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

237

Plan Performance and Risk Processes

CHAPTER ◆ 26

The inputs into a trading/investment system, that is, into the trade selection, execution and

position, and risk management algorithms, are data—price data, valuation data, fundamen-

tal data, calculated data, and economic data. In basic statistical terms, all the trading algo-

rithms are trying to forecast future performance using stochastic, independent variables. In

manufacturing, effective coefficient of determination ( r 2 ) is very high, in excess of .95. In

finance, they are much lower, normally less than .60. The goal is to use both risk manage-

ment and Six Sigma techniques to increase the effective r 2. The unexplained variation is the

cause of uncontrollable risk. This risk can and has been catastrophic to trading/money man-

agement firms. The variation a model does not explain is itself stochastic—what we call

common cause variation in the outputs of the system, or the unpredictable part of the profit.

Define

performance

controls

Perform

SPC

analysis

Determine

causes of

variation

STAGE IV Manage Portfolio and Risk

1

2

3

Plan

performance

and risk

processes

FIGURE 26-1

A risk manager attempts to understand how and why a trading/investment system is

making and losing money and what the potential for loss is. He or she currently does this

using three tool sets:

● Single performance measurements.

● Performance attribution.

● Value at Risk.