Research Article Effects of the Interest Rate and Reserve...

Research ArticleEffects of the Interest Rate and Reserve RequirementRatio on Bank Risk in China A Panel Smooth TransitionRegression Approach

Zhongyuan Geng12 and Xue Zhai3

1School of Finance Zhejiang University of Finance and Economics Hangzhou 310018 China2Coordinated Innovation Centre ofWealthManagement and Quantitative Investment Zhejiang University of Finance and EconomicsHangzhou 310018 China3The Global Banking and Markets Hongkong and Shanghai Banking Corporation Limited Hong Kong

Correspondence should be addressed to Zhongyuan Geng zhygengzufeeducn

Received 27 March 2015 Accepted 24 June 2015

Academic Editor Rigoberto Medina

Copyright copy 2015 Z Geng and X ZhaiThis is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

This paper applies the Panel Smooth Transition Regression (PSTR) model to simulate the effects of the interest rate and reserverequirement ratio on bank risk in China The results reveal the nonlinearity embedded in the interest rate reserve requirementratio and bank risk nexus Both the interest rate and reserve requirement ratio exert a positive impact on bank risk for the lowregime and a negative impact for the high regimeThe interest rate performs a significant effect while the reserve requirement ratioshows an insignificant effect on bank risk on a statistical basis for both the high and low regimes

1 Introduction

Loose monetary conditions such as low interest rates oftenresult in excessive credit expansion which can largely explainthe financial imbalances and economic fluctuations Follow-ing the burst of the dotcom bubble many central bankspreferred a soft monetary policy and exerted a low interestrate over an extended period to ease the potential recessionsContinuous low interest rates can boost the increase in assetprices and securitized credit and push financial entities totake more risks [1] It seems that bank risk is thereforeincreased Although it is not the time to attribute this typeof monetary policy to the 2008 global financial crisis itmay have contributed to its build-up Thus more and moreacademic and practical debates are around the effect ofmonetary policy on bank risk which has become a centralissue off the back of 2008 global financial crisis [2 3]

The Chinese banking system contributes the most tothe financial system in China and risks exhibited by com-mercial banks are the biggest threat to the nationrsquos finan-cial stability The impact of monetary policy on bank riskis thus an essential issue to which should be paid great

attentionwhen establishing itsmacroprudentialmanagementframework Unlike advanced economies in which standardone-instrument (a policy interest rate) operating proceduredominates the monetary policy tools Peoplersquos Bank of China(PBC Chinarsquos central bank) makes a frequent adjustment onthe interest rate (a main price-based instrument) and thereserve requirement ratio (a main quantitative instrument)simultaneously to achieve its goals During 2007ndash2012 Peo-plersquos Bank of China had adjusted the RMB 1-year benchmarkdeposit rate and the reserve requirement ratio for seventeentimes (six times in 2007 four times in 2008 two times in2010 three times in 2011 and two times in 2012) and thirty-four times (ten times in 2007 nine times in 2008 six times in2010 seven times in 2011 and two times in 2012) respectivelywhich is quite uncommon in the international practiceTherecome some interesting questions does interest rate and thereserve requirement ratio have different effects on the bankriskWill it firm up the financial stability and price stability ifChina sticks to the frequent and simultaneous manipulationon the interest rate and the reserve requirement ratio (Aboutthe relationship between price stability and financial stabil-ity there are two conflicting viewpoints One is ldquosynergyrdquo

Hindawi Publishing CorporationDiscrete Dynamics in Nature and SocietyVolume 2015 Article ID 571384 8 pageshttpdxdoiorg1011552015571384

2 Discrete Dynamics in Nature and Society

viewpoint that monetary policy aiming at price stability willbe conducive to financial stability [4]The other is ldquotrade-offrdquoviewpoint thatmonetary policy aiming at price stability is notnecessarily helpful for financial stability and a trade-off rela-tionship exists between price stability and financial stability[5 6]) Is Chinarsquos experience in monetary policy worth learn-ing for other economic entities in tailoring their monetarypolicyThe answers of the above questions require a deep divein the empirical test of the effects of the interest rate and thereserve requirement ratio on the bank risk in China

A linear model has been the main focus of most researchon the effect of monetary policy on bank risk while from ourperspective monetary policy instrument delivers a nonlineareffect on bank risk due to the subjective and irrational prop-erty of bank risk behavior The bankrsquos risk appetite risk per-ception and risk decision-making behavior changes slowlygradually and continuously following the implementationof monetary policy Moreover different monetary policyinstruments have various impacts on bank risk in termsof different macro environments and bank characteristicswhich is subject to uncertainty thanks to the counteractingdeterminants as well Therefore the relationship betweenmonetary policy instruments and bank risk follows a non-linear path It is much more reasonable to utilize a nonlinearmodel to analyze the effects of monetary policy instrumentson bank risk in avoiding assumption errors and bias

Nonlinear theories and models have matured graduallysince the 1970s which sparks scholars in accepting the factthat nonlinear model can fit economic phenomena andeconomic laws in a better manner [7] Among multipleavailable nonlinear models regime-switching models are themost popular ones Common nonlinear regime-switchingmodels include the following three models Markov regime-switching (MRS) model threshold regression (TR) modeland smooth transition regression (STR) model MRS and TRmodels are based on the assumption that the transition fromone regime to anther is discrete which is inconsistent withthe reality in many cases and thus limits their applicationin practice Hansen [8] made an initial effort on introducingthreshold effects together with a panel threshold regression(PTR) model which assumes a jumping transition throughdifferent regimes In improving the practicability Gonzalezet al [9] developed the Panel Smooth Transition Regression(PSTR) model extending a smooth transition regression(STR) model to panel data heterogeneity across the panelmembers and over time [10] The merged PSTR version ofcombining both STR and panel data enables the transition toswitch between regimes over time as smooth as it can be

