to -earnings (Left Axis) Price-to -book (Right Axis)
2012 2014 20169
13
17
21
1
2
3
4
U.S. equit y valuat ions: Shiller CAPE (rat io)
Note s: CAPE is the ratio of the monthly S&P 500 price le ve l to trailing te n-ye arave rage e arnings (inflation adjuste d).S ources: Haver Analytics, Robert S hiller
1920 1930 1940 1950 1960 1970 1980 1990 2000 20100
10
20
30
40
50
0
10
20
30
40
50
S&P 500 implied volat ilit y and opt ion skew(percent )
Note s: Option ske w is the diffe re nce be twe e n thre e -month implie d volatility ofout of the mone y puts and calls with strike s e qual distance from the spot price(+/- 20 pe rce nt). Highe r value s re fle ct gre ate r de mand for downside riskprote ction.S ource: Bloomberg Finance L.P .
VIX 80% - 120% Skew
Jul2016
O ct2016
Ja n2017
Ap r2017
Jul2017
0
10
20
30
5
Volatility
Implied volat ilit y by asset class (Z-score)
Note s: Z-score re pre se nts the distance from the ave rage , e xpre sse d instandard de viations. Standardization use s data going back to January 01, 1993.S ources: Bloomberg Finance L.P ., OFR analysis
U.S. equities (VIX) Global currencies (JPMVXYGL)U.S. Treasuries (MOVE) Average
2013 2014 2015 2016 2017-3
-2
-1
0
1
2
3
Realized volat ilit y by asset class (Z-score)
Note s: T hirty-day re alize d volatility. Equitie s base d on S&P 500 inde x, inte re strate s base d on we ighte d ave rage of T re asury yie ld curve , FX base d on we ightsfrom JPMVXY inde x. Standardization use s data going back to January 01, 1993.S ources: Bloomberg Finance L.P ., OFR analysis
Global FX U.S. interest rates U.S. equitiesAverage
2013 2014 2015 2016 2017-3
-2
-1
0
1
2
3
Global equit y indexes 1-mont h implied volat ilit y (percent )
S ource: Bloomberg Finance L.P .
Eurostoxx 50 SP500 German DaxUK FTSE 300 Japan Nikkei MSCI EM
Jul2015
Sep2015
No v2015
Ja n2016
Ma r2016
Ma y2016
Jul2016
Sep2016
No v2016
Ja n2017
Ma r2017
Ma y2017
Jul2017
0
10
20
30
40
50
60
Slopes of implied volat ilit y curves(basis point s)
Note s: Se ve n-day moving ave rage . Slope re pre se nts diffe re nce be twe e n one -ye ar and one -month maturitie s. G10 FX base d on we ights from De utsche Bank'sCVIX inde x.S ources: Bloomberg Finance L.P ., OFR analysis
G10 FX (Left Axis) S&P 500 (Left Axis)2-year USD Swaption Rate (Right Axis)10-year USD Swaption Rate (Right Axis)
Jul2016
Sep2016
No v2016
Ja n2017
Ma r2017
Ma y2017
Jul2017
-600
-400
-200
0
200
400
600
800
-30
-20
-10
0
10
20
30
40
Opt ion skew by asset class (z-score)
Note s: Option ske w is the diffe re nce be twe e n thre e -month implie d volatility ofout of the mone y puts and calls with strike s e qual distance from the spot price(+/- 10 pe rce nt). Highe r value s re fle ct gre ate r de mand for downside riskprote ction. Equitie s re pre se nts S&P500 inde x. Inte re st rate s re pre se ntwe ighte d ave rage ske w of T re asury future s curve . Curre ncie s re pre se nt dollarske w against major curre ncie s base d on JPMVXY inde x we ights. Z-scorestandardization use s data going back to January 01, 2006.S ources: Bloomberg Finance L.P ., OFR analysis
U.S. equities U.S. interest ratesGlobal currencies Average
Jul2016
O ct2016
Ja n2017
Ap r2017
Jul2017
-3
-2
-1
0
1
2
3
Volat ilit y of equit y volat ilit y
Note : VVIX Inde x me asure s the e xpe cte d volatility of the 30-day forward priceof the CBOE VIX Inde x.S ource: Bloomberg Finance L.P .
Jul2016
O ct2016
Ja n2017
Ap r2017
Jul2017
60
80
100
120
140
6
Advanced Economies
2-year sovereign bond yields (percent )
S ource: Bloomberg Finance L.P .
U.S. Germany U.K. Japan
Jul2016
O ct2016
Ja n2017
Ap r2017
Jul2017
-1.5
-1
-0.5
0
0.5
1
1.5
2
10-year sovereign bond yields (percent )
S ource: Bloomberg Finance L.P .
U.S. Germany U.K. Japan France
Jul2016
O ct2016
Ja n2017
Ap r2017
Jul2017
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Breakeven inflat ion (percent )
S ource: Bloomberg Finance L.P .
U.S. ten-year Germany ten-year U.K. ten-yearJapan ten-year
Jul2016
O ct2016
Ja n2017
Ap r2017
Jul2017
0
1
2
3
4
10-year euro area periphery government bond spreadsover German bunds (basis point s)
S ource: Bloomberg Finance L.P .
Italian govt (Left Axis) Spanish govt (Left Axis)Portuguese govt (Left Axis) Greek govt (Right Axis)
Jul2016
O ct2016
Ja n2017
Ap r2017
Jul2017
80
120
160
200
240
280
320
360
400
0
400
800
1200
1600
2000
Major currency indexes
Note s: Fore ign curre ncy incre ase s re pre se nt gre ate r stre ngth ve rsus the U.S.dollar. DXY incre ase s re pre se nt gre ate r stre ngth of the U.S. dollar ve rsus abaske t of major world curre ncie s. Inde x 100 = July 01, 2016.S ource: Bloomberg Finance L.P .
