Economic capital: implementation practices and methodologies

Good news for the industry: A recent study from RMA confirms far more commonalities than differences in economic capital implementation practices among leading institutions.

Recent developments in the use of economic capital as a management tool have been obscured by the activity surrounding revisions to the Basel Accord. But although it's true that adoption of economic capital may have been hastened by the pressures of Basel, its use as an everyday instrument is based on a deeper need.

Economic capital's value to management is that it provides a common vocabulary and set of metrics for managing risk across the entire enterprise. Elements driving adoption of this technology include management's realization of the following:

* Improved loss experience, especially in the credit sector, can reduce earnings volatility and therefore have a positive impact on share value.

* Modern Portfolio Theory can be applied to the management of risk in financial institutions.

* There is a competitive advantage to be gained from employing economic capital as a management tool.

Many of these elements have been addressed individually in earlier RMA research projects and publications, as well as in the literature. However, in 2004, a group of major banks participating in RMNs Global Risk Policy Round Table asked RMA to study the extent to which economic capital measures have been adopted by the group members and to see whether any commonalitics in definition and practice exist.

This study of implementation practices and methodologies assisted RMA in defining the scope of the research and in developing the questionnaire used to survey the participants.

The findings showed many areas of commonality in terms of approach. They are important both as an indicator of common practice among large, internationally active institutions and as a directional marker for future development. Methodologies differ among these firms, however, and underline the fact that it is too early to categorize "best practices." Therefore, RMA intends to repeat the survey to better identify and confirm these trends and common practices within the industry.

Definition of Economic Capital

There are two views on the definition of economic capital. Ten of the respondents define it in terms of extreme losses: on the order of 99.95%, with a one-year time horizon. This view is essentially the one adopted by regulators in the Basel II capital framework. Meanwhile, two respondents define economic capital in a "going concern" framework. In this view, the firm is subjected to a far less extreme stress loss and is expected to emerge from such stress with an unimpaired ability to conduct business.

There is a consensus that economic capital should include only unexpected losses. Only one firm includes expected future profitability, in some circumstances, as an element of economic capital.

Elements of Capital

Capital itself is defined by all respondents to include common equity. Two-thirds include preferred stock, while one-third include quasi-equity securities that count as Tier I capital. Almost all exclude from the definition senior and subordinated debt, loan loss reserves, hidden reserves, and expected future earnings. And 50% of respondents considered goodwill to be a reduction of available capital (while a third considered it an increase in economic capital).

Uses of the Economic Capital Framework

The application of economic capital as a management tool has proliferated in recent years. All institutions that participated in the survey report institution-wide use of economic capital for assessing capital adequacy and for profitability reporting. Some 92% of firms use the technology for strategic planning and resource allocation to businesses, while 75% use it in transaction pricing for credit products.

In contrast with the heavy use of economic capital for profit-center reporting, only a third of the firms report economic capital by country.

Most firms (75%) adjust business-unit profitability with a capital charge based on the level of economic capital. This practice is consistent with the economic value-added framework as put forth by Stern and Stewart. Most firms (75%) base the cost of capital (actually, the cost of equity) on the Capital Asset Pricing Model and use the same cost of capital for almost all business traits. Any variations that exist are based on the cost of capital of other firms in the same line of business, or differing risk-free rates for different businesses.

Despite the increased use of economic capital, only 25% of respondents use it for senior management compensation. Not surprisingly, even fewer firms use it to drive compensation for lower-level managers. This finding may indicate a lack of confidence in economic capital results: a belief that economic capital utilization is outside management control or questions as to the relevance of the measure at lower organizational levels.

What is clear from the findings is that economic capital has become the vocabulary of enterprise-wide risk management at these large institutions, despite variations in methodologies, and that the key definitional elements of tenor and confidence levels are common currency with them. The study findings next looked at the applications of economic capital across risk types--credit, market, and operational-and in business units.

Primary Risk Types

All firms allocate economic capital to the traditional risk factors: market risk, credit risk, and operational risk.

* Credit risk still attracts the greatest amount of economic capital, for an (unweighted) average of 47% of the total.

* Market risk follows at 16%.

* Operational risk accounts for 13%.

In addition, the majority of firms define two additional risk types: equity investment risk and business risk. A significant 92% of firms attribute economic capital to equity investments (average 12%). And some 75% measure capital for business risk as a separate risk category attracting an average of 9% of total economic capital. This result represents a relatively new risk category compared to the traditional position-based risk factors.

Credit Risk: Corporate

All respondents except one employ a model-based approach to determine the distribution of potential losses on corporate loans. The KMV approach, which measures default distances, is the most popular, followed by Credit Risk+ (default volatility of cohorts), CreditMetrics[TM] (rating transitions and asset correlations), and in-house methods. In 92% of cases, the methodology includes the economic impact of downgrade risk, as well as that of default.

All respondents include counter-party derivative exposures in their model for corporate credit exposures. Respondents were evenly divided among those who 1) base their calculations on mean positive exposure; 2) base their calculations on mean positive exposure plus an adjustment factor; and 3) perform a full co-simulation of market scenarios and default losses. Only one firm uses peak exposure, but it intends to discontinue this practice.

