This is the third installment of our seven-part series. In this series, we have taken a white paper that we wrote together and broken it into standalone issues so that you can easily read them in one sitting. In this article, we ask, “Why did it appear that firms took a narrow view of the risk versus reward tradeoffs in their portfolios?”
Based on our collective experience and conversations with banks and other financial services companies, three key issues continue to surface:
- There was limited use of extensive analytics; instead, the industry relied on a few quantitative measures, which they thought were sufficient.
- There was an overreliance on diversification.
- There was an assumption that capital could be raised quickly, if necessary, and that the markets had abundant liquidity that would remain available if needed.
Initially, the reliance on a few quantitative techniques that were directed at determining the maximum exposure at the portfolio level created the common view that these techniques captured the risks of the entire firm or the collection of instruments; what was ignored was considered immaterial or just noise. For example, although value at risk (VaR) failed miserably during the economic crisis of 2008, VaR still remains one of the leading exposure quantification techniques.
This reliance on a limited set of exposure measures resulted in an understated view of how large an exposure could potentially become. Additionally, the calculation of exposures looked at an isolated portfolio view only – e.g., a division of the bank such as the mortgage portfolio – and did not account for any cross-portfolio interactions or dependencies. The reason firms measured risk this way was partly due to the nonadditive nature of VaR (single portfolio view of exposures) and limitations in a firm’s ability to calculate firmwide exposures. To implement VaR, they had to deal with the traditional limitations in computational technology and the ability to gather and aggregate portfolio data, and the inability to model exposures across disparate financial products. To gain more precise views of the portfolio exposures, firms narrowed the measured “path of risk” to something they viewed as accurate within a narrowed set of products and instruments.
The view was flawed
Confidence in this narrowed view of risk led to the premise that if a firm was to extend its current path of risk, it could be quantified and managed via a combination of capital reserves, hedging and derivative instruments to limit potential losses. The market volatility was estimated by narrowing the market components of the underlying risk. It was assumed that market effects would be neutralized by the shock absorbers of capital and hedging strategy.
The second view was that the risk could be diversified across many instruments, product portfolios or paths of risk, and that the paths from a market volatility view would never align in a congruent manner. The thought was that correlations between market sectors would remain low, and if they were to increase, an appropriate response to the volatility could be both estimated and enacted within a time frame that would allow the firm to control or hedge the risk. If the firm could not respond quickly enough, then VaR delineated the maximum exposure.
In the extreme case of multiple losses across divisions of the firm, the hedging mechanism and capital cushions were in place to act as shock absorbers. The prevailing thought was that one could retain the alpha components and hedge or shed the beta components of investment instruments via a set of structured financial products. Because volatility in the market was assumed to be low and would remain so, little consideration was given to the possibility that the “shedding” mechanisms were incomplete, and during shocks the risks of the structured investments would return to the firm. These risks were assumed to be inconsequential and manageable.
The third view was that market liquidity was ample and was largely available, or that the probability that liquidity would diminish or be unavailable was extremely small. If asset values fell and additional capital was needed, it could be generated via debt structures or by selling or structuring asset holdings or, if need be, by increases in preferred or equity capital. If an investment or risk position was to be exited, the market’s ability to absorb the instruments used to structure the products was assumed to be available. The notion of having to wind out of a particular position, instrument or product type when markets ceased to function was not a prime consideration. This cost of adjustment was assumed to be small because of capital cushions, adequate hedges or other structured investments.
Clearing the view
The notion that market liquidity is highly correlated to the overall performance of the market was not self-evident. When markets are performing well, capital and liquidity flow in and are readily available; when markets underperform or are in shock, capital exits the overall system. This arises because intermediaries are no longer confident in their ability to assess values or that liquidity demands are temporary. They are unable to distinguish between price movements caused by liquidity or valuation changes.
Additionally, leverage ratios at financial firms tripled over the last decade, and this debt-based nature of the capital used to add to on- and off-balance sheet assets was not sustainable. Once the market declined, the capital that provided the liquidity in the market was not replaceable. And, as we have seen in the first part of 2010, sovereign nation debt concerns have constrained the flow of capital within the system, limiting access to capital in some markets. Any forward-looking framework for the future estimation of risk versus reward will not only have to account for the amount of available capital, but also incorporate leverage adjustment costs and the duration mismatches of assets and liabilities. Estimations of market, sector and firm-level leverage should be accounted for, along with where the capital is being “pushed” in the market, within which sectors of the market or simply out of the market. This view will also help to provide better estimates of market volatility.
Future regulatory changes will likely address these issues. In our next article, “Risk management: Shifting from a portfolio-theory view to an optimization view of risk management,” we begin discussing the changes that must take place in the evolution of risk management. Management of risk must evolve to include advanced analytics. Our next article explores that paradigm shift.
If you missed the first two articles in the series, here are links to help you catch up: