Risk:
Risk is one of the most misunderstood notions. This is paradoxical because a primary purpose of the financial services industry is the control of financial risks. Popular misconceptions include:
Let's start with a definition: Risk is exposure to uncertainty. Accordingly, it has two components: (1) Uncertainty and (2) Exposure to that uncertainty.
For example, if a man jumps out of an airplane with a parachute on his back, he may be uncertain as to whether or not the chute will open. He is taking risk because he is exposed to that uncertainty if the chute fails to open, he will suffer personally.
In this example, a typical spectator on the ground would not be taking risk. They may be equally uncertain as to whether the chute will open, but they have no personal exposure to that uncertainty. Exceptions might include:
Such spectators do face risk because they may suffer financially and/or emotionally should the man's chute fail to open they are exposed to the uncertainty. The financial services industry is primarily concerned with financial risk that is financial exposure to uncertainty.
A synonym for uncertainty is ignorance. We face risk because we are ignorant about the future after all, if we were omniscient, there would be no risk. Because ignorance is a personal experience, risk is necessarily subjective. Consider another example:
In this example, there are two individuals exposed to the same event. Because they have different levels of uncertainty, they face different levels of risk. Risk is subjective.
Some of the most significant risks which organizations face are highly subjective.
These include:
Institutions can actually reduce these risks simply by researching them. A bank can reduce its credit risk by getting to know its borrowers. A brokerage firm can reduce market risk by being knowledgeable about the markets it operates in.
In financial applications, quantitative models are often used to try to give objective estimates of risk. Examples include the Capital Asset Pricing Model and value at risk models. While such models are extremely valuable in managing risk, they are just objective estimates based on the model's subjective assumptions. Those assumptions, such as lognormality, homoscedasticity or continuous trading, may be reasonable in any given instance, but are still subjective. For example, if a model says that there is no risk, does this really mean that there is no risk? What if the model is wrong? This subjective component that exists for every objective measure of risk is called model risk.
Risk is a personal experience, not only because it is subjective, but also because it is individuals who suffer the consequences of risk. Although we may speak of organizations taking risk, in actuality, organizations are merely conduits for risk. Ultimately, all risks that flow through an organization accrue to individuals: stockholders, creditors, employees, and customers, board members, etc.
One of the fundamental challenges of enterprise risk management is the fact that individuals who take risks on behalf of an organization are not always the same people who suffer the ultimate consequences of those risks.
Finally, the notion that expected excess returns are proportional to risks taken is simply wrong. This popular notion stems from the Capital Asset Pricing Model which, based on some broad simplifying assumptions, and draws the conclusion that the market compensates investors for taking systematic risk. While the theory provides valuable insight into the workings of financial markets, it was never intended to support companies in managing complex or illiquid risks. Just as there is no positive expected return from gambling or jumping out of airplanes, there is no positive expected return from specific risk or from corporate risks resulting from fraud, mismanagement, inefficiency or misunderstanding. These are some of the significant risks which enterprise risk management seeks to address.
Capital Asset Pricing Model divides equity risk into two components:
For example, if the stock market rises upon the release of good economic news, all stocks are more or less affected. This is systematic risk. On the other hand, if a company's stock rises upon the success of a new product line, that is specific risk. It affects only that company's stock.
Specific risk can be diversified away. As more and more different stocks are added to a portfolio, the random fluctuations unique to each stock start to offset one another. If diversification is taken to an extreme, the investor is left with a portfolio who's composition corresponds identically to that of the overall market. Such a portfolio has no specific risk. Because its composition is the same as the market's, by definition, all of its risk is systematic. Systematic risk can never be diversified away.
Beta measures a portfolio's (or an individual stock's) systematic risk. It is defined as:
Beta = (s p/s m) r p,m [1]
Where s p and s m are the return volatilities of the portfolio and market respectively, and r p,m is their correlation. Beta is usually calculated from daily return data. For this purpose, a broad market index such as the S&P 500 or BSE Sensex or NSE Nifty are often used as a proxy for the market.
Beta measures the tendency of a portfolio to participate in market moves. For example, suppose a portfolio is 1.5 times as volatile as the market, and has a correlation of .4 with the market. Then, by Equation [1], the portfolio's beta will be 0.6. The portfolio will tend to gain 6% for each 10% gain in the market—or lose 3% for each 5% loss in the market.
Suppose a portfolio has twice the volatility of the market and has a correlation of .8 with the market. Then its beta will be 1.6. It will tend to participate in market moves 160% of the extent to which the market moves.
Obviously, the beta of the market is exactly 1.0.
The Capital Asset Pricing Model states that, because specific risk can be diversified away, the market will not compensate investors for taking it. A stock's excess expected return (above the risk free rate) will be proportional to its systematic risk—its beta. This is expressed mathematically:
E(Rp) = Rf + beta (E(Rm) - Rf) [2]
Where Rf is the risk free rate, and E(Rp) and E(Rm) are the expected returns on the portfolio and the market respectively.
Beta is sometimes used as a measure of a portfolio's risk. For highly diversified portfolios, this can be appropriate because systematic risk is the primary source of risk for such portfolios. For less diversified portfolios, however, specific risk is more significant. For such portfolios, beta can be a misleading measure of total risk. Although the Capital Asset Pricing Model states that the market will not compensate investors for taking specific risk, this does not mean that specific risk is not real. A dollar lost to specific risk costs just as much as a dollar lost to systematic risk.