This paper utilizes a PSTRmodel to study the nonlinearitybetween monetary policy instruments (ie the interest rateand the reserve requirement ratio) and bank risk Our studydiffers from the previous literature in the following waysFirst in the PSTRmodel established for the analysis of effectsof the interest rate and the reserve requirement ratio onbank risk the empirical results indicate that the interest rateand the reserve requirement ratio have nonlinear impacts onbank risk Second the result demonstrates that the objectivefunction of the central bank should be nonlinear-orientedand bring in financial sectorThe standard textbook approach

adopts a linear-quadratic (LQ) framework in analyzing opti-mal monetary policy where the dynamic behavior of theeconomy is described as linear and the objective functionstressing the policy goals is quadratic Monetary policy isalways keeping a balance by seeking an optimal match pointat which the loss function isminimized and the squared valueof the inflation gap and the squared value of the output gapare comprised at the same time [11] Our empirical resultsgive evidence to the fact that monetary policy instrumentsproduce nonlinear effects on bank risk In that case thecentral bank should incorporate nonlinear elements and afinancial stability variable into its objective function Thirdwe try to differentiate the effects of the interest rate and thereserve requirement ratio on bank risks thereby providingcomprehensive policy guidance in Chinarsquos implementationAdditionally someuseful information can be dug up from theresult as well facilitating the policymakers of other countriesin designing their monetary policy

The remainder of the paper is organized as followsSection 2 reviews the literatures Section 3 sketches out thegeneral empirical model to be estimated and describes thedata Section 4 presents the empirical results and relatedcomments Section 5 provides the conclusive remarks

2 Literature Review

The effects of monetary policy on bank risk are a partof although distinguishable from the relationship betweenmonetary policy and financial stability which is detailed byOosterloo and de Haan [12] and is not discussed here Sofar there is very limited theoretical support and empiricalevidence about the effects ofmonetary policy on bank risk [1]

Advanced economies typically use a policy interest rate asthe monetary policy instrument So the studies on the effectsof monetary policy on bank risk concentrate on the effectsof interest rate on bank risk The theoretical research in theliterature suggests several channels that interest rate affectsbank risk [13 14] (1) ldquoAsset valuationrdquo channel A reductionin the interest rate boosts asset and collateral values which inturn can modify banksrsquo estimation of probabilities of defaultloss given default and volatilities and it incents banks totake on risk (2) ldquoSearch for yieldrdquo channel Low interest ratescause banksrsquo target revenue to decline which provokes banksto invest in high-margin and high-risk areas or financialinstruments (3) ldquoAsset substitutionrdquo channel The decline ininterest rates will lead to a low proportion of safe assets inthe bank assets portfolio Risk-neutral banks will increase thedemand for risky assets until a new equilibrium arises in theratio of safe assets and risky assets (4) ldquoConstant leveragerdquochannel Commercial banks target a constant leverage ratioLow interest rates will boost the assets prices Bank equitywill increase and banks will respond to the fall in leverageby increasing their demand for risky assets This reactionreinforces the initial boost to asset values and so on Theresult is a more fragile banking system that is more exposedto negative shocks to asset values and thus riskier (5)ldquoCentral bank communicationrdquo channel If the central bankhas transparent policy and credible commitment low interestrate is an implicit commitment that will induce collective

Discrete Dynamics in Nature and Society 3

moral hazard Low interest rates mean loose monetary andregulatory environment which stimulates banks to take onmore risk (6) ldquoAsset-liability mismatchrdquo channel Wheninterest rates are low banks can only absorb short-termdeposits The mismatch between short-term deposits andlong-term project finance tends to high leverage The moreleveraged the banks are the higher the risk of failure is (7)ldquoHabit formationrdquo channel If the interest rate is low investorstend to consume more and the expected credit spread ishigh Thus investors are willing and able to get more loansfrom the bank or invest in high-risk financial instrumentswhich results in higher bank risk In addition some studiessuggest that interest rates have an uncertain effect on bankrisk which depends on many factors that affect the mutuallycountervailing forces [1]The effect of changes in interest rateson bank riskmay change over time alongwith a change in thebanking system or a change in the characteristics of the bankitself [15]

Empirical research shows some conflicting findings Onesuch finding claims that low interest rates lead to an increasein bank risk and high interest rates can prevent its accumula-tion [16ndash19] while others [20 21] claim that the reverse is trueInterestingly Thakor [22] Jimenez et al [23] and MarthaLopez et al [24] document an uncertain effect of interestrates on bank risk Interest rates have a smaller impact on therisky assets of banks with higher capital but a larger effecton the banks with more off-balance business Certain bankscan react heterogeneously to interest rate changes Bankswith a high capital adequacy rate and income diversificationperform more radically in their risk-taking

Some observation can be noted from the literature citedabove First most of the previous studies have employed theordinary least squares and generalized least squares method-ology to establish a linear model to study the impact ofmonetary policy on bank risk in the context of cross-sectionalor time series Yener et al [25] studied the nonlinear effectsof monetary policy on bank risk by simply incorporating thequadratic term of an explanatory variable-credit expansioninto the linear regression equation without establishing acutting-edge nonlinear model Second few studies comparethe different effects of the interest rate and the reserverequirement ratio on bank risk In advanced economiesinterest rate is the major monetary policy instrument Thusthey focus primarily on the effect of interest rates on bankrisk In China some scholars concentrate on the effects of theinterest rate and the reserve requirement ratio on bank riskbut they all draw the same conclusion that both the interestrate and the reserve requirement ratio have a negative effecton bank risk [26 27]

3 Model Specification and Data

When the sample size is not sufficiently large the intro-duction of too many explanatory variables will result inthe decline in the degrees of freedom and multicollinearityTherefore this study will only concentrate on the impact ofmacroeconomic factors on bank risk and does not considereffects of the bank-level micro factors on bank risk Fromthe angle of monetary policy instruments we construct

the following PSTR model to study the effects of interestrate and reserve requirement ratio on the bank risk (aboutthe detailed methodology of PSTR model see Granger andTerasvirta [28] Terasvirta [29] Eitrheim and Terasvirta [30]Hansen [8] and Gonzalez et al [9]) Consider