DXY (U.S. do llar) euro British poundJapanese yen Swiss franc
Jul2016
O ct2016
Ja n2017
Ap r2017
Jul2017
80
90
100
110
U.S. dollar long posit ioning vs. major currencies(net speculat ive posit ions, t housands of cont ract s)
Note s: Positive value s re pre se nt ne t U.S. dollar long positions. T he DollarInde x (DXY) is a future s contract base d on the U.S. dollar's value against abaske t of major world curre ncie s. T o e xpre ss a U.S. dollar long position in non-U.S. dollar contract, the contract must be shorte d.S ource: Bloomberg Finance L.P .
a
DXY (U.S. do llar) euro British poundJapanese yen Total
Jul2016
O ct2016
Ja n2017
Ap r2017
Jul2017
-100
0
100
200
300
400
7
Emerging Markets
Emerging market currencies(U.S. dollars per foreign currency unit )
Note s: Incre asing value s indicate stre ngthe ning ve rsus the U.S. dollar. Inde x100=July 01, 2016.S ource: Bloomberg Finance L.P .
RussiaChina o ffshoreBrazilEM currency Index
Jul2016
O ct2016
Ja n2017
Ap r2017
Jul2017
90
100
110
120
Emerging market sovereign debt yield
S ource: Bloomberg Finance L.P .
EM local currency
Jul2016
O ct2016
Ja n2017
Ap r2017
Jul2017
3.5
4.0
4.5
5.0
5.5
Equit y price indexes
Note s: T he US e quity inde x is the S&P 500 Inde x. T he Chine se e quity inde x isthe Shanghai Composite Inde x. T he De ve lope d Economie s inde x is the MSCIWorld Inde x and the Eme rging Marke ts inde x is the MSCI EM Inde x (both are inlocal te rms). Inde x 100 = July 01, 2016.S ource: Bloomberg L.P .
U.S. China Developed economiesEmerging markets
Jul2016
O ct2016
Ja n2017
Ap r2017
Jul2017
80
90
100
110
120
130
1-mont h realized emerging market s volat ilit y (percent )
Note s: Re alize d volatility is the annualize d standard de viation. Hard curre ncysove re ign de bt base d on the J.P. Morgan Eme rging Bonds - Global Price Inde xand curre ncie s base d on a we ighte d ave rage of EM curre ncy re turns against thedollar using we ights from J.P. Morgan VXY-EM curre ncy volatility inde x.S ources: Bloomberg L.P ., OFR analysis
EM sovereign hard currency debt EM currencies
2010 2011 2012 2013 2014 2015 2016 20170
10
20
IIF port folio f lows t o emerging market s ($ billion)
Note s: Data re pre se nt the Institute of Inte rnational Finance 's monthly e stimate sof non-re side nt flows into thirty EM countrie s. Data for late st obse rvations arede rive d from IIF's e mpirical e stimate s using data from a smalle r subse t ofcountrie s, ne t issuance , and othe r financial marke t indicators.S ource: Bloomberg
Debt flows Equity flows
2010 2011 2012 2013 2014 2015 2016 2017-30
-20
-10
0
10
20
30
40
50
China's Foreign Exchange Reserves ($ t rillion)
S ource: Bloomberg
FX reserves
Ja n2006
Ja n2007
Ja n2008
Ja n2009
Ja n2010
Ja n2011
Ja n2012
Ja n2013
Ja n2014
Ja n2015
Ja n2016
Ja n2017
$0.5
$1
$1.5
$2
$2.5
$3
$3.5
$4
$4.5
8
Commodit ies
Major commodit ies prices
Note s: Inde x 100 = January 01, 2010S ource: Bloomberg Finance L.P .
Bloomberg commodities indexCrude o il front month (Brent) Gold front month
2014 2015 2016 20170
25
50
75
100
125
Crude oil
Note : WT I and Bre nt are front-month contracts.S ource: Bloomberg Finance L.P .
$/Barrel Million BarrelsWTI (Left Axis)Brent (Left Axis)U.S. inventories (Right Axis)
2014 2015 2016 201720
50
80
110
140
300
350
400
450
500
550
Oil and nat ural gas fut ures curves
Note : Data as of July 05, 2017.S ources: Bloomberg Finance L.P ., OFR analysis
$/barrel $/mmbtu
Brent (Left Axis) Natural gas (Right Axis)
2M 6M 12M 24M 60M45
50
55
60
2.0
2.5
3.0
3.5
4.0
Oil supply and demand fact ors
Note : Global production and consumption are e stimate s by the Inte rnationalEne rgy Age ncy.S ource: Bloomberg Finance L.P .
Million barrels per dayGlobal production (Left Axis)Global consumption (Left Axis)U.S. rig count (Right Axis)
2014 2015 2016 201785
90
95
100
105
110
115
120
200
400
600
800
1000
1200
1400
1600
1800
Speculat ive fut ures posit ioning(t housands of cont ract s)
Note s: Positive value s re pre se nt ne t long positions. Ne gative value s re pre se ntne t short positions.S ource: Bloomberg Finance L.P .
Brent (Left Axis) Gold (Left Axis)Copper (Left Axis) Steel (Right Axis)
Jul2015
Ja n2016
Jul2016
Ja n2017
Jul2017
-140
-70
0
70
140
210
280
350
-20
-10
0
10
20
30
40
50
Met als spot price indexes
Note : Inde x 100 = January 01, 2010.S ource: Bloomberg Finance L.P .
Copper Steel Precious metals
Jul2015
Ja n2016
Jul2016
Ja n2017
Jul2017
0
50
100
150
200