Here is an essay on ‘Treasury Risk Management in Banks’ especially written for school and banking students.
Essay # 1. Supervision and Control of Treasury:
1. Treasury Risk Management:
Treasury risk management assumes importance for two reasons:
(a) The nature of treasury activity is such that profits are generated out of market opportunities and market risk is present at every step;
(b) Treasury is also responsible for balance sheet management, i.e. market risk generated by other operational departments. We will deal with the first aspect a little more elaborately.
2. Concern for Treasury Risks:
Bank management is highly sensitive to treasury risk, as the risk arises out of high leverage the treasury business enjoys. The risk of losing capital is much higher than, say, in the credit business. Bank’s capacity to extend loans is limited by the resources at its command, that is, deposits and other borrowings.
In case of a loan, the risk is limited to the principal and interest, which may be lost, fully or partly, over a period of time. Most of the loans are also secured by tangible assets. The risk is ‘capped’ by the amount invested in the loan asset. Potential loss in loan assets is known as credit risk.
Treasury on the other hand, has a very low funding requirement, which we call as high leverage. For instance, treasury can buy and sell foreign exchange of value Rs. 100 crore without any direct investment of funds, except for allocation of risk capital as per capital adequacy requirement of RBI. At the same time, an adverse movement of the exchange rate by Re. 1 may result in a loss of over Rs. 1 crore to the bank – which is a straight loss of capital.
A second reason for management concern is large size of transactions done at the sole discretion of the Treasurer. As we have learnt earlier, whether it is foreign exchange or money market or investment business, the value of a single transaction may range from Rs. 5 crore to Rs. 50 crore or even more in larger banks. The limits are delegated to the Treasurer in advance, and individual market deals rarely need specific approval from the management. If the Treasurer commits an error of judgment, consequent losses to the bank would be enormous.
A third factor closely connected to the above is that the losses in treasury business materialize in very short-term, and the transactions, once confirmed, are irrevocable – hence no corrective action is possible. Particularly in foreign exchange, markets react so fast that profits or losses on trade deals are almost instantaneous. Traders are generally not allowed to hold open positions for long, as the risk of loss increases with time.
The source of risk in treasury activity is variation in the market price of currency or security, when there is a gap between the buy leg and sell leg of the transaction. The risk is hence termed as market risk, as opposed to credit risk of loan assets of the bank. The variability of the price, upward or downward, is known as volatility. In case of currency, it is known as volatility of exchange rate and in case of securities, it is volatility of interest rates (security prices have inverse relationship with interest rate movement).
Treasury also faces funding risk, or liquidity risk, as all ‘settlements’ need to be funded. Treasury is also required to bridge the funding gaps of the bank by borrowing funds in the market at short notice. Liquidity involves managing cash flow mismatches, and if the liquidity is not readily maintained, interest costs will go up, sometimes threatening the viability of banking operations.
Liquidity and interest rate risk are two sides of the same coin, but the risks are dealt separately, as banks are highly sensitive to liquidity risk. Asset liability management (ALM) of the bank is also closely connected to market risk.
Treasury risks are primarily managed by conventional control and supervisory measures, mostly in the nature of preventive steps, which may be divided into three parts:
i. Organisational Controls:
The organisational controls refer to the checks and balances within the system. Treasury is basically divided into three parts: the front office, back office and the mid office.
The front office generates deals with counter-party banks (purchase and sale of foreign exchange, securities etc. & lending and borrowing operations). Treasury may enter in to currency dealings either on its own (Trading Book), or on behalf of clients (Merchant Book) or for bank’s internal requirement.
Security deals are either for Bank’s SLR requirement, or for Treasury’s own trading book. Though Treasury may also buy and sell securities on behalf of its retail clients, the activity as on date is not very significant – many banks have associate companies specializing in fund management. Treasury’s money market activity is exclusively to meet bank’s own requirements.
Front office is headed by Chief Dealer, assisted by other dealers in foreign exchange, securities market and money market. Larger banks may have dealers specializing in specific activities, such as corporate dealers, cross-currency dealers, equity traders etc.
The back office is responsible for confirmation, accounting and settlement of the deals. The back office obtains independent confirmation of each and every deal from the counterparty and settles the deal only if it is within the exposure limits allowed for the counterparty. Back office also verifies that the rates/prices mentioned in the deal slips are conforming to the market at the time the deal is entered into.