EDF119894119905= 120583119894+1205730IR119894119905 +1205731119867119894119905 + (120573

1015840

0IR119894119905 +1205731015840

1119867119894119905)

sdot 119892 (119875119894119905 120574 119888) + 120576

119894119905

(1)

EDF119894119905= 120583119894+1205730RR119894119905 +1205731119867119894119905 + (120573

1015840

0RR119894119905 +1205731015840

1119867119894119905)

sdot 119892 (119875119894119905 120574 119888) + 120576

119894119905

(2)

where 119894 = 1 119873 119905 = 1 119879 and 119873 and 119879 denote thecross section and time-dimension of the panel respectivelyEDF119894119905= expected default frequency which is the dependent

variable 120583119894represents the fixed effects IR

119894119905= interest rate

RR119894119905

= reserve requirement ratio 119867119894119905

= real estate priceindex119875

119894119905= purchasingmanagersrsquo index a threshold variable

The 119892 is the transition functions normalized to be boundedbetween 0 and 1 (When the transition function equals 0 or1 the corresponding model is resp called low regime orhigh regimeThe values of transition function transit between0 and 1 smoothly which makes the model transit betweenlow regime and high regime smoothly) 120574 slope parameterdenotes the speed of transition from one regime to the other119888 the threshold parameters 120576 the residual term and 120573 theregression coefficients

Before carrying out the empirical analysis we shoulddiscuss the variables used and the dataset In view of theavailability of the data we use the quarterly (from 042007to 032012) data of thirteen Chinese listed banks (The dataof listed banks is from the 119860 share market rather than119867 share market We exclude Agricultural Bank of ChinaChina Everbright Bank and China Construction Bank Thereasons are as follows Agricultural Bank of China and ChinaEverbright Bank became a listed bank in 2010 and there isnot much data available In December 2011 the number oftotal shares 119860 shares and 119867-shares of China ConstructionBank is resp 25001 billion 9593 billion and 21483 billionThe ratio of 119860 shares to total shares is only 384 producingthe mismatch between 119860 share market value (too less) andliabilities (too more) This mismatch will result in the factthat calculated value of EDF cannot objectively reflect theexpected default frequency of China Construction Bank)Considering that the calculation of EDF requires bankrsquos stockreturns and market value data which can be present onlyafter the bank lists and that the Bank of Communicationsthe Industrial Bank the CITIC Bank the Bank of Ningbothe Bank of Nanjing and the Bank of Beijing became listedbanks in 2007 our sample starts from the fourth quarterof 2007 so as to obtain sample data as much as possibleThe sample contains three large commercial banks sevenjoint-stock commercial banks and three city commercialbanks Among those the large commercial banks includethe Industrial and Commercial Bank of China (ICBC) theBank of China (BOC) and the Bank of Communications(BOCOM) the joint-stock commercial banks include CITIC


Bank Huaxia Bank (HB) Pingan Bank (PB) China Mer-chants Bank (CMB) Shanghai Pudong Development Bank(SPDB) Industrial Bank (IB) and China Minsheng BankingCo (CMSB) the city commercial banks include Beijing Bank(BB) Nanjing Bank (NJB) and Ningbo Bank (NBB)

Table 1 provides descriptive statistics for the variablesused in the empirical analysis Table 2 reports correlationcoefficients between these variables According to GujaratiDamodar [31] if the zero-order correlation coefficient of tworegressors is over 08 the multicollinearity problem will besevere Correlations in our study are at acceptable levels asshown in Table 2

In what follows we analyze the choice of the dependentexplanatory transition and control variables

31 Bank Risk As for the dependent variablemeasuring bankrisk we chose the expected default frequency (EDF) Thisindicator has become a popular measure of bank soundnessin related empirical work on financial stability Theoreticallyaccording to the nature of the risk EDF that utilizes stockprice and earnings volatility to characterize the risk behaviourof the bank is undoubtedly the ideal choice [25] The reasonsare listed below First EDF is relatively objective because it iscalculated on the basis of stock transaction data and financialdata found in the financial statements of the listed banksSecond EDF is a dynamic index and can be updated basedon changes of stock transaction data and regularly publishedfinancial statements of listed banks Therefore EDF canreflect changes of bank risk over timeThird EDF overcomesthe bias caused by applying historical data to represent futuretrends EDF is calculated on the basis of real-time situationsof the stockmarket Changes of the yields andmarket value inthe stock market can reflect the bankrsquos performance marketexpectations and future trends

EDFs are the outcome of Moodyrsquos KMV model whichestablishes a functional relationship between distance todefault and the probability of default The EDF of a companyvaries over time reflecting the changing economic prosperityof the firm or its industry sector A detailed description ofthe mapping between the distance to default and the EDFmeasure can also refer to Crouhy et al [32]

We use the method of Brandimarte [33] to calculatethe EDF Results of EDF calculations by Matlab70 softwareare reported in Table 3 The risk-free interest rate needsto be used to calculate EDF and it is based on the dailyweighted average of the RMB 1-year benchmark depositrate (The following data is from RESSET Financial ResearchDatabase and RESSET is Chinarsquos leading provider of financialdatabases thirteen listed banksrsquo daily yield daily total marketvalue quarterly long-term liabilities and quarterly short-term liabilities that are used to calculate EDF broad moneyquarterly real estate price index RMB 1-year benchmarkdeposit rate The data of purchasing managerrsquos index (PMI)is derived from CEInet statistics database (CEI China Eco-nomic Information))

32 Monetary Policy Instruments About monetary policyinstruments we choose the reserve requirement ratio (amain quantitative instrument) and interest rate (a price-based

Table 1 Descriptive statistics

Variable Mean Standard deviation Minimum MaximumEDF 0011 0064 4351119864 minus 22 0810IR 0030 0007 0023 0041RR 0176 0024 0131 0215119875 52007 3144 42118 56504119867 103816 6186 91072 111510

Table 2 Correlations

IR RR 119875 119867

IR 1RR 1119875 minus0164 minus0344 1119867 minus0023 minus0547 0517 1

instrument) because they are two main policy instrumentsapplied by Peoplersquos Bank of China (PBC Chinarsquos centralbank)

We adopt the reserve requirement ratio for the largefinancial institutions (denoted by RR) with the reason thatlarge financial institutions are the majority of Chinese bank-ing industry and the reserve requirement ratios for thelarge small and medium-sized financial institutions have abasically uniform trend

The interest rate IR is denoted by the RMB 1-year bench-mark deposit rate rather than the RMB 1-year benchmarkloan rate because the former is mainly set by the PBC andthe latter can float greatly according to the will of commercialbanks and cannot fully be controlled by the PBC

Since the reserve requirement ratio and the RMB 1-yearbenchmark deposit rate may be changed for several timeswithin one quarter or one year the daily weighed means ofboth is used

33 Transition Variable The purchasing managerrsquos index(PMI) denoted as 119875 measures the macroeconomic boomor bust As an indicator of the economic health of themanufacturing sector purchasing managerrsquos index (PMI) iswell known andwidely used in theworldThePMI is based onthe key indicators new orders inventory levels productionsupplier deliveries and the employment environment A PMIof more than 50 represents expansion in business activitycompared with the previous month A reading under 50represents a contraction while a reading at 50 indicates nochange The PMI is usually released at the start of the monthmuch before most of the official data on industrial outputmanufacturing and GDP growth becomes available It istherefore considered as a good leading indicator of economicactivity It is reasonable for us to use the PMI with thereason that China is a big manufacturing country and theboom and bust of manufacturing can be representative of themacroeconomic cycle

We choose the PMI as a transition variable with theaim of studying the effect of monetary policy in differentmacroeconomic situations on bank risk


Table3ED

Fof

thirteenlistedbank

sinCh

ina

2007Q4

2008Q1

2008Q2

2008Q3

2008Q4

2009Q1

2009Q2

2009Q3

2009Q4

2010Q1

2010Q2

2010Q3

2010Q4

2011Q1

2011Q2

2011Q3

2011Q4

2012Q1

2012Q2

2012Q3

ICBC

0000171000011100002290003059000011965537119864minus07980119864minus10115119864minus08622119864minus14425119864minus18108119864minus12435119864minus22318119864minus05574119864minus17112119864minus12577119864minus13805119864minus16600119864minus19113119864minus20173119864minus15

BOC212119864minus05221119864minus0600002010004331000058905843119864minus05710119864minus08106119864minus07734119864minus12354119864minus17354119864minus09305119864minus16369119864minus12213119864minus17383119864minus13163119864minus09493119864minus08131119864minus07800119864minus17487119864minus09

BOCM000015000098400075450035141010998000402300003130000363259119864minus06157119864minus070000271854119864minus080000208760119864minus060001803012675038616042537081019016587

CITIC191119864minus060000171000208700077680005042782119864minus06798119864minus08413119864minus050000122365119864minus070000279294119864minus09147119864minus05112119864minus09316119864minus0800077950000103140119864minus05686119864minus05309119864minus07

HB

000192100081930077918013274011939000499100001580000929208119864minus05362119864minus05000206137119864minus0700011930000329950119864minus050000953000032581119864minus0600001780001853

PB00005250005862001441002067300695170005127662119864minus050000372865119864minus06183119864minus060000124292119864minus09409119864minus06148119864minus07413119864minus050000123431119864minus06349119864minus05460119864minus05882119864minus07

CMB155119864minus0500010960000829000817800264270001509318119864minus05701119864minus06725119864minus07896119864minus09146119864minus07228119864minus12202119864minus07624119864minus10159119864minus11274119864minus07346119864minus07683119864minus09184119864minus09113119864minus07

SPDB000013300102560010277003173011217000325600001040000476622119864minus07124119864minus07929119864minus06664119864minus10346119864minus07274119864minus09202119864minus07134119864minus05148119864minus06203119864minus06128119864minus07532119864minus06

IB964119864minus05000510500052710013240054480000711132119864minus050000691266119864minus06339119864minus070000402577119864minus07511119864minus05914119864minus06291119864minus060000197510119864minus05587119864minus05929119864minus05229119864minus06

CMSB155119864minus05000119100065870018314001467700001350000178378119864minus05387119864minus07464119864minus11999119864minus07313119864minus15517119864minus09799119864minus11951119864minus060000243113119864minus05204119864minus06156119864minus06292119864minus06

BB184119864minus0700001380002205000583500051621278119864minus06268119864minus070000278127119864minus08276119864minus08220119864minus05125119864minus09696119864minus07182119864minus10816119864minus10690119864minus06544119864minus06518119864minus06380119864minus08706119864minus12

NJB107119864minus07294119864minus05000048500019850003723637119864minus05222119864minus06739119864minus05849119864minus09341119864minus08379119864minus05463119864minus10759119864minus06163119864minus09157119864minus11102119864minus08282119864minus07172119864minus08257119864minus07191119864minus12

NBB872119864minus05000019000099900033840002692492119864minus05443119864minus08175119864minus05518119864minus06138119864minus070000157656119864minus09997119864minus07403119864minus08447119864minus12742119864minus08132119864minus09546119864minus10134119864minus05798119864minus10


34 Control Variable In recent years the real estate markethas been hot and real estate loans have accounted for a signif-icant portion of bank credit in China introducing potentialriskWe select the real estate price index represented by119867 asa control variable to reflect the effect of the real estate marketon the risk of banks

4 Results and Analysis

41 Linearity and No Remaining Nonlinearity Results Theresults of the linearity tests are presented in Table 4 andshow that the null hypotheses that model (1) and model(2) are both linear are rejected at the 5 significance levelfor the Wald test implying that the relationship betweeninterest rate reserve requirement ratio and bank risk isindeed nonlinear Table 5 presents the test for no remainingnonlinearity after assuming a two-regime model The resultsindicate that the null hypothesis cannot be rejected implyingthat model (1) andmodel (2) have both only one threshold ortwo regimesThis implies that there is only one threshold levelof interest rate or reserve requirement ratio which separatesthe low and highmoney supply regimes inmodel (1) ormodel(2)

