Now that we have looked at various kinds of risk here and here, the question arises as to how we can measure or quantify some of them so as to enable us to monitor and manage these risks. There is, however, one key concept that we would like to address before we get there – the difference between ex-ante and ex-post risk.
Ex-ante (before the event) and ex-post (after the event) are terms used to describe the timing of risk measurement. Imagine that an analyst is evaluating a company and she looks at all the past historical data available to her and comes up with risks that she sees based on them, these would be ex-post risks. She would also make some scenarios and assumptions about how the future might pan out and how certain aspects of the business may or may not be affected going forward – these would be ex-ante risks. Ex-ante is forward-looking- it is basically an estimate of a portfolio’s risk given the current positioning of the portfolio; while ex-post risk measures are calculated after the fact, based on the actual performance of the portfolio.
Ex-ante risk measures can be useful for planning and decision-making, while ex-post risk measures can be useful for performance evaluation. To simplify, one could think of them as expected and realised risks, and always appreciate the fact that these two can differ widely, in some cases even for long periods of time. Now let’s take a look at some of the popular risk measures that are used to measure portfolio risk, most of which can be used ex-ante and ex-post.
Volatility: The most common proxy for risk across asset classes and investment products, this is generally calculated using the standard deviation of returns. This is easy to calculate as one only needs prices or returns for a period of time. This measure basically tells you how much the performance is likely to vary from the mean and a lower number implies lower risk. There are certain modifications like semi-standard deviation which are also used by certain practitioners.
Value at Risk (VaR): This is a useful measure to look at in an ex-ante setting as it calculates the potential magnitude of an extreme loss in the portfolio. VaR is calculated using a specified time period for risk assessment, historical data for portfolio returns and a confidence interval. If the portfolio has a 99% 1-day VaR of -100,000, there’s a 1% probability that the portfolio will lose more than 100,000 over a day. The general argument against VaR is that it greatly underestimates tail risks.
Drawdown: One of the favourite measures for any portfolio manager, drawdown measures the decline in portfolio value from its prior peak and serves as a very useful measure for comparing strategies and products. Maximum drawdown is the historical maximum peak to trough decline that the portfolio might have had and gives an indication of the risk inherent in the strategy. The advantage of a drawdown measure, is that it is not dependent on a time window for calculation and gives you an estimate (with some confidence) of how much the portfolio value can fall from a given point.
Beta: Beta gives an estimate of the riskiness of a security or portfolio relative to a market-based benchmark. For e.g., a stock with a beta of 1.5 to an equity index is likely to be more volatile than a stock with a beta of 1.2 and each of them individually are likely to be more volatile than the index as a whole. It is a useful measure to compare long only portfolios as it gives a sense of how much additional systematic risk is present in the portfolio.
Tracking error: Commonly used in passive portfolios, this measures how different the portfolio is from a reference benchmark and is calculated as the standard deviation of excess returns relative to the benchmark. This gives an estimate of the range of out performance or under performance that is likely to occur relative to the benchmark. For passive funds, a lower tracking error is always better.
Duration: Typically applicable to portfolios that invest in debt securities, duration measures the sensitivity of the portfolio to changes in interest rates. Longer maturity securities generally have higher duration risk when compared to shorter maturity ones and hence a rise in interest rates will lead to a larger fall for the former than the latter.
Shock risk: This is the risk of there being a large discontinuity between 2 consecutive trades (prices) of a security which might happen between the closing and the opening of the next day (or session), but can also occur during continuous trading hours due to an extreme market event or due to market disruption. Option writers would calculate anticipated losses in various such ‘gap’ scenarios and adjust their position sizes or manage risk accordingly.
An investor might use one or more of the above to analyse the historical performance of her portfolio or to gauge the risk of a prospective investment product. Investment managers on the other hand would use multiple different measures to assess and analyse prospective risks in their portfolios and hedge out or reduce certain risks.