Forum Documents
   Equity at risk
     
     5 - Alternative methodologies for establ...
     
















 

Equity at risk

5 - Alternative methodologies for establishing regulatory capital standards

The preceding sections have set out the central considerations which we believe should characterise any new regulatory capital system. In summary they are

  • Regulatory capital requirements should seek to provide protection against unexpected and stress loss events at firms and not against expected losses. Expected losses should be treated as business costs and either charged to the p&l account or provisioned against

  • The risk which a regulatory capital requirement seeks to protect against is the risk that a firm's net earnings will be negative in any period, so resulting in a depletion of capital. Consequently the regulatory capital regime needs to protect against any material threat to a firm's earnings

  • The quantum of regulatory capital needs to be sufficient both to protect against the risk to the firm's earnings of an adverse stress event and to ensure that the firm has sufficient capital following such an event to allow an orderly work out

  • The time horizon for setting the regulatory capital requirement should be the length of time which it would take for a crisis to emerge together with the length of the work out period, during which time the firm can be assumed to have no means of access to additional capital. We assume that this period is around three years

  • The regulatory capital requirement must be supplemented by effective supervision of firms' liquidity management policies, and the determination of the adequacy of a firm's capital must take into account the potential loss that the firm might incur in the event of a funding crisis

Given these requirements, we can now consider the extent to which the current Basel Committee methodology for setting regulatory capital as well as the methodology which some firms use internally for determining their own capital requirements are consistent with these requirements.

- Balance sheet mapping - The Basel Committee approach

The underlying premise which underpins the Basel Committee's approach to setting capital requirements for banks as well as most of the literature about capital management within financial firms is the belief that it is possible to relate the total capital needs of a bank to the risk which is embedded in the components of the firm's balance sheet. Implicit in this approach is the assumption that once a firm is able to measure the economic risk in its balance sheet it is relatively easy to scale up by a confidence factor or probability of default to reach the total capital which the firm requires. It is then often asserted that the firm needs to operate a process of capital allocation between its business areas in order to ensure that it is maximising its return on equity.

(This approach to capital allocation is also impossible to reconcile with any dynamic approach to risk management, since it assumes that the risk always adds to the capital irrespective of any action the firm may take to alter its risk profile. It is also counter intuitive, since a firm which used such a capital allocation methodology and which then reduced its risk profile would end up having to charge more capital to the business for each unit of risk assumed as a 'reward' for the risk reduction achieved. Further, the approach pays no attention to the capital creation process . A firm's capital is never static, but is the increased or depleted as a result of the business dynamics which are driven by considerations which are largely independent of any balance sheet measure of risk. )

This approach, which we refer to as 'balance sheet mapping', has historically proved a reasonably reliable methodology for measuring the exposure which a bank had to volatility in its published earnings. The main reason for this is that, where a bank's primary source of revenue is the net interest income from its loan portfolio, it is a reasonable simplifying assumption that the level of earnings is directly proportionate to the measure of risk which is calculated by reference to that same balance sheet. This was particularly true at a time when banks did not distinguish between the expected and the unexpected loss in the portfolio.
(In this situation, the volatility measure dominated the estimate of the mean to such an extent that the failure accurately to estimate the mean was relatively insignificant to the outcome of the estimation process ) A further advantage of this approach in the past was that it allowed for a common definition of the appropriate level of regulatory capital for banks - the 8% standard which was originally set by the Basel Committee - which could be related in a relatively intuitive manner to the observed volatility of banks' earnings to give an implied regulatory safety margin.

For the future, however, it is questionable whether the balance sheet mapping approach will remain a reliable estimate of the risk to a bank's earnings even for a bank that undertakes no other business than traditional credit activities. This follows because where banks properly account for expected losses it is much less likely that the unexpected or stress loss in the portfolio will be directly proportionate to the net interest income which the bank is earning. As a result, a balance sheet mapping approach, which seeks to measure the potential risk to a firm's earnings directly from its balance sheet without factoring in any measure of quality of its earnings is likely to become an increasingly unreliable measure of risk. Indeed it could even be seen to be counterproductive from a regulatory perspective because it will tend to overstate the risk and required capital for a bank which generates superior rates of return from its portfolio and will understate risk and capital for a bank which generates inferior rates of return.

Moreover, the balance sheet mapping approach becomes even less reliable when it is applied to risks other than portfolio type risks or is applied to firms which have a materially different business mix and earnings stream from a traditional commercial bank. In these cases the link between the risk measure and the firm's earnings is even more tenuous. Indeed the risks to the earnings of many firms which are active in international financial markets are increasingly unlikely to be the traditional banking problem of an adverse credit environment resulting in unanticipated provisions which in turn results in a depletion of capital. Instead the potential for loss derives from other sources of risk to the firm's earnings, whether related to liquidity risk, to operational risks or simply to the fact that other sources of business income are less reliable than had been assumed.

