Chapter 52 Computing Skills in Forecasting for Liquidity Risk Management in the Indian Banking Industry

Rituparna Das

National Law University, India

ABSTRACT

Liquidity Risk Management (LRM) in the banking industry happens at two levels: (1) the Central Bank (i.e. the regulator) and (2) the commercial banks.

1. INTRODUCTION

1.1 Concept of ‘Liquidity’ for a Bank

The term ‘liquidity’ is defined for a bank as “the ability of a bank to fund increases in assets and meet obligations as they come due, without incurring unacceptable losses” by the Basel Committee on Banking Supervision (2008). In the above defi­nition, the term “without incurring unacceptable losses” tallies with the term “at a reasonable cost” in the definition of ‘liquidity’ provided by the Reserve Bank of India (2012). With the above

DOI: 10.4018/978-1-4666-6268-1.ch052

ability a bank manages the mismatches in the volumes between the maturing liabilities and the maturing assets at the end of different common terms to maturity. If such mismatch affects the cash availability in a bank for the depositors at any point of time it would lead to bank-run. In this context Sethuraman (2008) distinguished be­tween liquidity and solvency as follows. Solvency means a non-zero net worth.

.

1.2 Concept of ‘Liquidity’ for the Central Bank

For the Central Bank, the term ‘liquidity’ is signi­fied by Mitra and Abhilasha (2012) as “financial flows of various kinds, ranging from that originat­ing from the Central Bank to the overall existent financing available in the banking system” or more simply by the term ‘macroeconomic liquidity’. Here ‘liquidity’ is described to be determined by the net change in the bank reserves through the interaction among the various autonomous factors that drive liquidity and its management by the RBI through the impacts of the RBI actions and the autonomous factors such as the government financial flows, the foreign capital flows and the demand for currency on the reserves of the banks with the RBI as a result of the central banking functions of the RBI in the Indian Rupee market as well as the foreign currency markets apart from its functions as the monetary authority of the coun­try and the daily outstanding figure of liquidity adjustment facility though repo and reverse repo (LAF) is officially considered to be a measure of available funding in the system. The LAF is a process whereby the banks borrow from the RBI against the government security collateral, called ‘repo’ process. Thus the LAF, introduced in the middle of 2000, is a monetary policy tool as per tquantity based monetary targeting described by the Reserve Bank of India (2003), in the hand of the RBI to pump liquidity into the system.

The aforesaid monetary policy tools include the cash reserve ratio (CRR) and the open market operations (OMO). Higher the CRR, higher are the bank deposits with the RBI and vice versa.

2. BACKGROUND

Econometrics emerged as a distinct discipline in 1930s and gradually crossing the border of the social sciences entered the realm of computational finance and risk management in 1990s, where it churns laser generated data as also computer generated data apart from time series data like currency-exchange data. There are three broad categories of econometric analysis - (a) linear regression models, (b) time series models and (c) dynamic econometric methods. The time series models like the GARCH are widely used by the practitioners in the modeling of volatility. Practitioners also use analytical statistics, e.g. the probability distributions are used in computing the limits. In addition, a range of softwares based on Excel is used in constructing the yield curve.

Liquidity for a commercial bank means cash in hand, gold reserve, marketable securities and unencumbered government securities which are by large covered by the statutory liquidity ratio (SLR) requirement. In the forecasting of its liquidity requirement a bank needs computing skill in the case of the following components of the forecasting process: (i) forecasting the volumes of the liquid assets, the liquidity coverage ratio, the net stable funding ratio, the borrowing needs, the volumes of collateral, the premature deposit withdrawals, the variance analysis and (ii) the stress testing for the purpose of calculating the internal limits and developing a contingency funding plan. This is an exercise of analytical statistics involving fitting the field data with the appropriate prob­ability distribution functions and then generating samples. These samples include the values of the variables during both the periods of stability and stress.

Academics expressed the fear of superficial advices from the applied research and consultancy born out of blind application of software with an intention to quickly resolve the risk management issues in the banks.

Against the backdrop of intense liquidity crisis, wild volatility in the interest rates and spiraling inflation following one after another in the Indian financial system during the post-crisis period, this chapter would revisit the econometric models with the causality analysis used by the practitioners in the Central Bank and the commercial banks and also the way they are used.

