The Basic Methods of Risk Measurement and Assessment
Spurred by the financial crisis of late 2008, risk management has experienced increased importance and prominence as a function within the financial services industry. Accordingly, familiarity with the basic methodologies for measuring, assessing and controlling risk is vital for those wishing to get ahead in finance. Here we present a quick primer on key concepts in this field.
Money at Risk
The crudest, yet most conservative, measurement of risk is the total sum of money invested or loaned.
The worst possible outcome is that the entire investment becomes worthless or that the borrower defaults. A refinement is the introduction of probabilities to the analysis, but often doing requires a number of assumptions that are not strictly amenable to precise measurement. See our explanation of Monte Carlo simulations.
Limitations on the size of positions that can be held by securities traders or the amount of funds that loan officers can extend to a given borrower are, essentially, applications of this same risk reduction strategy.
Volatility and Variability
These are common measures of risk with respect to publicly-traded securities and classes of securities. Historical data can be mined to make assessments of possible future price movements, in light of past fluctuations in price. Risk measurement with respect to individual securities and classes of securities is frequently put in the context of correlations between them, among them, and with reference to broader economic indicators.
Much of modern portfolio theory, for example, involves developing strategies to reduce the amplitude of aggregate price fluctuations in an investment portfolio by selecting a mix of investments whose individual prices tend to be either uncorrelated or, better yet, to be negatively correlated (that is, their prices tend to move in opposite directions, with one being up when the other is down, and vice versa).
It has applications for financial advisors, money managers, and financial planners.
Predictive Power of History
The standard legal boilerplate on investment prospectuses cautions that "past performance is no guarantee of future results." Likewise, correlations and statistical relationships measured in some historical period offer only imperfect indications of what the future may hold for the same security or class of securities. Extrapolating historical trends and relationships into the future thus should be done with extreme caution.
Counterparty risk is the risk that the other party to a transaction, such as another firm in the financial services industry, will prove unable to fulfill its obligations on time. Examples of these obligations include delivering securities or cash to settle trades and repaying short-term loans as scheduled.
Assessments of counterparty risk often are made based on the analyses of companies' financial strength provided by rating agencies. However, as the financial crisis of late 2008 demonstrated, the methodologies used by the rating agencies are deeply flawed (as are consumer FICO scores) and subject to grave error. Additionally, in a general financial panic, events can spiral out of control very swiftly, and small counterparty failures can rapidly accumulate to the point where large firms with supposedly ample financial cushions are rendered insolvent.
Lehman Brothers, Merrill Lynch, and Wachovia were such casualties of the 2008 crisis; the first went out of business, and the others were acquired by stronger firms.
A large part of the problem with assessing counterparty risk is that the analyses performed by rating agencies are not dynamic enough. They typically adjust to new realities only relatively slowly. Furthermore, once a counterparty that previously was deemed sound suddenly lurches toward insolvency, it is extremely difficult, if not impossible, to unwind obligations and transactions already entered into under the favorable circumstances that held in the past.
The Role of Actuaries
Actuarial science, as it is often called, is an application of advanced statistical techniques to huge data sets which themselves have high degrees of measurement accuracy.
Additionally, the risk assessments made by life insurance actuaries are based on data that is almost completely uncorrelated with the financial system and movements in the financial markets. By contrast, measurements of counterparty risk, the future behavior of investment securities and the outlook for specific business initiatives are not amenable to such precise, scientific analysis. Thus, risk managers (and the management science professionals who lend them quantitative support) probably will never have the ability to develop predictive models that have anywhere near the degree of confidence that one can place in those estimated by a life insurance actuary.