At the same time it is important to recognise that the availability of other sources of earnings at these firms provides a significant measure of protection against the risk of loss, which in turn reduces the risk which these firms pose to the financial system. This inability to recognise the risk mitigating effect of these other income sources is a fundamental weakness of the balance sheet mapping approach and is a further reason to support our view that it is an inappropriate basis for establishing a risk based capital standards in the future.

- The earnings volatility approach - emerging practice at firms

The weaknesses of the balance sheet mapping approach emphasise the importance of developing an approach to risk measurement and regulatory capital which captures all the potential sources of volatility in a firm's earnings stream and not just those that relate directly to the valuation of balance sheet items. In order to achieve this it is helpful to consider how financial firms themselves integrate risk measures into their overall financial and capital management processes.

As we have seen in Section 3, from the perspective of the management of the firm the measure of risk that has to be managed is the possibility that the firm's earnings will differ in any adverse respect from the shareholder's expectation of those earnings. The challenge for the management is therefore not only to ensure that the firm is adequately capitalised to ensure that it can continue in business and to survive a potential systemic shock, but also that it can continue to maximise its shareholders expected utility from holding its stock.

In practice, this means that the firm is engaged in two complementary exercises:

  • risk return optimisation for 'normal' risk; and

  • limitation of stress loss exposure in order to ensure that 'tail' risk is properly factored into the risk decision taking process

In the context of risk return optimisation, the firm must take into account both its expected earnings and the perceived volatility of those earnings as seen from the perspective of the shareholder. In determining the amount of risk which the firm is prepared to assume, the firm's management must be able to measure the potential volatility of those earnings and to seek to maximise the ratio of risk to return either by reducing the risk relative to the expected level of earnings or by increasing the expected return relative to a given level of risk. What is important to note here, however, is that the measure of risk which is most relevant to this exercise is the potential volatility of earnings in normal circumstances, sometimes referred to as unexpected or statistical loss or 'earnings at risk', and not the scale of loss that the firm might face in extreme adverse circumstances. Moreover, it is neither necessary from the perspective of risk return optimisation, nor is it possible at present in statistical terms, to show any systematic mapping between these measures of risk and the total capital of a firm.

(For a fuller account of the argumentation behind this statement see Shepheard-Walwyn and Litterman 'Building a coherent risk measurement and capital optimisation model for financial firms' )

In the context of stress loss limitation, by contrast, the considerations which the firm faces are different. The process of setting stress loss limits expressly recognises that there are situations where the statistically estimated measures of variance which the firm uses to manage its earnings volatility are in many cases relatively poor estimators of the extreme or tail events which the firm may face on rare occasions. The firm therefore has to attempt to estimate its potential exposure to these extreme events using a range of techniques and to place an overall limit on this exposure. The estimation techniques which firms are using vary from firm to firm, but are generally referred to as stress tests or scenario analyses. The purpose of these is to examine the firm's risk profile for its exposure to a predefined set of events which the firm recognises could happen - even if there is no historical precedent for the event in question - and then to place a limit on the financial loss which the firm might face if those events were to occur.

(It should be noted that stress testing and scenario analysis are relatively new disciplines for financial firms, and a variety of techniques are being examined. One such approach is the use of statistical estimates which are based on existing statistical estimates of unexpected loss. An example of such a stress test is the current Basel standard for market risk (3 times ten day 99% VaR). Another approach which is being explored by a number of firms in this context is the application of extreme value theory. The point which is relevant here, however, is that stress testing and scenario analysis are necessary processes to enable firms to derive their overall stress loss exposures and to assist them in determining their ability to withstand such events.)

Moreover in determining appropriate levels for the firm's exposure to stress loss, the management of the firm needs to consider how large a loss would be acceptable to the shareholders, and to require stress loss limits to be implemented which will seek to ensure that this amount, which we term the firm's risk bearing capacity, will not be exceeded if the firm encounters a particular stress event. It is therefore a key consideration for the senior management of the firm to be able to specify the firm's risk bearing capacity by reference to a value which is acceptable to the shareholders. Indeed in this context it is interesting to note that, while in the traditional analysis of capital in banking the presumption has always been that the firm's equity exists to absorb stress losses, it is our impression that in most cases large international firms do not share this view. Instead they set their risk bearing capacity at a value which relates not to the amount of their total equity capital but to their annual earnings in order to ensure that even in the event of an extreme loss their shareholders will not in fact have their equity holding in the firm diluted.

If this is an accurate representation of the process which firms are themselves adopting to ensure that they are adequately protected against potential systemic shocks, it would seem desirable that regulatory capital requirements, which exist to protect against the same type of risk should be calculated using a similar methodology. This implies that the regulatory capital requirement should be based on estimates of the firm's potential exposure to stress loss events and the extent to which the firm's earnings capability would protect it against such a loss.

With the benefit of these conclusions we can now seek to develop a framework for calculating a minimum regulatory capital standard which satisfies the conditions set out in Section 5 above

Risk Library * Forum Documents * Equity at risk