Perspectives on Risk

Risk and reward are the central concepts of investing.

There is a lot to unpack in such an innocuous seeming statement.

First, it is absolutely true that it is the focus on risk that separates good investors from poor ones. No matter how good your edge is, if you go bankrupt through poor risk control, you won't be able to capture the money. Successful trading is a long term game. 

But from that fact, a number of incorrect conclusions get drawn. Risk control is not all you need. A seemingly prudent rule like "only risk 2% on any trade" is both wrong in its arbitrariness and misses the point that if the trade has no intrinsic edge, this rule will just lead to bankruptcy one 2% loss at a time. A related piece of stupidity is that discipline is what will lead to success. Executing a trade with no edge in a disciplined manner is worse than executing it in a haphazard manner. If you are inconsistent, at least you will avoid some of the inevitable losses.

But let's assume you have some edge. What is risk?

Practitioners tend to think of risk as the probability of losing too much money. There are several problems with this definition:

  • The definition is entirely subjective. What is "too much"?

  • Even in retrospect we cannot know risk. Clearly if we make money on an investment it doesn’t mean that there was no risk involved.

  • It is tough to solve any concrete problems with this definition.

Academics usually define risk mathematically, e.g. as the standard deviation of returns, i.e. the fluctuations around the mean return. This also has problems:

  • We need to choose the estimator, the length of time series, and the time scale (e.g. daily or monthly returns).

  • We need to estimate future risk from historical returns, and we know that the future is never like the past.

But this definition has the advantage that the mathematics of many interesting situations such as portfolio optimization and option pricing become tractable. The researchers who proposed this definition were not stupid. They understood its limitations but felt that the ability to create stable models outweighed the problems.

In practice we tend to combine the two concepts anyway. We conceptualize in terms of the practitioners view but do calculations using the academics’ version. Also, academic approaches tend to be about measurement rather than management.

Over the next few weeks I will be writing about risk. Specifically,

  • Why it is essential to know the historical numbers.

  • How to really manage market risk.

  • Why market risk is far from the riskiest risk.

Stay tuned!

Euan Sinclair1 Comment