Such verification is done by scanning the Reuter/Bloomberg screens and noting down the market rates at regular intervals during the day (generally every two hours or more frequently). The lending and borrowing rates are also similarly verified. The back office has the overall responsibility for compliance with exposure limits and position limits imposed by the Management and RBI, as well as for accuracy and objectivity of the transaction detail.
Banks also have a Middle Office (mid-office) which is responsible for risk management and management information system (MIS). Mid-office would ensure treasury’s compliance with Board approved policies bearing upon FX risk management, investment management and liquidity management.
Some of the key responsibilities of Mid-office are: monitoring compliance with risk limits set in the respective policies, ensuring compliance with regulatory requirement, daily mark-to-market (MTM) valuation of Treasury positions, verification of pricing of treasury products, including derivatives, and periodic reports to top management.
Mid-office maintains the overall risk profile of Treasury and monitors the liquidity and interest rate risks closely, in line with Asset Liability Management (ALM) guidelines. In quite a few banks, the ALM support group is a part of Mid-office, or works closely with the Mid-office.
Front Office and Back Office functions need to be segregated totally, with both offices reporting independently to the Treasury Head. Mid office may report directly to the Treasury Head or to a senior executive, outside Treasury, such as Chief Risk Officer, to ensure better risk control.
ii. Internal Controls:
The most important of the internal controls are position limits and stop loss limits. The limits are imposed on the dealers who trade in foreign exchange and securities. Trading is a high risk area, vulnerable to sudden market fluctuations and the limits imposed by management are preventive measures to avoid or contain losses in adverse market conditions.
The trading limits in the context of foreign exchange are of three kinds:
(i) Limits on deal size
(ii) Limits on open positions and
(iii) Stop-loss limits.
All limits are expressed in absolute amounts.
Limits on deal size prescribe the maximum value for a buy/sell transaction. The limit is a protection against potential losses on the deal. The limit generally corresponds to the marketable size of the transaction, and applies only to trade deals (and not to merchant deals).
Open positions refer to the trading positions, where the buy/sell positions are not matched. The Treasury may buy USD 1 million, and hold on to the position with an intention to sell when the USD appreciates against the Rupee. Not only there is a potential loss if the US dollar does not appreciate, but there is also a ‘carry’ cost, as the Treasury loses interest on the USD funds during the holding period. Treasury may also take forward positions expecting a rise or fall in the exchange rate.
The management, therefore, limits the size of open or unmatched positions. The limits in foreign exchange trade are defined as day light and over-night – the day light limits pertain to the intra-day positions, say if the dealer purchases currency in the morning and sell it in the afternoon. As the forex market is very active, the currency prices may move from moment to moment, the currency may lose value in the mean-time. The overnight limits are smaller as the dealers may continue to hold the position for next day.
Position limits are prescribed currency-wise as also for aggregate position expressed in Rupees. For the purpose of aggregation, currency-wise net position is first translated into USD at the day-end rate and then converted into Rupees. The overnight limits are to be pre-approved by RBI, who also prescribes the method of arriving at the aggregate position.
Even when there are matching positions, there is scope for loss if the delivery is at different points of time. In a swap deal, the dealer may purchase USD at spot and sell it forward, say, after three months.
It is a matched deal as the purchase and sale prices are prefixed and hence there is no exchange risk. However, the forward prices are derived out of interest rate differentials and there is an opportunity cost if interest rates move adversely during the transaction period.
The risk in forward positions is measured by ‘gaps’ (residual time for completion of the transaction) which are then capped with a gap limit – akin to position limits on spot trading positions. All the forwards are re-valued periodically (generally monthly, but in most computer systems, daily valuations will be available) and the outstanding positions in each time bucket are subjected to gap limits. The gap limits are internally approved by the management.
Position limits and gap limits attract capital requirement, as prescribed by RBI.
Similar controls exist for securities trading, where the size of the deal, maximum value of securities held for trading and holding period are defined by the Management. For non-SLR securities, minimum credit rating requirement is also prescribed by the Management.
