In earlier parts of this series I discussed various types of fund managers and how to spot effective ones. Another key role of a CIO is to monitor and judge risk, in all its forms.
There seems to be as many ways to think about risk as there are stocks to choose from, but inevitably one must focus on a few measures that prove reliable over time.
Surprisingly, these are not all statistical in nature. These need to be supplemented with intuition, rules of thumb and above all else, common sense. It is important to know not only the risk in the portfolio, but how a fund manager views and thinks about that risk. This can help us understand the difference between confidence and arrogance.
Value-at-risk (VAR), beta, tracking error and correlations all come from the same family of statistical practice. They take a specific historical period and describe the movement in returns over that period using statistical techniques.
The period chosen and its length can have a large impact on the result of the measurement. Add this to the fact that short periods, such as 12 months, are almost meaningless statistically, and one can see that knowing what lies behind the numbers is vital to understanding their usefulness.
Measuring risk is one thing, but understanding how it is perceived is also important. Two perspectives are important here, the way the fund manager perceives risk and the way the client does. Often, these can be at odds.
It’s not all about you
The CIO’s task is to make sure that the manager understands risk from the client perspective as well. If the client thinks they are on a pleasure cruise and the fund manager believes they purchased a rollercoaster ticket, the experience will not be happy for anyone involved.
It is remarkable how many times I have had to tell managers that while they may believe firmly in their views and positions, their risk positions were not in keeping with the clients’ expectations. I actually had one fund manager tell me, ‘well then, they are the wrong clients’!
Another manager was famous for having huge upswings and huge downswings, holding on to positions doggedly until they came right.
On average, they added some value, but the clients attracted during the up years were inevitably shaken off during the down years. The assets he managed would grow dramatically and then shrink dramatically.
True, there was a core group of clients who knew what they had bought, but it was too small a clientele to form a profitable core, and the result can be damage to the reputation of the firm overall. The best managers know their clients’ view of risk and incorporate that into their portfolio construction.
Of course, the flip side of this is the manager who takes little or no risk. These are the closet indexers that are becoming the focus of regulators.
It is an unfortunate truth that a little alpha goes a long way, and as long as you don’t raise eyebrows, you can run a profitable fund for a very long period of time.
Profitable from a business point of view, but not necessarily a client’s one. A good CIO needs to encourage appropriate risk taking, which can result in decreasing or increasing risk levels.
Risk should reflect the confidence of the manager in their views and be scaled appropriately. I am not of the school that says managers should always have a given level of risk, though. I have heard people argue that ‘our clients pay us to take risk’.
Actually, they pay us to add value. At any given time, if we do not believe the value available warrants the risk, low risk might be appropriate.
Over time, however, risk should reflect the alpha target. The oft quoted fundamental law of active management (Richard Grinold, 1989) basically states that the return per unit risk is directly related to the accuracy of the manager and the number of positions held.
To generate a given amount of return per unit risk, a manager with a small amount of information (say they are correct 52% of the time) needs many positions in their portfolios.
A manager with a higher degree of insight (say a heroic 70%) needs fewer positions in their portfolios to gain the same return per unit risk – on average. The key is the ‘on average’ part. You have to be able to take these risks over and over again.
If your client fires you mid-way through the process, your actual accuracy is irrelevant. I have had to have discussions with managers pointing out to them the fact that the size of a given position reflects a degree of accuracy that is highly unlikely if not impossible.
Sensible scaling of risk within reasonable bounds of confidence is appropriate. Anything beyond that is hubris and is usually appropriately rewarded with the anger of the gods.
Risk is often spoken about but rarely means the same thing to two people. Between the inherent problems of the statistical measures, to the difference in perspectives of managers and clients, it is useful for someone like the CIO to take a more holistic view of portfolio risk, considering the rules of thumb their experience has taught them, the knowledge of the client’s (or the product’s) risk appetite, the fund manager’s degree of insight, and the general market environment.
Ultimately, portfolio managers are focused on adding value and having someone between them and the client is an important part of the checks and balances which make for successful investment products.
The previous three parts of the series can be found below: