In the hedge fund marketplace, nothing is more central than risk. Historical data is widely used as a basis for risk assessment, particularly to predict the current stress/beta/Value-at-Risk characteristics and future potential paths of the assets and portfolio. However, solely relying on historical data is not enough and thus it’s very important to rely on expert opinion to interpret the data in context with the investment strategy and market conditions. Here are some ways historical data can help you asses risk:
- It helps you determine how a portfolio behaves under stress so you can predict future performance, however its accuracy depends on the volume and quality of historical data.
- It helps you identify key betas that need to be avoided in context with investment objectives and market sentiment.
However, historical data has a number of limitations that need to be considered when being used to asses risk for example:
- Beta dependent stress testing of an asset and portfolio by using market conditions today and stress charecteristics from historical data might only be marginally useful as correlations between assets change over time..
- Statistical information is not always available on every kind of incident, so one can’t determine the exact rate and severity of occurrences of all incidents. Furthermore, the impact of the consequences is often quite difficult to evaluate for intangible assets.
- The accuracy and cleanliness of the information collected. Professionally usable data needs to be clearly distinguished out of the massive yet inapplicable information and one has to be careful to not “data mine.”
In order to improve the way we assess risk when working with historical data, risk measures should be complemented by information from hypothetical scenarios. Forward-looking information reflecting expectations of market participants such as implied volatilities should be used together with statistical estimates (which are necessarily based on past information).
Finally, you need a trained eye to interpret the data and thus utilizing an educated opinion of a risk management expert can play a crucial role, when working with historical data.