Stop-loss limits represent the final stage of controlling trading operations. When the market moves adversely, the open positions will result in loss. A dealer typically would like to wait till the market turns around, so that he can close the position with a profit. There is an added risk in that the market correction may not take place as anticipated, and the losses may continue to accumulate in the mean-time.
The stop-loss limits prevent the dealer from waiting indefinitely and limit the losses to a level which is acceptable to the management (which the bank is in a position to absorb). If the stop loss limit is USD 5000, the dealer must close his position (i.e. sell or buy at a loss). Any violation of stop-loss limit is viewed seriously by the management.
The stop-loss limits are prescribed per deal, per day, per month as also an aggregate loss limit per year. Back office need to monitor all the limits meticulously. Back office is also responsible for valuation of deals, and if it is found out that the open positions upon valuation are found to be resulting in loss, the front office should be immediately informed for applying the stop-loss limits, (normally, it is front office who monitor the positions, and back office only keeps a second check). The forex trading positions need to be valued (marked to market) daily for this purpose.
Stop-loss limits are applied on securities trading in a different manner, as price movements in securities market are larger and more irregular and as the market is less liquid as compared to forex market. The market risk is controlled in terms of mark-to-market value of the trading portfolio, applying risk measures such as VaR and duration.
iii. Exposure Ceiling Limits:
Exposure limits are kept in place to protect the bank from credit risk. Credit risk in Treasury may be split into default risk and settlement risk. Default risk is typically when the bank lends in the money market (mainly to other banks), the borrowing bank may fail to repay the amount on due date. Similar risk is there in repo transactions also.
Even though inter-bank market is considered to be relatively risk free, it is not uncommon that a weak bank may suddenly become bankrupt, or, there is a run on the bank squeezing its liquidity. Even assuming that there is no credit risk in short-term lending, it is not prudent that Treasury lends its entire surpluses to a single bank or to a handful of banks.
The settlement risk refers to the possible failure of the counterparty to the transaction (which is generally a bank or a financial institution) to deliver/settle their part of the transaction. While ideally all deals should take place in DvP (Delivery vs. Payment) mode, it is not always possible to achieve the standard, either for want of institutional mechanism, or due to physical barriers (such as different time zones).
Delivery of government securities is already taking place against payment, as the banks have both the securities account (SGL) and funding account with RBI, so that debit and credit can take place simultaneously. Similar sophistication is also present in exchange of non-SLR or corporate securities depository participants CCIL mechanism.
On the other hand, in foreign exchange transactions, the time gap between Rupee/settlement and FC settlement is unavoidable when the FC settlement takes place in a different time zone (e.g. USD in New York). The settlement risk is bankruptcy, or inability of the counterparty for whatever reason to complete their leg of transaction.
It is for the above reasons that banks fix exposure limits for counterparties, including other banks, financial institutions, mutual funds, primary dealers, forex and security brokers etc. The exposure limits are fixed on the basis of the counterparty’s net worth, market reputation, track record and/or credit rating.
The limits also take into account the size of treasury’s operations, so that the business is spread over several counterparties and there is no concentration of risk. The limits vary in relation to the period of exposure; whether the obligation is for overnight, 3 months or 6 months – the longer the exposure, greater is the risk and the limits are adjusted accordingly.
The exposure limits are also fixed for foreign exchange and money market brokers, in order to avoid business concentration, even though they are only intermediaries and are not counterparties. While the limits are generally left to the banks’ discretion, Reserve Bank of India has imposed a ceiling of 5% of total business in a year for individual brokers, subject to exceptions being reported to the bank’s management for ratification. All the limits are to be reviewed at least once in a year.
It is the responsibility of back office/Mid-office to ensure that Treasury complies with the exposure limits meticulously.
Essay # 2. Market Risk and Credit Risk:
We have frequently referred to default by a counterparty and market movements. In that we are indirectly referring to two types of risks faced by the bank in all spheres of its activity. One is credit risk, or the risk of losing funds invested together with interest, fully or partly, on account of failure of the counterparty to honour its obligations. Then there is market risk, where the price of a security, interest rates or exchange rates move in such a way that the value of an asset diminishes or the liability under an existing obligation increases.