42 Model Estimation Results We utilize the nonlinear leastsquaresmethod to estimate parameters Before the estimationof the parameter we should apply the grid search methodto determine the initial value of the transition speed (120574) andlocation parameters (119888) The higher the number of iterationsis the better the initial value is For accuracy and time-savingreason the number of iterations is set as 20000 Estimatedmodel (1) and model (2) parameters are presented in Table 6

421 Model (1) This is a two-regime PSTR model the tran-sition speed is positive and low and the location parameter(48748) is within the changing interval of transition variable(PMI) When the PMI is over 48748 the model graduallymoves towards the high regime with an increasing transitionvariable When the PMI is under 48748 the model graduallyfalls to a low regime with the decrease of the transitionvariable The effects of interest rate on bank risk transitsmoothly and gradually between the high and low regimeswith the change in the value of the transition variableWithinthe time interval of 20 quarters the PMI is solely under thelocation parameter in the fourth quarter of 2008 while in theremaining quarters the PMI exceeds the location parameterThis shows that PMI affects bank risk mainly in the highregime

Coefficient for the interest rate is statistically significantand positive for the low regime (1205730) and statistically signifi-cant and negative for the high regime (1205730+120573

1015840

0)This indicatesthat when PMI is under 48748 interest rate has a positiveeffect on bank risk and when PMI is above 48748 interestrate is negatively correlated with bank risk The probablereasons are as followsWhen PMI is under 48748 the centralbank will reduce the interest rate to stimulate the economywhichmeans there is a down economy In this case the banksare cautious and dare not be involved in the highly risky field

Table 4 Linearity tests

Test Statistic 119875 valueModel (1) Lagrange multiplier-Wald 15221 0048Model (2) Lagrange multiplier-Wald 19489 0026Note1198670 linear model1198671 PSTR model with at least one threshold

Table 5 Tests of no remaining nonlinearity (test for the number ofregimes)

Test Statistic 119875 valueModel (1) Lagrange multiplier-Wald 0017 0992Model (2) Lagrange multiplier-Wald 0047 0977Note1198670 PSTR with one threshold1198671 PSTR with at least two thresholds

Table 6 PSTR models estimation Dependent variable EDF

Parameters Model (1) Model (2)1205730

6075lowastlowastlowast (2340) 3480 (1346)1205731

minus0003 (minus1587) minus0006 (minus1220)120573

1015840

0minus6089lowastlowastlowast (minus2164) minus3595 (minus1284)

120573

1015840

10002lowastlowast (1771) 0006 (1202)

120574 0483 0515119888 48748 47152RSS 0827 0835Note the values in parentheses are 119905-statistics lowastlowastlowastlowastlowast denote significanceat the 1 and 10 levels respectively

or business and the bank risk is low PMI above 48748 showsthat there will be an up economy the central bank tends toincrease the interest rate to suppress the probable economicoverheating In our sample period RMB 1-year benchmarkdeposit rate and ROE (Return on Equity) of the banks arepositively correlated (their correlation coefficient is 0101) Inthis case RMB 1-year benchmark deposit rate is high and thebanks also have a high profit Thus the banks do not takemore risk to make a lot of profit Therefore bank risk is low

422 Model (2) This is also a two-regime PSTR modelthe transition speed is positive and low and the locationparameter (47152) is within the changing interval to tran-sition variable (PMI) When the PMI is over 47152 themodel gradually moves closer to the high regime with anincreasing transition variable When the PMI is under 47152the model gradually falls to a low regime with the decrease ofthe transition variable The effects of the reserve requirementratio on bank risk transit smoothly and gradually betweenthe high and low regimes with the change in the value of thetransition variable Within the time interval of 20 quartersthe PMI is under the location parameter only in the fourthquarter of 2008 while in the remaining quarters the PMIexceeds the location parameter This shows that the reserverequirement ratio affects bank riskmainly in the high regime

Coefficient for the reserve requirement ratio is positive forthe low regime (1205730) and negative for the high regime (1205730+120573

1015840

0)but is statistically insignificant in both regimes


To sum up if the transition variable is the PMI the inter-est rate has a significant effect while the reserve requirementratio has an insignificant effect on bank risk in China

5 Concluding Remarks

We adopted the Panel Smooth Transition Regression (PSTR)approach to analyse the effects of the interest rate and thereserve requirement ratio on bank risk empirically basedupon the Chinese bank quarterly data from 042007 to032012The outcome of the exercise evidenced the nonlinearnexus between monetary policy instruments and bank riskwhich was assumed by the utilized model Our findingsmanifest that the consecutive change in the value of the PMIthreshold level enables the impact of the interest rate andthe reserve of requirement ratio on bank risk to undertakea smooth and gradual transition from high to low regimeThe interest rate has a positive and statistically significanteffect on bank risk for the low regime and a negative andstatistically significant effect for the high regime The effectsof the reserve requirement ratio on bank risk are positiveand statistically insignificant in low regime and negative andstatistically insignificant in high regime

The empirical results deliver theoretical implication andpractical significance It should improve upon the standardtextbook approach to analyzing optimal monetary policy thelinear-quadratic (LQ) framework only focusing on outputand inflation rather than financial stability The reactionfunction of monetary authorities should seek a way toinvolve nonlinear effects as well as financial sector in themacroeconomic decision-making model

In achieving financial stability and monetary policyeffectiveness the monetary authorities should concentrateon the nonlinear effects of the interest rate on bank riskin both the high and low regimes of PMI Strengthenedmacroprudential supervision and active cooperation withregulatory authorities aid attention in bank risk monitoringand research to avoid any potential risk against financialimbalances Finally commercial banks should devise anenhanced early risk-warning system which is incorporatedwith the nonlinear effects of the interest rate on bank risk inboth the high and low regimes of the PMI