There is no consensus regarding economic capital for defaulted loans. A wide variety of approaches were mentioned by the respondents.

Credit Risk: Retail and Consumer

The survey examined several asset classes--auto, unsecured personal, credit card, and residential mortgage. Those institutions most active in this segment use simulation models in their computation of economic capital. Others report calculations using the carrying value of the defaulted obligation, net of specific charge-offs and any risk-free collateral, multiplied by a segment-specific factor. In almost all cases, the allocated capital will vary based on the underlying risk characteristics of the asset class.

Credit Risk: Country Risk

Some 92% of respondents include country risk in their economic capital calculations for credit risk. The majority do this at the country level, determining additional economic capital requirements for cross-border exposures to measure losses due to country events.

A second approach is to use higher correlations between obligors in a particular country based on their increased chance of joint default. In some cases, regional "contagion" is taken into account, as is a scaling factor, in recognition that economic capital cycles are of different magnitudes in different countries.

Market Risk

Value at Risk (VaR) remains the most popular method of determining economic capital levels for market risk; however, 25% of respondents have moved to stress testing (in one case, in combination with VaR). None of the firms report using revenue volatility or observed losses or drawdowns as tools. Some 83% go on to use multiples to convert risk measures to the desired confidence interval, using theoretical relationships (60%) or simulation models.

Most firms (83%) calculate economic capital for activities other than trading account assets, most commonly interest rate mismatches in the loan portfolio (73%), interest rate mismatches arising from demand or time deposits (no percentage given for this), security portfolios not marked to market (55%), and foreign exchange mismatches within a business (36%). Meanwhile, 58% give an economic capital requirement to the trading account different from that given the investment security account, because of different holding periods. However, liquidity is taken into account in only a third of cases. A diversification benefit of some sort is taken by 92% of respondents.

Operational Risk

Surprisingly, given the newness of technology for the operational risk segment, 75% of the respondents report using quantitative models to compute economic capital for all or some major risk factors in all businesses. (This may be driven by Basel II considerations, at the enterprise level.) A full 58.3% use quantitative models to compute economic capital for all major risk factors in all businesses. Those not using quantitative models rely on basic indicators or qualitative factors.

Calibration of these models, however, varies considerably. Of those using quantitative models, 40% rely entirely on internal data and consider external data qualitatively; 30% explicitly include external data; and the remainder do one of three things: 1) use a combination of the aforementioned approaches, depending on the business segment; 2) rely entirely on internal data; or 3) use expert judgment to determine scenario parameters, assisted by internal and external data. This variation in response may also reflect the lack of robust, consistent pooled loss-experience data in the market.

The bottom-up and top-down approaches to the computation of economic capital for operational risk each found an equal number of adherents. Others straddled the fence, computing on a bottom-up basis for some risk factors and a firm-wide basis for others, while one respondent reported starting in the middle and allocating down while diversifying up!

Given the increasing size and frequency of fines and settlements across the industry, it is not surprising that 83% of firms include legal risk in their calculations of operational risk.

Business Risk

As might be expected with a new risk category, the definition of business risk varies widely across firms. Generally, it represents the potential for downward changes in profit caused by changing business volumes. Most firms include cost factors in this determination, as well as potential changes in revenue. And most define business risk to avoid double-counting with the traditional risk factors.

Other Risk Factors

There is little consensus regarding the treatment of reputation risk. Almost half the firms do not include it as a risk factor, and only two define it as a separate risk category. The remaining firms state that it is implicitly included in operating risk or business risk.

Given the importance of liquidity to financial institutions, it is surprising that no firms have begun to define economic capital for liquidity risk. One likely reason is the lack of a metric (such as VaR) to quantify liquidity risk. A second reason is the lack of academic research on liquidity risk's relationship to the amount or cost of a firm's capital structure.

Insurance risk, meanwhile, was included by only three firms as a risk factor, probably because only a few respondents have a material insurance business. Bank premises and equipment were noted as a risk factor by 25% of respondents. Other risk factors, such as the level of assets, were noted as a risk factor by only one firm.

Diversification Across Risk Factors

Two-thirds of the reporting organizations stated that they reduce bottom-up economic capital calculations (by an average of 21%) to reflect the diversification benefit computed across risk factors. These companies are split evenly between those that then retain this benefit at the corporate level, and those that allocate the reduction back to individual business units. These differences may have pricing implications at the business-unit level.

Conclusion

Economic capital has moved with amazing rapidity from academic construct to management tool. It provides a vital common language for those who manage risk across the enterprise and down into the business units. While credit risk remains the major element in the risk equation, other risk elements--market, operational, equity investment, and business--are now managed using a common metric, with this technology being increasingly driven down to broader levels within the organization. Nevertheless, methodologies still vary.

Follow-up surveys will further document this evolution and its impact on management in the financial services industry.

Nicholas Hayes can be contacted by e-mail at nhayes@rmahq.org.

COPYRIGHT 2005 The Risk Management Association
COPYRIGHT 2005 Gale Group