3. MAIN FOCUS OF THE CHAPTER

3.1 Link between the RBI’s Liquidity Adjustment and the Call Money Market

For controlling the demand for liquidity the RBI has introduced marginal standing facility (MSF) whereby a bank may borrow from the RBI at a rate 100 basis points higher than the repo rate although sometimes the RBI accommodates the excess demand for liquidity by allowing a second LAF, e.g. at the financial year 2011-12 ending on 31st March 2012 the second LAF was scheduled after the time of the normal LAF on the same day. With the LAF the RBI maintains stability in the overnight call money market. Thus the LAF helps the RBI to replace the reserve targeting with the interest rate targeting. The LAF is also considered to be a narrow measure of money supply and sub­ject to some autonomous drivers of liquidity like the volume of currency with the public, the cash balance with the government and the foreign ex­change operations by the private businesses, which are normally constrained by various monetary policy tool parameters of the RBI. There is enough room in the post crisis period for hypothesizing a positive feedback from the overnight call rate to the repo volume, i.e.

On March 30, 2012 the call rate shot up by 324 basis points. In the last quarter of the finan­cial year the liquidity-injection by the RBI goes up. For example during the entire second week of February 2012, the RBI injected daily four digit figures of liquidity in billion INR. Pairwise Granger causality test as described by Hamilton (1994) with 1 day lag shows that these variables Granger cause each other. The result is given in Table 1.

If we assume that on the one hand the Central B ank responds to the c all rate of the previous night by manoeuvring the LAF and on the other hand a participant in the overnight call market responds to the tenor of LAF of the previous day, i.e. if LAF

Table 1. Pairwise Granger causality tests

has injection mood it will report multiple positive figures on a particular day or positive figures for a number of days successively. In terms ofHamilton (1994) if a particular observed series of daily call rate (DCR) Granger-causes net LAF (NLAF) one should use auto-regressive specification and test the null hypothesis H0: DCR does not Granger cause NLAF.

Table 2. Pairwise Granger causality tests

up to be little different. H0: DCR does not Granger cause NLAF is rejected at 5% F-Test here. The result is given at Table 2.

This means the RBI was not much concerned about the systemic liquidity conditions when financial year end is not nearing.

Comparison between the level graphs Figure 1 and Figure 2 of NLAF and DCR for the periods from June 1, 2011 to May 19, 2011 and from May 29, 2011 to October 30, 2012 respectively here reveals that in the vicinity of financial year end the peak is by and large higher than the preceding one but when the financial year end is a distant future, every peak is by and large lower than the preceding.

Table 3 furnished the result of cointegration exercise between NLAF and DCR during the sample period including the financial year end.

Table 3 demonstrates long term equilibrium and non-spurious relationship between above variables in line with Granger (1986) quoted by Gujarati (2003).

The LAF repo rate is the floor of the delivery of the excess reserves by the RBI to the banks. In presence of a continuing liquidity surplus the LAF causes emanation of a distorted interest rate structure across the economy. Similarly in pres­ence of a continuing liquidity crunch MSF rate works as the ceiling and helps the RBI keep the interest rate structure under its control. This is a part of the RBI’s liquidity management operations process of overseeing that the functioning and the flows of the financial as well as real markets are not affected by the systemic liquidity fluctuations.

Ferhani (2005) took note of liquidity man­agement by the Central Banks of the developed countries followed by their emerging economy counterparts with the monetary policy imple­mentation in 1970s as a measure of the direct control and later their shift towards the reserve money management. The direct control was lost because of the distribution of resources by the financial markets and increasing number of non bank intermediaries. In the interest of efficiency

Table 3. Pairwise Granger causality tests

*(**) denotes rejection of the hypothesis at 5%(1%) significance level L.R. test indicates 2 cointegrating equation(s) at 5% significance level

and growth the market forces were being allowed to allocate resources.

The link between liquidity management for a bank and the same for the Central Bank is the possibility that liquidity problem of one bank can have repercussions on other banks and hence may spread over the system.

3.2 Liquidity Risk Management Process in a Bank

The Central Bank of a country takes care of the in­terest of the depositors as also the lenders. Because the liquidity risk arises out of the potential asset liability volume mismatches across the maturity structure - as discussed above, the RBI circulates guidelines from time to time about how to manage such mismatches and also warns the banks from time to time whenever it feels such mismatches to be in an alarming magnitude and may lead to systemic liquidity risk, e.g. in the last week of January 2011 the RBI warned banks against the fast shortening deposit maturity vis-a-vis the lend­ing terms especially in the infrastructure lending. This risk is known to be the funding liquidity risk.