Our empirical results can make helpful suggestion topolicymakers in China and other countries in monetarypolicy formulation The empirical results mean the interestrate has a significant effect (positive in a low regime andnegative in a high regime)while the reserve requirement ratiohas an insignificant effect on bank risk in China Thereforeapplying the interest rate and the reserve requirement ratioon a frequent and controllable basis is flexible for Chinato realize financial and price stability For example whenthe economy is down the central bank can cut the interestrate or (and) the reserve requirement ratio to stimulate theeconomy and make the price level upward without beingafraid of the adverse effects of the interest rate and the reserverequirement ratio on the bank risk with the reason that theformer has a significant and positive effect and the latter hasan insignificant effect on bank risk When the economy is upthe central bank will increase the interest rate or (and) the

reserve requirement ratio to suppress the economy andmakethe price level downward without being afraid of the adverseeffects of the interest rate and the reserve requirement ratio onthe bank risk with the reason that the former has a significantand negative effect and the latter has an insignificant effect onbank risk

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The authors acknowledge the financial support of theNational Natural Science Foundation of China (Grant no71103048 and Grant no 71273224) and the State Schol-arship Fund of the China Scholarship Council (File no201203070348) Thanks are due to Shangwei Fan (Passportno G52596733) for helping the authors collect and sort thedata

References

[1] G De Nicolo G DellrsquoAriccia L Laeven and F ValencialdquoMonetary policy and bank risk takingrdquo IMF Staff PositionNotes SPN1009 International Monetary Fund Home 2010

[2] Z Chang ldquoLiquidity shocks monetary policy mistakes andfinancial crisismdasha reflection of the US financial crisisrdquo Journalof Financial Research vol 7 pp 18ndash33 2010 (Chinese)

[3] M Hongxia and S Xuefen ldquoAcademic debate on the financialcrisis and monetary policyrdquo Economic Perspectives vol 8 pp119ndash124 2010 (Chinese)

[4] A Schwartz ldquoSystemic risk and themacroeconomyrdquo in BankingFinancial Markets and Systemic Risk G Kaufman Ed vol 7 ofResearch in Financial Services Private and Public Policy pp 19ndash30 JAI Press 1995

[5] C Borio and P Lowe ldquoAsset prices financial and monetarystability exploring the nexusrdquo BIS Working Papers 114 BIS2002

[6] G J Schinasi ldquoResponsibility of central banks for stability infinancial marketsrdquo IMFWorking Paper WP03121 2003

[7] Z Chang ldquoApplication of nonlinear dynamics in the field ofmacroeconomics a surveyrdquo Economic Research Journal vol 9pp 117ndash128 2006 (Chinese)

[8] B E Hansen ldquoThreshold effects in non-dynamic panels esti-mation testing and inferencerdquo Journal of Econometrics vol 93no 2 pp 345ndash368 1999

[9] A Gonzalez T Terasvirta and D van Dijk ldquoPanel smoothtransition regression modelsrdquo Research Paper 165 QuantitativeFinance Research Centre University of Technology SidneyAustralia 2005

[10] T Chang and G Chiang ldquoRegime-switching effects of debt onreal GDP per capita the case of Latin American and Caribbeancountriesrdquo Economic Modelling vol 28 no 6 pp 2404ndash24082011

[11] F S Mishkin ldquoMonetary policy flexibility risk managementand financial disruptionsrdquo Journal of Asian Economics vol 21no 3 pp 242ndash246 2010


[12] S Oosterloo and J de Haan ldquoCentral banks and financialstability a surveyrdquo Journal of Financial Stability vol 1 no 2 pp257ndash273 2005

[13] T Adrian and H S Shin ldquoMoney liquidity and monetarypolicyrdquo American Economic Review vol 99 no 2 pp 600ndash6052009

[14] C Borio and H Zhu ldquoCapital regulation risk-taking and mon-etary policy a missing link in the transmission mechanismrdquoJournal of Financial Stability vol 8 no 4 pp 236ndash251 2012

[15] G DellrsquoAriccia L Laeven and R Marquez ldquoReal interest ratesleverage and bank risk-takingrdquo Journal of EconomicTheory vol149 pp 65ndash99 2014

[16] A Maddaloni and J-L Peydro ldquoBank risk-taking securitiza-tion supervision and low interest rates evidence from theEuro-area and the US lending standardsrdquo Review of FinancialStudies vol 24 no 6 pp 2121ndash2165 2011

[17] V Ioannidou S Ongena and J-L Peydro ldquoMonetary policyrisk-taking and pricing evidence from a quasi-natural experi-mentrdquoThe Review of Finance vol 19 no 1 pp 95ndash114 2014

[18] M D Delis and G P Kouretas ldquoInterest rates and bank risk-takingrdquo Journal of Banking amp Finance vol 35 no 4 pp 840ndash855 2011

[19] Y Yu and W He ldquoEmpirical test of monetary policy creditquality and bank risk appetiterdquo Studies of International Financevol 12 pp 59ndash68 2011 (Chinese)

[20] M Lucchetta ldquoWhat do data say about monetary policy bankliquidity and bank risk takingrdquo Economic Notes vol 36 no 2pp 189ndash203 2007

[21] T Zhong and S Fang ldquoEmpirical analysis of monetary policymarket discipline and bank risk taking behaviourrdquo Journal ofShanghai University of Finance and Economics vol 11 pp 57ndash65 2011 (Chinese)

[22] A V Thakor ldquoCapital requirements monetary policy andaggregate bank lending theory and empirical evidencerdquo Journalof Finance vol 51 no 1 pp 279ndash324 1996

[23] G Jimenez S Ongena J-L Peydro and J Saurina ldquoHazardoustimes for monetary policy what do twenty-three million bankloans say about the effects of monetary policy on credit risk-takingrdquo Econometrica vol 82 no 2 pp 463ndash505 2014

