Market Risk Management: A Brief Discussion

by professor Eben Maré, University of Pretoria

Multibillion dollar losses in the financial markets over the last two decades have served to focus companies’ attention on management of market risks. In many cases these losses have arisen from new investment related product innovation or a mere misunderstanding of the risks inherent in a specific new business endeavour involving the financial markets. In this short note we shall aim to present the essential ideas underlying a market risk management programme.

Firstly, what is market risk? Market risk arises by virtue of our income statement or balance sheet changing in value as a result of changes in the values of traded securities, these typically being:

1. Equities,

2. Foreign currency,

3. Interest rate(s),

4. Commodities,

5. Credit spreads,

6. Changes in volatilities on the above instruments or associated correlations.

It should be clear therefore that if we understand the market risks inherent in our company that we are placing ourselves in a position to hedge or mitigate these risks or even gear up on these risks!

Our total market risk management ethos consists of the following key

elements:

• Risk identification,

• Risk measurement,

• Monitoring,

• Control,

• Testing.

We shall briefly elaborate on these elements below. [[[PAGE]]]

Risk identification

It should come as no surprise that the first element of our management programme is the identification of all risks. This means a thorough understanding of our business and proper examination of our income statement and balance sheet to understand what market variables give rise to changes in these. Ideally this is done by examination of daily profit/loss statements for the overall business and comparing those with changes in underlying market variables.

Risk measurement

Having identified sources of risk it becomes crucial to quantify the extent of such risk. This is usually done through several risk measurement metrics ranging from simple spot equivalent positions to value-at-risk and stress testing, as examples, depending on the sophistication of the company involved with the analysis.

If a company is exposed to a set of risks which are very similar a proxy could be made to perform a ‘back-of-the-envelope’ calculation to quantify the risk. An example might be to approximate a basket of equities by an index. The advantage of such an approach is simplicity and yielding an ‘effective’ position which management relates to. The obvious disadvantage is that we are making some serious assumptions as regards basis risks between different (but related) positions. These measures are also frequently applied at a global level where we would look at an overall exposure, say the amount of exposure to interest rates increasing by a basis point. This would look across the whole yield curve and again ignore the fact that different areas of the curve do not always move by the same amount.

Having identified the sources of risk, it becomes crucial to quantify the extent of such risk.

Stress tests typically consider the effect of some hypothetical scenario applied to our overall portfolio. An example could be to see how much we make/lose if the exchange rate depreciates by 20% overnight. Stress tests are very useful to identify risk but frequently also ignore basis risks. The results from stress tests could also be ignored because management finds the scenario totally implausible. The advice should be that the quantum of the profit/loss effect is not as important as the identification of exposure that we have. Stress tests are very useful to identify embedded optionality.

Value at Risk (VaR) presents another measure of market risk frequently used by banks and regulators. VaR has been defined as the loss (stated with a specified probability) from adverse market movements over a fixed time horizon, assuming the portfolio is not managed during this time. Hence VaR is measured as the lower percentile of a distribution for theoretical profit and loss that arises from possible movements of the market risk factors over a fixed time horizon.

It is important to understand that no risk measurement metric is perfect - the interpretation thereof is very important, though, as well as how we put it to further use.

Monitoring and control

If we are able to attach a metric to an identified risk it means that we have a sense of the economic damage that could be caused as a result of that risk. We should judge against that the economic benefit from having the risk - such as the amount of profit or other advantage that we would derive.

Typically we would not want to have unbounded amounts of any risk and would impose limits to control these risks. It is important therefore to ensure that we set our controls (limits) to be reflective of the identified risks and that these accurately control the exposure that will ultimately feed through to our income statement.

Our process is therefore to monitor the measured risks against limits on a periodic basis and to ensure the correct level of control of our risks.

Testing

Market risk management is  most successful in an environment where the risks are transparently discussed and understood. It is of the utmost importance to dissect the risks and to understand whether our understanding of its impact to the balance sheet and income statement is correct, whether our measurement is correct and if we are adequately compensated for the risk.

Part of the process of challenging and testing risk measurements would be to ensure that we have captured all our risks. Throughout the whole process we need to continually revisit all our assumptions. We started off by investigating our income statement - it is therefore only fair that we return there; our risk measurements must balance back to the income statement in the sense that market moves applied to market sensitivities should lead to profit/loss numbers; if this is not the case it probably means that we are missing something from our risk model or possibly some operational error.

We need to ensure that we understand hidden risks and question these. These are frequently basis risks, an example being a long position in a government bond against a paid position in a swap - on an overall interest rate risk basis we could have no exposure as long as both instruments move in tandem, but the two underlying base curves could move in different ways creating basis risk; elementary measurement models would not always capture these type of risks. [[[PAGE]]]

Potential problems

There are many pitfalls in market risk management. As an example, when identifying risks we often have incomplete profit and loss information resulting from inadequate systems and processes. When measuring risks we frequently encounter insufficient data or inadequate proxies - typically when valuing derivative based contracts.

While systems and processes can always be improved it is of the utmost importance that we have the ability to interpret results, management can sometimes be impressed by very sophisticated concepts but in a crunch-time when heavy-hitting decisions need to be taken it is often the simple concepts that provide the most intuition and management information.

Conclusion

This note is effectively a coffee-break discussion on market risks and the management thereof. There are significant benefits from managing market risk within a proper framework, in particular a real understanding of cause and effect of financial market forces on our financial results.