The LRM in a bank as per the RBI is the first and foremost component of asset liability manage­ment (ALM), for, even the cash equivalent assets held by the banks at the instance of the RBI as SLR like government securities and treasury bills may depreciate in a financial crisis. Hence liquidity risk is another form of market risk, where all market participants are willing to offload their portfolios or the secondary market is lacking requisite depth to absorb them. This is also called market liquidity risk. As guidelines on LRM for the ALM purpose, to begin with within Basel I frame, much before the 2007-08 crisis, the RBI suggested in 1999 the banks to report (a) the statement of structural liquidity having two components - (i) total out­flow in eight different time buckets and (ii) total inflows in these buckets and the statement of Gap between the rate sensitive assets (RSA) and the rate sensitive liabilities (RSL). A decrease in the market rates of interest may lead to premature closure of the RSA whereas a hike in the same may lead to the higher interest payments on the RSL. Both these phenomena may cause a surge in the outflows. In 2012, as a part of implementing the Basel III guidelines the RBI asks the banks to measure liquidity with the two approaches - the stock approach and the flow approach. The concept of ‘stock’ has been explained in contrast with the concept of ‘flow’ in Bank Management Literature by Vento and Ganga (2009). Balance sheet items like ‘Balances with banks and money at call and short notice’ and ‘advances’ are called the stock items in the sense their figures are enumerated for a particular financial year with a single numerical value for every variable whereas the items in the structural liquidity statement like ‘term deposit’ within the maturity bracket of 29 days to 3 months and ‘tier II bonds’ within the maturity bracket of over 1 year and up to 3 years in a format similar to different fund flow statements for different maturity buckets aggregated into a single one are called the flow items. These approaches emerged in the circumstance where the pre-crisis period liquidity risk models were reported by the same to have failed to track the tailed events. Prior to the Basel III, the flow approach calls for the structural liquidity statements as discussed above but bifurcated into two components - one for the domestic operations and another for the offshore operations. But in Basel III, the structural liquid­ity statements have three more components based on combined onshore and offshore operations, foreign country wise operations and consolidated operations including aforesaid four components.

At the Board level a bank need to declare the liquidity risk appetite of the bank, i.e. the magnitude of liquidity risk the bank is willing to bear. The Federal Reserve Bank of San Francisco has given an example of above magnitude - the proportion of debt obligation not funded yet.

3.3 The Liquidity Issue Regarding Fixed Income Products

In India banks hold in their investment portfolio treasury bills and long term dated government securities (G-sec) within and beyond the SLR requirement stipulated by the RBI. If the fair price (to be denoted by ‘price’ in this context) of a se­curity is considered to be proxy for liquidity, in a situation where the security is selling at premium, as the security approaches towards maturity, both the maturity and the price have to decrease and duration decreases and vice versa. Duration means the sensitivity of the price of the security towards fluctuation of the benchmark rate (Grandville, 2001). Longer the distance of the security from maturity, the larger is the duration value and vice versa because a larger number of coupons are due and not paying more than yield. As an example is picked up from CCIL (2012) the security 8.33% GOVT.STOCK 2026, traded with the maximum volume on November 30, 2012. Figure 3 here shows that other things remaining the same, as the maturity decreases over time, but price is increas­ing initially because of the lingering excess demand attributable to the liquidity premium theory and then decreases from January 13, 2013. The reader may herself check that the duration of the above security starts falling after remaining constant till the second week of July 2013 because of the combination of coupon and yield unique to itself other things remaining unchanged.

On the other hand the security 8.15% GOVT. STOCK 2022 experiences a decreasing price with a decreasing maturity because it is paying a less coupon than does the yield or alternative return - then the holders of this security like to dispose it in lieu of incurring interest loss for next nine years of remaining maturity (Figure 4 here). The reader may check that the price starts increasing in the third week of March 2013 other things remaining unchanged.

The former security is more liquid because it will be paying a higher coupon for a longer period. Here the term ‘duration’ may be interpreted as the period within which the dirty price of the security would be recovered by the buyer.

3.4 Liquidity Risk Management by the Central Bank

In a country like India, where the banks are increas­ingly involving the international operations, the Central Bank, as a part of the systemic liquidity management, needs to engage in the cross border coordination, interventions in the banks’ money market operations and the foreign exchange de­rivatives dealings by the banks, and adjustment in its liquidity operations. In India the RBI on behalf of the government de facto guarantees the bank debts, but does not buy the distressed assets against the suggestion from the International Monetary Fund (2010). The same also prescribed a systemic view of liquidity management instead of a bank-centric view.