[24] PMartha Lopez G Fernando Tenjo and S Hector Zarate ldquoTherisk-taking channel and monetary transmission mechanism inColombiardquo Borradores de Economia 616 Banco de la Republicade Colombia 2010

[25] A Yener L Gambacorta and D Marquez-Ibanez ldquoDoesmonetary policy affect bank risk-takingrdquo Working Paper 1166European Central Bank 2010

[26] J Shuxia and C Yuchan ldquoMonetary policy bank capital andrisk-takingrdquo Journal of Financial Research no 4 pp 1ndash15 2012(Chinese)

[27] F Yi Z Shengmin and X Xiaowen ldquoAn analysis of bearingbank risks in monetary policies on the coordination betweenthemonetary policy and themacro-prudential policyrdquoManage-ment World no 11 pp 9ndash19 2012 (Chinese)

[28] C Granger and T Terasvirta Modelling Nonlinear EconomicRelationships Oxford University Press New York NY USA1993

[29] T Terasvirta ldquoSpecification estimation and evaluation ofsmooth transition autoregressive modelsrdquo Journal of the Amer-ican Statistical Association vol 89 no 425 pp 208ndash218 1994

[30] Oslash Eitrheim and T Terasvirta ldquoTesting the adequacy of smoothtransition autoregressive modelsrdquo Journal of Econometrics vol74 no 1 pp 59ndash75 1996

[31] N Gujarati Damodar Basic Econometrics China Renmin Uni-versity Press Beijing China 3rd edition 2000

[32] M Crouhy D Galai and R Mark Risk Management McGraw-Hill New York NY USA 2001

[33] P Brandimarte Numerical Methods in Finance and EconomicsA MATLAB-Based Introduction JohnWiley amp Sons New YorkNY USA 2006

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

MathematicsJournal of


Mathematical Problems in Engineering

Hindawi Publishing Corporationhttpwwwhindawicom

Differential EquationsInternational Journal of

Volume 2014

Applied MathematicsJournal of


Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of


Mathematical PhysicsAdvances in

Complex AnalysisJournal of


OptimizationJournal of


CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of


Operations ResearchAdvances in

Journal of


Function Spaces

Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of Mathematics and Mathematical Sciences


The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014


Algebra

Discrete Dynamics in Nature and Society



Decision SciencesAdvances in

Discrete MathematicsJournal of


Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Stochastic AnalysisInternational Journal of


viewpoint that monetary policy aiming at price stability willbe conducive to financial stability [4]The other is ldquotrade-offrdquoviewpoint thatmonetary policy aiming at price stability is notnecessarily helpful for financial stability and a trade-off rela-tionship exists between price stability and financial stability[5 6]) Is Chinarsquos experience in monetary policy worth learn-ing for other economic entities in tailoring their monetarypolicyThe answers of the above questions require a deep divein the empirical test of the effects of the interest rate and thereserve requirement ratio on the bank risk in China

A linear model has been the main focus of most researchon the effect of monetary policy on bank risk while from ourperspective monetary policy instrument delivers a nonlineareffect on bank risk due to the subjective and irrational prop-erty of bank risk behavior The bankrsquos risk appetite risk per-ception and risk decision-making behavior changes slowlygradually and continuously following the implementationof monetary policy Moreover different monetary policyinstruments have various impacts on bank risk in termsof different macro environments and bank characteristicswhich is subject to uncertainty thanks to the counteractingdeterminants as well Therefore the relationship betweenmonetary policy instruments and bank risk follows a non-linear path It is much more reasonable to utilize a nonlinearmodel to analyze the effects of monetary policy instrumentson bank risk in avoiding assumption errors and bias

Nonlinear theories and models have matured graduallysince the 1970s which sparks scholars in accepting the factthat nonlinear model can fit economic phenomena andeconomic laws in a better manner [7] Among multipleavailable nonlinear models regime-switching models are themost popular ones Common nonlinear regime-switchingmodels include the following three models Markov regime-switching (MRS) model threshold regression (TR) modeland smooth transition regression (STR) model MRS and TRmodels are based on the assumption that the transition fromone regime to anther is discrete which is inconsistent withthe reality in many cases and thus limits their applicationin practice Hansen [8] made an initial effort on introducingthreshold effects together with a panel threshold regression(PTR) model which assumes a jumping transition throughdifferent regimes In improving the practicability Gonzalezet al [9] developed the Panel Smooth Transition Regression(PSTR) model extending a smooth transition regression(STR) model to panel data heterogeneity across the panelmembers and over time [10] The merged PSTR version ofcombining both STR and panel data enables the transition toswitch between regimes over time as smooth as it can be

This paper utilizes a PSTRmodel to study the nonlinearitybetween monetary policy instruments (ie the interest rateand the reserve requirement ratio) and bank risk Our studydiffers from the previous literature in the following waysFirst in the PSTRmodel established for the analysis of effectsof the interest rate and the reserve requirement ratio onbank risk the empirical results indicate that the interest rateand the reserve requirement ratio have nonlinear impacts onbank risk Second the result demonstrates that the objectivefunction of the central bank should be nonlinear-orientedand bring in financial sectorThe standard textbook approach

adopts a linear-quadratic (LQ) framework in analyzing opti-mal monetary policy where the dynamic behavior of theeconomy is described as linear and the objective functionstressing the policy goals is quadratic Monetary policy isalways keeping a balance by seeking an optimal match pointat which the loss function isminimized and the squared valueof the inflation gap and the squared value of the output gapare comprised at the same time [11] Our empirical resultsgive evidence to the fact that monetary policy instrumentsproduce nonlinear effects on bank risk In that case thecentral bank should incorporate nonlinear elements and afinancial stability variable into its objective function Thirdwe try to differentiate the effects of the interest rate and thereserve requirement ratio on bank risks thereby providingcomprehensive policy guidance in Chinarsquos implementationAdditionally someuseful information can be dug up from theresult as well facilitating the policymakers of other countriesin designing their monetary policy