Thus market risk is a confluence of liquidity risk, interest rate risk, exchange rate risk, equity risk and commodity risk. Considering all these factors, the Bank for International Settlements (BIS) defines market risk as “the risk that the value of on-or off-balance sheet positions will be adversely affected by movements in equity and interest rate markets, currency exchange rates and commodity prices”. The market risk is also known as price risk.
The three main components of market risk are liquidity risk, interest rate risk and currency risk.
1. Liquidity Risk:
Liquidity risk refers to cash flow gaps which could not be bridged. Let us assume that the Treasurer has borrowed in call market and purchased a 5-year government security, assuming the bond prices would go up next day and he can sell the security with profit. Let us further assume that the bond market collapses next day and the Treasurer could not dispose of the security. Though the bank is solvent, the treasury has faced liquidity risk, as he needs to borrow funds in the market at whatever cost, if he has to avoid default or delay in repayment of the call borrowings. Liquidity risk thus translates in to interest rate risk.
Treasury is generally prepared to meet the known events, such as due date for a money market loan, or for a deposit. However there are unforeseen events such as invocation of a guarantee or premature payment of a large deposit which would strain the bank’s liquidity. The Treasury needs to have a contingency plan to meet any liquidity crisis.
2. Interest Rate Risk:
Interest rate risk refers to rise in interest costs (of a liability) or fall in interest earnings (from assets) eroding the business profits.
Treasury deals in financial assets, value of which is highly sensitive to interest rate movements. A steep rise in interest rates may cause a crash of bond market, eroding the value of securities held by Treasury. If liquidity is not planned ahead, Treasury may need to borrow at higher cost to meet its obligations.
The interest rate risk is present wherever there is a mismatch between assets (cash inflows) and liabilities (cash outflows). The incremental deposit funds of the bank, say with an average maturity of 1 year, to the extent they are not lent, are invested by Treasury, say, in 3-month T-bills. If the yield on T-bills, which changes every three months, does not match with the cost of the deposit, the net earnings of the bank will be negative.
3. Currency Risk:
Currency risk or exchange rate risk is also a manifestation of interest rate risk, although for the sake of clarity, it is identified as a component of market risk.. Interest rates are influenced by factors like domestic money supply, rate of inflation, activity in debt and equity markets etc. which also influence exchange rates.
However, exchange rates are influenced more by external trade, global interest rates and capital flows. As globalization progresses, exchange rates and interest rates are increasingly influenced by similar factors, most prominent being GDP growth rate, global interest rates and capital flows.
To put it simply, interest rate is domestic cost of currency, while exchange rate is external cost of currency. Forward exchange rate of two currencies actually reflects interest rate differentials of respective currencies (though in India, Rupee being not fully convertible on capital account, forward rates is distorted by supply and demand for the currency). You will appreciate the relationship between interest rates and exchange rates better when you study purchasing power parity and interest rate parity theories in Foreign Exchange module.
Although bank treasuries have limited exposure to equity markets, movement of stock prices is an important factor influencing financial markets. Stock market also reflects the overall liquidity in the system, interest rates and the exchange rate movements. If the currency is convertible, the exchange rate and interest rate changes play even greater role in attracting foreign investment inflows into the secondary market.
There are also commodity markets where price movements are affected by not only supply and demand, but also by economic factors like interest rates and exchange rates. Treasuries with exposure to commodity futures and options are particularly vulnerable to changes in commodity prices. Treasurers even when they are not participating in commodity markets can observe close linkages between, say, price changes of gold and USD/Re exchange rate.
All the free markets are highly susceptible to speculation, which in fact is the essence of Treasury’s trading positions. It is therefore perception of future changes in interest rates and exchange rates, rather than the actual changes that direct the market movements.
The market risks directly affect the transaction values and thereby profits of the treasury. Additionally Treasury also plays an important part in the risk management of the bank as a whole. Market risk translates into balance sheet risk of the bank, and treasury is closely connected with the asset – liability management. (ALM).Treasury provides inputs to ALM and is also instrumental in implementing the risk management solutions.
Essay # 3. Measures of Risk:
The movement in currency prices or security prices cannot be accurately predicted and the uncertainty associated with their price movements gives rise to price risk. At the same time the treasurer should have some idea of the inherent risks and the way they would affect his positions. This quest for risk solutions, led to two important measures of risk, known as value at risk and duration.