Since some select simple and weighted addi­tion formulae of the diverse components of money supply are considered by academics as measure of liquidity also and liquidity crisis is deemed by the same as another version of monetary instabil­ity, the line of distinction between the systemic LRM policy and the monetary policy must be very thin. Though there was reportedly a hot but academically rich debate between the banking economists inside and outside the RBI during the pre-reform period around the applicability of bal­ancesheet approach vis-a-vis multiplier regarding the RBI’s arriving at a agreeable measure of the money supply, in the monetary policy operations the RBI stance was chronologically the reserve money targeting because of its perceived limited efficacy of money multiplier, after emergence of the new Keynesian ideas described by Goyal (2011), interest rate targeting and recently call rate targeting. The call rate targeting is keeping in view the systemic LRM also.

3.5 Forecasting Liquidity for a Bank

The literature on liquidity forecasting by a bank from academic perspective is in a nascent stage and yet to appear in public in its full fledged form. Every bank has their economists who perform the job of forecasting. Depending on the technological sophistication and human expertise level of the bank, the nature of forecasting and the softwares used varies from bank to bank because the choice of forecasting methodology is left to the banks in India - there is no guideline or regulation of the same.

To start from the scratch, let us consider the Annexure - i of the first alm guideline of the RBI circulated in February 1999. Liquidity forecasting for a bank means the forecasting of the variables appearing in the columns with the titles ‘INFLOWS’ and ‘OUTFLOWS’. This requires collection and compilation of daily data on these variables by the concerned bank. Since these data need not be public as per the extant financial dis­closure norms, for academics it is not possible to access the same. But periodical aggregate figures of these variables are published by the RBI.

As per the most general academic notion, forecasting methodology is of two types - (i) qualitative and (ii) quantitative. In dire paucity of quants in the bank, qualitative method is used based on the experience and the acumen of the senior executives. The quantitative forecasting of above variables comprises use of one or a combi­nation of the following frameworks:

1. Classical bi-variate or multi-variate regres­sion models involving time series data of policy/independent variables and target/ dependent variables,

2. Time series econometric models built up exclusively for time series data like auto regression, moving average, auto regressive moving average (ARMA), auto regressive integrated moving average (ARIMA), au­toregressive conditional heteroskedasticity (ARCH) and generalized autoregressive conditional heteroskedasticity (GARCH),

3. Hendry’s General to Specific Method and

4. Bayesian Framework.

These are all for detecting causality between the target variable on the one hand and on the other hand policy variables, own lagged values of the target variable or a combination of the two and then performing simulation to check whether the estimated causality reproduces the past behaviour of the target variable described in Pindyck and Rubinfeld (1998). Above exercises could be performed either in a spreadsheet with Analysis Value Pak or using software like Eviews® version 7 or TSP® version 4.5 with looking glass interface. Automaticity into the functions in a spreadsheet may be inserted into the above models. Above models have limitation that they do not forecast extreme or stress events. But the bankers need stress testing for different purposes like understanding the daily changes in 10 year government securities yield in the stock dealing desk or in the risk management desk quantifying market risk described in Das (2010).

After J P Morgan popularized value at risk (VaR), practitioners started ascertaining the prob­ability distribution of the target variable and gen­erate samples using the distribution parameters. There are three types of VaR:

1. The normal VaR,

2. The historical VaR, and

3. The simulated VaR.

The normal VaR is computed using the features of normal distribution based on a confidence inter­val in the daily change of a variable like premature withdrawal during the sample period of, say, 1000 business days, with probability, say, 99%. This gives an estimate of the maximum withdrawal amount during the sample period. The historical VaR is computed by arranging the sample data in an ascending or a descending sequencing, calcu­lating percentile of each figure and then spotting the one with, say, 99% or 99.5%. Suppose there are 1000 sample points. After arranging them in the above mentioned style, we need to locate the fifth largest value of the daily change and then need to multiply it with the exposure value in order to obtain 99% VaR. In order to generate the simulated VaR the first step is to fit the sample in fitting softwares like Easyfit® and run it. Check the probability distribution, note the values of the mean and the moments and using the same run several, say, 100 rounds of simulation in order to generate alternative samples. From each sample compute 99% VaR using both of previous two methods. Then choose the maximum of these. This is to remember that after fitting the data in above software the probability distribution often comes out to be other than the normal distribution like, the lognormal distribution.

Before feeding into any model, the time series data needs filtering from the outliers and the property of non-stationarity. Otherwise regres­sion parameters may not be significant or may be spurious. Non-stationarity most of the times is detected with the unit root tests and removed with the differencing of order 1, 2 etc. Any reli­able results from the regression analysis call for both the regressor and the regressand to be non-stationary or stationary of the same order. If regression between two variables of same order of integration, i.e. non-stationarity provides robust adjusted R2 and t values, these variables are said to be correlated. All the methods revolve around the least square method or alternatively the maximum likelihood method.

3.6 Forecasting Liquidity for the Central Bank

The Central Bank should be ready to prevent or take an immediate action against the eruption of any liquidity crisis, for, the liquidity crisis may convert into a solvency crisis. In the process, as described by the Money Market Contact Group (2009), the Central Bank needs to oversee the pay­ments system and capture the common liquidity risk factors across banks.

Forecasting the macroeconomic liquidity by the Central Bank follows definition and measure of the same. In India the first hint of the link be­tween the concepts of liquidity vis-a-vis money dates back to Bhole (1987). There was elaboration on the same in the Working Group on Money Supply (1988). Keeping in view the property of ‘moneyness’ or liquidity of certain instruments which compete with money the Working Group on Money Supply (1998) reduced the difference between the broadest measures of money and liquidity having compiled the liquidity aggregates by including the liabilities of the post office savings banks and the non-bank financial institutions with the liabilities of a specified number of financial institutions, and excluding certain non-money instruments from the measure of broad money. Since the above liquidity measure is subject to a number of autonomous factors and the repo rate is a policy variable which the RBI cannot change daily, the RBI needs to operationalize the liquidity mismanagement by steering the call rate. At this juncture a number of studies during the period 2009-2011 tried to explore the genesis of liquidity conditions in the factors driving violations of the following four arbitrage relationships -

1. The covered Interest Parity,

2. The CDS-bond basis for non-bank corporations;

3. The on-the-run U.S. Treasuries versus the off-the-run U.S. Treasuries, and

4. The interest rate swap spread and the con­structed systemic liquidity risk index (SLRI) based on these violations- described by Severo (2011).

His findings which are applicable to India are the SLRI varies directly with the net stable funding ratio and the SLRI affects inversely the volatility of the bank stocks. In India a study by a group of RBI economists - Mishra, Mohan and Singh (2012) made a similar attempt after a year to construct the systemic liquidity index (SLI).

They reported that (i) the SLI encompassed the major components of the financial markets like the banking sector in the form of a difference between call and repo rates, the corporate sector in the form of a difference between the commercial paper rate and the certificate of deposit rate, the foreign exchange market in the form of the implied deposit rate and the expectations about the liquid­ity conditions in the form of the steepness of the overnight index swap curve, (ii) the variance-equal method is found to be the most appropriate relative to other methods in the Indian context, (iii) the liquidity condition has impact on return on bank stocks, (iv) the LAF has impact on the SLI and (v) the SLI varies directly with the credit growth vis-a-vis the deposit growth. But DEPR (2012) did not perform contingency claim analysis, the valu­ation of the implicit liquidity guarantees and the computation of the liquidity insurance premium because these do not apply in India.

Their liquidity indicators were (i) the weighted average call rate (WACR) minus the RBI repo rate, (ii) the three month commercial paper (CP) rate minus the three month certificate of deposit (CD) rate, (iii) the three month CD rate minus the three month implied deposit rate, and (iv) the WACR minus the three month overnight indexed swap (OIS) rate. The RBI in its database on Indian economy stores above data for free distribution to the public.

Following are problems with the data on above variables:

1. The RBI repo rate varies over specified periods, not daily, i.e. since April 18, 2012 it is 8% till the date of submission of this chapter in November 2012. For any sample lying within this period, the variance of above rate is nil and the standard normal method of the SLI computation comes to a halt here.

2. The RBI provides weighted average rates of the CPs, not for the specific three month maturity.

3. Because the volume of the overnight call lending is the maximum on any business day compared to all other maturities the WACR is tilted towards the overnight call rate and hence it is inconsistent to adjust the WACR for the three month OIS rate.

