LEVEL: INTERMEDIATE
Portfolio managers will think about risk management routinely as part of their investment strategy, and they will strategize about managing risk on a case-by-case basis. Sometimes this means minimizing risk, but sometimes it also means calculating how much risk is acceptable to yield a higher return on an investment. One of the most widely used metrics for managing risk is a concept called “value at risk” which describes the amount of capital that can be gained or lost given a specific fluctuation in a portfolio.
From a historical perspective, the concept surrounding value at risk was not actively used prior to the middle of the 1990’s. Prior to that, there was no set standardized method to risk management, but market participants felt the need to conceptualize the notion of value-at-risk as a response to a historical financial crisis.
To measure value at risk, portfolio managers and investors will examine the amount of a potential loss, the probability of that loss happening, and the time frame in which that loss may happen. With these three variables, investors can determine how much a portfolio or investor can expect to lose in a certain period of time.
The SEC initiated the first capital requirements to losses that would occur within a 2-standard deviation move, which equates to a 95-confidence band. Historical returns were the first data series used to calculate reserves and capital requirements. The reserves were then used to cover potential losses from an outside movement in financial markets.
Measuring Value at Risk
There are three basic methods used to determine VAR, along with a plethora of variations on those basic approaches. An analytical solution can be used by measuring the variances within a portfolio. VAR can also be estimated by running a historical data analysis or from Monte Carlo simulations. These are a random sampling of values within a time series. Monte Carlo methods are often used in simulating physical and mathematical systems to see the likelihood of them yielding a particular return.
Drawbacks of Value at Risk
While VAR is a quantitative tool built upon strong fundamental modeling data, it is by no means the final word in identifying just how likely a loss is to occur in a fixed period of time. This is because historical data is never a completely reliable predictor of future losses and gains, since markets and economies are always in flux due the unpredictable nature of human behavior.
While some behavior can be modeled, there is no way to guarantee that a certain portfolio will have a certain risk profile, because unexpected disruptions in the market and in individual investments can never be fully accounted for. When using historical simulations, the assumption is that the past reflects events in the future, but as we have seen with the recent financial crisis, that is not always the case. History may not a good predictor and every simulation uses some kind of historical data as a way of generating future data points.
This is one reason that several have been skeptical of VAR both in terms of accuracy and as a tool for managing risk. Nonetheless, the concept has acquired a strong following in the risk management community.
Beyond the unpredictability of markets, there are other reasons to be skeptical of the idea. There is no precise measure for Value at Risk, and each measurement comes with its own limitations. Value at risk can be the wrong approach for certain portfolios, since liquidity along with the lack of specific data can generate varying results. Every calculation makes an assumption about future outcomes, and the more assumptions there are in a model, the less reliable it tends to be. Additionally, value at risk generally deals with normally distributed instruments, which will not accurately reflect most portfolios.
Another problem with value at risk is that it has a narrow definition that fails to incorporate numerous other issues related to risk management. Using VAR as the only measure of risk can create a false sense of security about the positions held within a portfolio, which can sometimes be more dangerous than having no risk management strategy at all.
Risk Analysis
Value at Risk measures the likelihood of losses to an asset or portfolio due to market risk. This is a narrow definition, and it excludes credit risks and liquidity risk. Additionally, VAR is only looked at as a negative, which eliminates positive outcomes. A more powerful estimate of risk could be achieved if political risk, liquidity risk and regulatory risks where incorporated into the calculation. The problem with many of these, particularly political risk, is that they are not quantifiable.
An example of liquidity risk could be seen If bid/offer spreads become wider, causing a lack of liquidity for a product (as seen during the financial crisis when the Libor moved 500 basis points above T-bills). When this happens, the ability to exit a position becomes harder, so liquidity is reduced. This risk is exacerbated by the size of a portfolio; the larger the position in a security, the more vulnerable it is to a lack of market demand.
Efficient Measurements
Value at Risk can be computed over multiple time frames, with shorter term movements preferred over longer term movements. The need for specific data to be measured on a daily basis given recent events has created a risk environment that is cognizant of short-term movements within a market. Financial services firms use VAR to generate daily hedges which has generated a focus on daily data. Regulators have also requested daily updates since 2008, which also puts a focus on short-term data.
Summary
Value at risk is widely used within the risk management industry as a benchmark to gauge market risk. Although it has numerous drawbacks, it is only one of the risk measurements used by investors to assess the risk of a portfolio. Risks related to the beta of a portfolio along with benchmarking the returns are not incorporated into a standard VAR model. Given its wide breadth, it is important to understand the value of the calculation along with its abilities to assist in measuring a portfolio’s exposure.