Quality Money Management : Process Engineering and Best Practices for Systematic Trading and Investment

(Michael S) #1
Very interesting is that some of the best funded pension plans in the world do not
believe that a firm can significantly outperform its benchmark without the addition
of hidden risk, while individual investors believe a firm can shoot for the moon to
generate 50% returns a year with no additional risk. Our view is closer to the pension
fund world than the individual investor ’ s world. All successful trading/investment
ideas become known to smart investors quickly. Smart investors enter the area of
excessive returns quickly due to their superior ability to gather information. Very
quickly an informal benchmark is created by the smart money to measure which
firm to place capital with. Once a benchmark has been formed and outside capital
is put to use, the focus shifts from absolute returns to asset gathering to generating
fees. The best way to gather assets is to beat the benchmark with lower risk.
● Value at Risk (VaR). Value at Risk can be viewed as a forecast of the future return
of a system. Risk managers choose a distribution since they must believe a substantial
amount of the process variation remains in the system. The purpose of the forecast is
to determine what the worst case is given a predefined distribution and threshold. If the
trader hits this predefined threshold then the risk department forces the trading desk
or group to post additional margin or to close down the desk. This is done in the third
spiral since we will use this to determine if the underlying process variation is stable
or shifting. A process standard deviation should be stable as measured in the VaR cal-
culations. If a process ’ s standard deviation shifts from a low volatility regime to a high
volatility regime, risk managers and kaizen teams will need to quickly determine its
effect on the output of the algorithm or even better switch to a high volatility algorithm.

The real problem with all three levels of risk management is that many traders and money
managers view the risk department only as a necessary evil. They do not want to help the
risk department since this could lead to lower bonuses due to risk-based haircuts. Some trad-
ers believe a major part of trading is hiding profits and risks from the risk department, done
by shifting inputs in theoretical models. The goal is to have money for rainy days where the
market goes against you so you can smooth out your returns. This cat and mouse game has
been played for years by the traders and risk managers. In fact the risk managers unofficially
encourage this game since they are the ones that have to discuss with senior management any
unexpected losses. By smoothing out the returns everyone wins, according to many traders.
The problem with the smoothing out of the returns and managing expectations is we live
in a stochastic world. We should be using the stochastic outputs to improve the process by
identifying shifts in the investment environment early so we can do something about it before
everyone in the peer group loses money. Sometimes the way to win is not the best but may
avoid periods where everyone loses money by closing down or scaling back a trading algo-
rithm (during a period when it is no longer predictive) or switching investment styles.
Academics have a static view of risk, one that claims to explain the majority of the
variation of returns. In such a view, risk is a snapshot of a stochastic process that is
assumed to be stable and well defined. In the nonsystematic trading/investment world,
risk managers assign abnormal, or assignable, variation to a person. Large downside devi-
ations, that is, drawdowns, are pinned on people—a trader or a portfolio manager. The
normal reaction is to reduce trading limits until the person recovers his emotions. (Think
red bead experiment. In a red bead experiment, several players (employees) select ten
beads from a box containing 10% red and 90% white beads. A manager then yells at the
employee who draws the most red beads in the hope of improving his or her performance on
a subsequent trial. Deming used the red bead experiment to illustrate typical management
response to common variation in a process.) In systematic trading/investment, however,
there are no emotions in periods of drawdowns and no people to blame for outliers.


25.1. HISTORY OF RISK AND RETURN 231

Free download pdf