3.7 Forecasting Bond Liquidity

In the context of the securities market the bid-ask spread may be viewed as a proxy for the illiquid­ity of a security (Chacko, Das, and Fan, 2010). In this study the line of distinction between an investor and a market maker vanishes when the limit orders by the investors fall within the bid-ask quotes of the market makers. It compares the bid quotes and the ask quotes for buying and selling an exchange traded fund (ETF) of bonds to a call option and a put option respectively with the exercise prices at the net asset value (NAV). So whenever a transaction is taking place, this study proposes that, one of the options has a non-zero value. This value represents the illiquidity. He devised a very handy formula of the illiquidity measure of the above ETFs which is likely to appeal to the Indian bankers, barring a few, the most of which are not quants, once the bond ETF market develops.

3.8 Liquidity Adjusted Value at Risk

Keeping in view that the risks associated with the market illiquidity have not been effectively incor­porated into the VaR models the most emerging markets are less than perfectly liquid and many securities cannot be traded with an ease in the markets of the emerging economies like India having been subject to the process of the financial sector reform and deepening pari passu with a series of liquidity crises since summer 2004, Roy (2004) attempted to construct a liquidity adjusted value at risk (L-VaR) model that incorporates the liquidity risk in the VaR models. He tested the performance of L-VaR model vis-a-vis existing VaR models and finds that in the Indian context the liquidity risk is an important component of the aggregate risks absorbed by the financial institu­tions. While surveying the academic literature on L-VaR he enumerated following six approaches:

1. Ad-hoc Approach,

2. OptimalLiquidationApproachorTransaction Approach,

3. Liquidation Discount Approach, 3.

4. Exogenous Liquidity Approach,

5. Market Price Response Approach, and

6. Intraday Liquidity Risk Approach.

Having critiqued above models for their imple­mentation problems, he found a modified version of the exogenous approach put forth by Bangia, Diebold, Scheurmann and Stroughair (1999) to be suitable in the Indian context. The following facts raise question of implementation ofthis approach:

1. Bangia, Diebold, Scheurmann and Stroughair (1999) applied their approach to compute the L-VaR of the foreign exchange.

2. Roy (2004) applied this approach to the coupon paying dated government securities for the period 2003-04.

3. Currently the RBI keeps the information about only treasury bills regarding bids received and bids accepted in the public domain, i.e. the weekly statistical supple­ment and the RBI Bulletin. The link www. nds.rbi.org.in mentioned at the end of Table 3 in Roy (2004) does not exist now and perhaps is replaced by https://www.ndsind. com/accessible only by member institutions via login with username and password.

4. The RBI does not yet ask the banks and financial institutes to compute L-VaR.

4. RECOMMENDATIONS

1. The banks may form a liquidity policy sepa­rately for the period when the financial year end is coming near, e.g. the last quarter of the financial year and other three quarters after ascertaining the nature of the link between the LAF and the call rates.

2. The RBI should make the bid and ask spread available to the public for each dated security traded. This would assist the academia to work on the L-VaR.

Though the RBI does not make it manda­tory for the banks to compute the L-VaR, but nevertheless they may do it for specific securities as well as for the G-Sec portfolio with a view to managing the market risk capital charge.

4. The Association of Mutual Funds in India should compute the liquidity risks of the ETFs for the benefit of both the investors and the regulators.

5. FUTURE RESEARCH DIRECTIONS

A behavioural study of the daily volumes of the deposits to and the withdrawals from the com­mercial banks may be performed by the academia if the banks publish data on these variables in the public domain since the links between the cycles of the deposits and the withdrawals by the differ­ent sectors of the industry on the one hand and on the other hand systemic shocks like oil price hike and the rise in non-performing assets may be explored via an econometric modeling with the appropriate dummy variables.¹

6. CONCLUSION

Against the backdrop of the evolution of the mac­roeconometric models involving time series data and the compatible softwares used by the official economists as also the independent researchers, this chapter discussed the application of some of above to the liquidity forecasting as a part and parcel of liquidity management by banks, the financial institutions and the regulators. It distinguished between the concepts of liquidity and the practices of liquidity management from the separate angles of the banks and the Central Bank. It picked up the time series samples from the different parts of the financial year and analyzed the relationship between the market liquidity and the Central Bank’s monetary operations in India. It examined the important studies on modeling liquidity and the liquidity risk for exploring their applicability in India and accordingly made certain recommendations. Finally it suggests the way to study the behavioral aspects of withdrawals from and deposits into the banks such as to strengthen their liquidity management policy.