1. Value at Risk (VaR):
VaR is a statistical measure indicating worst possible movement of a market rate, over given period of time, under normal market conditions, at a defined confidence level. For instance, a overnight VaR of 45 bp for USD/INR rate at 95% confidence level implies that there is only 5% chance of the rate worsening beyond 45 bp next day. If today’s spot rate is 46.00, tomorrow the worst possible rate for exports can be assumed to be 45.55, with reasonable safety – there is only 5% chance of the rate being worse than 45.55.
Similarly, if overnight VaR of 1-year G-Sec yield is 0.35%, current yield of 7.75% is expected to fall/rise by not more than 0.35% by tomorrow. In the worst-case scenario, a prospective buyer of security may therefore expect the yields to fall to 7.75% – 0. 35% = 7.40% by next day, while a seller of security may expect rise in the yield to 7.75% + 0.35% = 8.10% by next day. At 95% confidence level, there is only 5% possibility of adverse change being higher than 0.35% (at 99% confidence level, there is only 1% possibility of loss being higher than VaR).
VaR is derived from a statistical formula based on volatility of the market. Volatility is the standard deviation from the mean of, say, USD/INR exchange rates (or any other asset prices) observed over a period. Volatility assumes a normal distribution curve and the no. of standard deviations from the mean denote the probability of reaching a target level. The volatility multiplied by the no of standard deviations required for a given confidence level results in the VaR.
There are a number of ways, with technical variations, to calculate the VaR. Three popular approaches to VaR are: parametric approach based on sensitivity of various risk components. For instance, say the price of a stock depends on its sensitivity to index changes, to interest rate changes and to changes in the exchange rates – all these components are built into a complex formula to arrive at the VaR of the stock.
The second approach is based on Monte Carlo simulation, where a number of scenarios are generated at random and their impact on the subject (stock price/exchange rate etc.) is studied. The third approach is to use historical data to arrive at the probable loss.
The historical data may simply be time series of data prevailing over a period (e g daily USD/INR exchange rate for last 90 days), or an index of changes (e g change in price over previous day). Progressive weights may also be assigned to the data, as more recent information has greater impact on future price movements.
While the methodology to arrive at VaR may appear to be complex, the utility lies in that the concept is easy to understand. The Management would like to know VaR of all risk positions, as it offers a single figure – an absolute amount – a potential loss which may affect bank’s earnings, or, net worth.
In Treasury, VaR is used to measure potential loss, or the worst case scenario, while holding a trading position either in foreign currency or in securities, i.e. VaR measure can be used to assess the currency risk as well as the interest rate and price risks. The VaR is used to measure the risk of a single investment, or more generally, a portfolio of investments.
VaR is most commonly used to measure overnight risk, or risk over short periods, say, over 1 month. VaR for longer periods is calculated as overnight VaR multiplied by √n (square root of n, where n is the period for which VaR is required). However, for longer periods, VaR is not a valid measure.
Duration is a measure widely used in investment business, though the concept of duration is applicable to all assets and liabilities, where interest rate risk is present. To understand Duration, we need to be familiar with the concept of YTM or Yield to Maturity of a bond.
Treasury invests in government securities and non-government securities of various descriptions, viz. bills, bonds and debentures – hereafter referred to as bonds. The bonds carry a coupon rate of interest which is payable on 100% value of the bond (par value, as at the time of issue). However, the bonds may be traded at a discount (<Rs. 100) or at a premium (>Rs. 100), depending on the interest rate trends in the market. The traded price is based on the market rate of interest for residual period of the bond and is constantly changing in the market.
The effective return on a bond (based on the coupon rate, market price and residual maturity) is known as yield. The yield is different from interest rate in that the yield takes into account the cash flows during the life of a bond, including interest payments (which are normally semi-annual or annual) and the payment of principal upon redemption.
All the cash flows are discounted to arrive at a present value (PV) and the rate of discount at which the present value equals the market price of a bond is known as the yield to maturity (YTM) or, simply, as yield on the bond. Conversely, it is the discount rate at which NPV (net present value = PV – market price of the bond) of the bond is zero. Yield is effective rate of return on amount invested in the bond. (The YTM is calculated on the bond calculator/built in formula in Excel/or from bond tables).
