Column: Risk Measurement, Risk Management

From aiCIO Magazine's June Issue: Institutional investors must understand that risk measurement is not the same as risk management. Dr. Arun Muralidhar reports. 

To see this article in digital magazine format, click here.

Mark Twain is on record stating that, “There are two times in a man’s life when he should not speculate: when he can’t afford it and when he can.” However, what distinguishes effective management of portfolios from speculation? The technical definition of speculation focuses either on the fact that it is a decision based on insufficient research or on the fact that one engages in a risky decision and is looking for a quick and considerable profit. Not much help there in distinguishing between the two as investors often make decisions without sufficient research and take very lumpy bets that could unintentionally lead to a quick and substantial profit or loss. The real distinguishing feature between the two is whether risks are managed or left unmanaged.

Sadly, too many investors have mistaken risk measurement for risk management and many vendors sell risk measurement systems promising risk management—and CIOs are left holding the bag. A risk measurement system gives you reams of reports at substantial cost, telling you how much Value-at-Risk you have in your portfolio and how much you are likely to lose at some given probability level. The fancier systems will allow you to simulate various scenarios to see how your portfolio will respond to another “Flash Crash” or “Black Monday.” This is equivalent to going to your doctor for an expensive annual checkup and being told what your probability of dying is under various scenarios, but being given no advice as to how to alter the probability.  

I recall the treasurer of the World Bank in 1997, Gary Perlin, asking us—his staff at the time—how much the World Bank’s pension fund would lose if emerging markets collapsed. Since we had already implemented what was probably the first LDI-based pension risk system at the time, we ran the numbers and gave him an estimate. When markets did collapse the following year, I went back to Gary to proudly show him how smart we had been: the actual result was very near our 1997 estimate. Gary’s prompt reply in his soft and extremely polite tone: “If you were so smart, why didn’t you make me money?” Never again did I confuse risk management with risk measurement.

A good risk management process distinguishes between good risk and bad risk or the likelihood that a position in the portfolio, whether overweight or underweight, is likely to add or detract value given the current economic and financial environment. Therefore, rather than blindly trying to reduce risk, whether defined as absolute risk or relative risk, a smart investor tries to enhance the good risk and reduce the bad risk just like a doctor encourages us to raise our good cholesterol and lower the bad cholesterol. In this issue, aiCIO profiles how San Bernardino County (SBCERA) altered their rebalancing process and details why their approach was effectively good risk management. The traditional approach to rebalancing was to roll up the total portfolio asset allocation tilts and if they did not breach the rebalancing ranges nothing was to be done. If the ranges were breached a rebalancing trade brought the portfolio back to the policy weights. Contrary to popular opinion, one could argue that traditional rebalancing is effectively speculation, for the act of doing nothing—or even selling/buying the overweight/underweight assets—is effectively taking a bet not based on an analysis of the relative attractiveness of the assets, but on some arbitrary ranges. A risk measurement system might give the appearance that these rebalancing trades are lowering the relative risk of the portfolio, but if the investor is buying an asset likely to plummet in value, this is terrible risk management. 

SBCERA effectively applied the same techniques to the asset allocation decisions that they expected of their asset managers; namely, research (ex-ante) the factors that influence the likely future performance of the assets in their portfolio and then use this research to make intelligent, explicit rebalancing decisions ex-post based on the relative attractiveness of the assets. Interestingly, all the ideas used by the staff to guide these decisions came from publicly available peer-reviewed articles. 

We live in a world of increased volatility. The need for good risk management has probably never been greater. Implementing a risk system focused on measurement—and not management—will lead to poor performance and wasted resources. A closing anecdote: A large public fund spent millions to set up their risk measurement system. When the frequency of the reports declined from daily to weekly to monthly, none of the senior officers complained. Effectively, the reports produced from this system did not inform investment decisions and hence were practically worthless. Instead, CIOs should conduct in-depth research to calibrate asset class and manager rebalancing decisions based on the current economic and financial data (not arbitrary ranges) to emphasize good risks and shed bad risks. To not do so is to succumb to Mark Twain’s equally well-known quote about speculation: “October. This is one of the particularly dangerous months to speculate in stocks. The others are July, January, September, April, November, May, March, June, December, August, and February.” 

Dr. Arun Muralidhar is Chairman of Mcube Investment Technologies LLC (www.mcubeit.com), and CIO of AlphaEngine Global Investment Solutions (AEGIS).

«