REFERENCES

Basel Committee on Banking Supervision. (2008). Principles for sound liquidity risk management and supervision. Bank ofInternational Settlement. Retrieved November 4, 2012 from http://www.bis. org∕publ∕bcbs144.pdf

Bhole, L. M. (1987). Definition, measurement and determination of money supply. Economic and Political Weekly, 22(23-24), 898-901, 945-950.

Business Standard. (2011). RBIagain warns banks on asset-liability mismatch. Retrieved November 7, 2012 from http://www.business-standard.com/ india/news/rbi-again-warns-banksasset-liability- mismatch/423227/

Clearing Corporation of India Ltd. (2012). G-sec settlement statistics: Maturity wise outright trades. Retrieved November 11, 2012 from https://www. ccilindia.com/SecuritiesSettlement/Lists/SM- Maturitywiseoutrighttrades/Attachments/3860/ Maturity-Wise%20Report_30-Nov-12.pdf

Das, R. (2010). Handbook of fixed income secu­rities: Indian banking perspective. Saarbrucken, Grmany: Verlag Dr. Muller.

European Central Bank. (2009). Retrieved No­vember 9, 2012 from http://www.ecb.int/paym/ groups/pdf/mmcg/Liquidity_supervision_D-FS. pdf?18861cbb8ff9edcf34bf6bb8c0986600

Federal Reserve Bank of San Francisco. (2008). What is liquidity risk? Retrieved November 9, 2012 from http://www.frbsf.org/publications/ economics/letter/2008/el2008-33.html

Ferhani, H. (2005). Liquidity management by central banks, the role of banks in credit allocation and thoughts on barriers to financial intermedia­tion. In Proceedings of CCM’s 9th Annual Senior Level Policy Seminar. Port of Spain, Trinidad and Tobago: CCM. Retrieved November 5, 2012 from http://www.ccmfuwi.org/files/publications/ seminar/hferhani05.pdf

Goyal, A. (2011). History of monetary policy in India since independence (IGIDR Working Paper). Retrieved November 9, 2012 from http://www. igidr.ac.in/pdf/publication/WP-2011-018.pdf

Grandville, O. D. (2001). Bond pricing and port­folio management. Cambridge, MA: MIT Press.

Gujarati, D. (2003). Basic econometrics. Boston: McGraw-Hill.

Hamilton, J. D. (1994). Time series analysis. Princeton, NJ: Princeton University Press.

International Monetary Fund. (2010). Systemic liquidity risk: Improving the resilience of financial institutions and markets. Retrieved November 9, 2012 from http://www.imf.org/external/pubs/ft/ gfsr/2010/02/pdf/chap2.pdf

Mishra, R. N., Mohan, G. J., & Singh, S. (2012). Systemic liquidity index for India (Reserve Bank of India Working Paper). Retrieved November 28, 2012 from http://www.rbi.org.in/scripts/Publica- tionsView.aspx?id=14304

Mitra, A. K., & Abhilasha. (2012). Determinants of liquidity and the relationship between liquid­ity and money: A primer (Reserve Bank of India Working Paper). Retrieved September 4, 2012 from http://www.rbi.org.in/scripts/Publications- View.aspx?id=14331

Money Market Contact Group. (2009). The devel­opment of robust and internationally consistent approaches for liquidity supervision: The current state of international initiatives. Author.

Pindyck, R. S., & Rubinfeld, D. L. (1998). Econo­metric models and economic forecasts. Boston: Irwin McGraw-Hill.

Reserve Bank of India. (1999). Asset-liability management system. Retrieved November 8, 2012 from http://rbidocs.rbi.org.in/rdocs/notification/ PDFs/5212.pdf

Reserve Bank of India. (2002). A short-term liquidity forecasting model for India. Retrieved September 1, 2012 from http://rbidocs.rbi.org.in/ rdocs/PublicationReport/Pdfs/30016.pdf

Reserve Bank of India. (2003). Report of the internal group on liquidity adjustment facility. Re­trieved November 5, 2012 from http://rbidocs.rbi. org.in/rdocs/PublicationReport/Pdfs/40456.pdf

Reserve Bank of India. (2008). RBI’s growth stimulus. Retrieved November 28, 2012 from http://rbi.org.in/scripts/BS_PressReleaseDisplay. aspx?prid=19612

Reserve Bank ofIndia. (2012). Draft guidelines on liquidity risk management and Basel IIIframework on liquidity standards. Retrieved September 4, 2012 from http://rbidocs.rbi.org.in/rdocs/content/ pdfsMURS210212FL_A.pdf

Roy, S. (2004). Liquidity adjustment in value at risk (VaR) model: Evidence from the Indian debt market. Reserve Bank ofIndia Occasional Papers, 25(1-3), 1-16.