For instance, a bond carrying a coupon rate of 5% with a balance maturity of 2 years is traded at a discount of 2%, i.e. at a price of 98. This indicates that market rate of interest is higher than 5% and hence the market price is less than the par value. Interest at 5% on a unit price of 98 will work out to 5.10% which is known as current yield of the bond. The YTM of the bond works out to 6.08%.
The yield is internal rate of return reflecting the ruling interest rates. Yield and price of a bond move in inverse proportion. If the yield rises, price of a bond falls. If the yield falls the bond price rises. It is the same relationship between interest rates and bond prices.
Bond yields tell us the rate of return at which the present value of cash flows equals the market price. However the YTM does not reveal how volatile are the bond prices and how they respond to changes in interest rates. The YTM of two bonds may be same, but the price risk associated with them may be different on account of maturity or frequency of coupon payments. The YTM of two bonds hence is not comparable.
Duration is a measure which helps us understand the impact of interest rate on the price, by taking into account periodicity of coupon flows.
Duration is weighted average measure of life of a bond, where the time of receipt of a cash flow is weighted by the present value of the cash flow. If the first cash flow (payment of interest) is occurring after 6 months from the date of investment, the period of 6 months is multiplied by present value of the cash flow (0.5*PV).
If the second cash flow is occurring after 12 months, the period (1 year) is multiplied by present value of the cash flow. The present value of final payment of interest and redemption of principal, occurring, say, after 5 years from the date of investment is similarly used to weigh the maturity period (5*PV).
The aggregate of results so obtained is divided by the total of weights (total of PV of each cash flow, which is also market price of the bond) to yield ‘Duration’. It is also known as Mecaulay Duration, following the originator of the concept, Frederick Mecaulay.
Duration is expressed in terms of years. The longer the duration, greater is the sensitivity of bond price to changes in interest rates. The duration thus helps compare bonds with different structures to find out which bond entails greater interest rate risk.
The duration of a zero coupon bond is equal to its time to maturity, as there are no interim cash flows, and redemption value of the bond includes interest on the initial investment.
To the Treasurer, duration of a bond portfolio is more important than the duration of a single bond. The weighted average duration of a number of bonds in the portfolio can be arrived at by adding weights to the duration of individual bonds, the weight being the market price of the bond (expressed as a % to the face value of the bond).
If the Treasurer has a target period of holding, say of 2 years, he can immunize his portfolio from interest rate risk by ensuring that at any point of time the average duration of his portfolio is equal to 2. This of course necessitates constant reshuffling of the bonds in the portfolio, as the duration changes with lapse of time (as a bond nears its maturity).
Modified Duration (MD) is a more direct method to measure the price sensitivity of a bond. The MD is arrived at by dividing the duration with the interest rate (which is actually principal plus interest for 1 year, expressed as 1+Y, where Y is the yield). If the duration of bond yielding 5% is 2.5, the MD = 2.5/(1 + 0.05), or 2.38.
Any change in yield multiplied by MD will give the likely percentage change in price of the bond. If the yield rises by 5 basis points, the change in the price of the above bond will be 2.38*0.05 = 12 bp. Since prices move in inverse proportion to yields, 5 bp rise in yield causes the price of bond to fall by 12 bp. The MD thus indicates price sensitivity of a bond per unit of change in the yield levels.
There are, however, certain limitations in application of the concept of duration. The modified duration is valid only for small changes in the price, and is also not uniform at different levels of the price. A proportionate change in prices corresponding to the change in yields is possible, only when the yield curve is linear i.e. the short-term and long-term interest rates increase or decrease in a fixed proportion to the term and the yield curve is a straight line steeping upward or downward.
In practice, term structure of interest rates is such that the long-term interest rates rise or fall more slowly as compared to the short-term rates and the yield curve is rarely a straight line. For the sake of greater accuracy, therefore, an adjustment is made to the duration measure for convexity of the curve.
While we have seen duration in the context of investment business, the concept is equally valid for any interest earning asset, or liability. In bank’s balance sheet, duration of assets (loans) and duration of liabilities (borrowings) can be calculated to find out how sensitive bank’s earnings are to changes in market rates of interest. Difference in the duration of assets and duration of liabilities is expressed as duration gap, which is useful for macro-hedging of balance sheet risk.