The remainder of the paper is organized as followsSection 2 reviews the literatures Section 3 sketches out thegeneral empirical model to be estimated and describes thedata Section 4 presents the empirical results and relatedcomments Section 5 provides the conclusive remarks

2 Literature Review

The effects of monetary policy on bank risk are a partof although distinguishable from the relationship betweenmonetary policy and financial stability which is detailed byOosterloo and de Haan [12] and is not discussed here Sofar there is very limited theoretical support and empiricalevidence about the effects ofmonetary policy on bank risk [1]

Advanced economies typically use a policy interest rate asthe monetary policy instrument So the studies on the effectsof monetary policy on bank risk concentrate on the effectsof interest rate on bank risk The theoretical research in theliterature suggests several channels that interest rate affectsbank risk [13 14] (1) ldquoAsset valuationrdquo channel A reductionin the interest rate boosts asset and collateral values which inturn can modify banksrsquo estimation of probabilities of defaultloss given default and volatilities and it incents banks totake on risk (2) ldquoSearch for yieldrdquo channel Low interest ratescause banksrsquo target revenue to decline which provokes banksto invest in high-margin and high-risk areas or financialinstruments (3) ldquoAsset substitutionrdquo channel The decline ininterest rates will lead to a low proportion of safe assets inthe bank assets portfolio Risk-neutral banks will increase thedemand for risky assets until a new equilibrium arises in theratio of safe assets and risky assets (4) ldquoConstant leveragerdquochannel Commercial banks target a constant leverage ratioLow interest rates will boost the assets prices Bank equitywill increase and banks will respond to the fall in leverageby increasing their demand for risky assets This reactionreinforces the initial boost to asset values and so on Theresult is a more fragile banking system that is more exposedto negative shocks to asset values and thus riskier (5)ldquoCentral bank communicationrdquo channel If the central bankhas transparent policy and credible commitment low interestrate is an implicit commitment that will induce collective








EDF119894119905= 120583119894+1205730IR119894119905 +1205731119867119894119905 + (120573

1015840

0IR119894119905 +1205731015840

1119867119894119905)

sdot 119892 (119875119894119905 120574 119888) + 120576

119894119905

(1)

EDF119894119905= 120583119894+1205730RR119894119905 +1205731119867119894119905 + (120573

1015840

0RR119894119905 +1205731015840

1119867119894119905)

sdot 119892 (119875119894119905 120574 119888) + 120576

119894119905

(2)




RR119894119905

















IR RR 119875 119867









Table3ED

Fof

thirteenlistedbank

sinCh

ina

2007Q4

2008Q1

2008Q2

2008Q3

2008Q4

2009Q1

2009Q2

2009Q3

2009Q4

2010Q1

2010Q2

2010Q3

2010Q4

2011Q1

2011Q2

2011Q3

2011Q4

2012Q1

2012Q2

2012Q3

ICBC





HB

















1015840










1015840


120573

1015840

10002lowastlowast (1771) 0006 (1202)





1015840












Acknowledgments


References










































Volume 2014




Journal of











Journal of


Function Spaces






Algebra
















EDF119894119905= 120583119894+1205730IR119894119905 +1205731119867119894119905 + (120573

1015840

0IR119894119905 +1205731015840

1119867119894119905)

sdot 119892 (119875119894119905 120574 119888) + 120576

119894119905

(1)

EDF119894119905= 120583119894+1205730RR119894119905 +1205731119867119894119905 + (120573

1015840

0RR119894119905 +1205731015840

1119867119894119905)

sdot 119892 (119875119894119905 120574 119888) + 120576

119894119905

(2)




RR119894119905

















IR RR 119875 119867









Table3ED

Fof

thirteenlistedbank

sinCh

ina

2007Q4

2008Q1

2008Q2

2008Q3

2008Q4

2009Q1

2009Q2

2009Q3

2009Q4

2010Q1

2010Q2

2010Q3

2010Q4

2011Q1

2011Q2

2011Q3

2011Q4

2012Q1

2012Q2

2012Q3

ICBC





HB

















1015840










1015840


120573

1015840

10002lowastlowast (1771) 0006 (1202)





1015840












Acknowledgments


References










































Volume 2014




Journal of











Journal of


Function Spaces






Algebra




















IR RR 119875 119867









Table3ED

Fof

thirteenlistedbank

sinCh

ina

2007Q4

2008Q1

2008Q2

2008Q3

2008Q4

2009Q1

2009Q2

2009Q3

2009Q4

2010Q1

2010Q2

2010Q3

2010Q4

2011Q1

2011Q2

2011Q3

2011Q4

2012Q1

2012Q2

2012Q3

ICBC





HB

















1015840










1015840


120573

1015840

10002lowastlowast (1771) 0006 (1202)





1015840












Acknowledgments


References










































Volume 2014




Journal of











Journal of


Function Spaces






Algebra










Table3ED

Fof

thirteenlistedbank

sinCh

ina

2007Q4

2008Q1

2008Q2

2008Q3

2008Q4

2009Q1

2009Q2

2009Q3

2009Q4

2010Q1

2010Q2

2010Q3

2010Q4

2011Q1

2011Q2

2011Q3

2011Q4

2012Q1

2012Q2

2012Q3

ICBC





HB

















1015840










1015840


120573

1015840

10002lowastlowast (1771) 0006 (1202)





1015840












Acknowledgments


References










































Volume 2014




Journal of











Journal of


Function Spaces






Algebra
















1015840










1015840


120573

1015840

10002lowastlowast (1771) 0006 (1202)





1015840












Acknowledgments


References










































Volume 2014




Journal of











Journal of


Function Spaces






Algebra



















Acknowledgments


References










































Volume 2014




Journal of











Journal of


Function Spaces






Algebra







































Volume 2014




Journal of











Journal of


Function Spaces






Algebra









Research Article Effects of the Interest Rate and Reserve...

Documents

Transcript of Research Article Effects of the Interest Rate and Reserve...