Saunders, A., & Cornett, M. M. (2008). Financial institutions management - A risk management approach. Boston: McGraw-Hill.

Sethuraman, V. N. (2008). Liquidity risk man­agement demystified. CAB CALLING. Retrieved November 1, 2012 from http://www.cab.org. in/Lists/Knowledge%20Bank/Attachments/53/ Liquidity%20Risk%20Management%20Demys- tified.pdf

Severo, T. (2011). Measuring systemic liquidity risk and the cost of liquidity insurance (IMF Work­ing Paper). Retrieved November 11, 2012 from http://www.imf.org/external/pubs/ft/wp/2012/ wp12194.pdf

Vento, G. A., & Ganga, P. L. (2009). Bank liquidity risk management and supervision: Which lessons from recent market turmoil?. Journal of Money, Investment and Banking, (10), 79-126.

Working Group on Money Supply. (1998). Money supply: Analytics and methodology of compilation. Mumbai, India: Reserve Bank of India.

KEY TERMS AND DEFINITIONS

Basel III: Stronger norms than B asel II in terms ofliquidity and capital.. Standards and regulations designed by B asel committee on B anking Supervi­sion (BCBS) regarding market risk and capital to be set aside as cushion against market and credit risk. Basel I focused on credit risk.

Basis: Basis means difference between spot rate and forward rate in the derivatives literature. In asset management liability literature it means the intra-bucket spread between asset rate and liability rate.

Call Money: Unsecured short term interbank loan.

Dirty Price: Clean price plus broken period interest. Clean price means present or fair value.

Duration: When interpreted as percentage of price, it means sensitivity of the fair price of a bond towards fluctuations in the benchmark rate like 10 year government securities rate. When interpreted in terms of number of years it means the period of recovery of security price. Duration is more popular for academicians but for practi­tioners modified duration is more accepted with similar interpretations.

Econometrics: Application of Statistics in Econometrics.

Fair Price: Present value of all the future cash flows from the security discounted with the rate of return or opportunity cost of capital avail­able from an alternative avenue of investment. If market price is different than fair price, arbitrage makes them equal. The concept originated from Marginal Efficiency of Capital (MEC) propounded by Keynes.

F-Test: A test to examine whether any signifi­cant regressor is omitted in the regression equation.

Inflation: Rise in prices, normally measured in terms of wholesale price index (WPI).

Null Hypothesis: Normally a statistical re­lationship under inquiry into its very existence. Null hypothesis assumes the relationship does not exist. Alternative hypothesis assumes the relationship exists.

Liquidity: Ability to procure cash without incurring any premium.

Liquidity Adjustment Facility (LAF): A monetary policy tool of the Central Bank to have reign on systemic liquidity.

Ordinary Least Square (OLS): The most basic but the widestly accepted tool of statically estimating relationship s between variables. Dif­ferent modified versions of OLS are found in almost all econometric models.

Rate Sensitive Assets (RSA): Floating rate loans, floating rate bonds etc, are examples of RSA. Fixed rate products are non-RSA.

Rate Sensitive Liabilities (RSL): Floating rate deposit is an example of RSL. Fixed rate deposit is an example of non-RSL.

Repo: Repurchase agreement between banks and the Central Bank.

Tail: The left end of the graph of normal probability distribution denoting extreme losses as happened to big US banks during the sub­prime crisis.

Simulation: An exercise of generating samples with pre-specified properties, expected to repeat historical behaviour of the variable under con­sideration.

Solvency: Excess of assets over liability in the balance sheet.

Structural Liquidity: Maturity wise statement of inflow and outflow of cash to be prepared by banks for reporting to the Central Bank as a part of asset liability management.

Value at Risk (VaR): The amount of the largest loss in the market value of the investment portfolio with a probability of say, 99% or 99.5%. It is a measure of market risk.

ENDNOTES

¹ Author acknowledges use comments from Ugam Raj Daga.

This work was previously published in Emerging Methods in Predictive Analytics, edited by William H. Hsu, pages 172-185, copyright 2014 by Information Science Reference (an imprint of IGI Global).