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

251

Perform SPC Analysis

Because of the uncontrollable variations of the inputs from the real world, a trading/investment

system can never achieve a stable output process. This is why we use SPC. The instability of

inputs leads to variation in the output factors of the system. (This is similar to variability in a

manufacturing process, such as the diameters and lengths of parts.) We apply the processes

used to control machinery in manufacturing to control trading/investment systems. Statistical

process controls around input data (CPI, PPI, VIX, trading volumes, etc. as defined in K | V

Stage 2) and outputs (percent winners, Sharpe ratio, sector weights, etc. as defined in K | V

4.2) can tell risk managers when the system is experiencing common variation and when the

system is out of control.

CHAPTER ◆ 28

FIGURE 28-1

Determine

causes of

variation

Manage portfolio and risk

Plan

performance

and risk

processes

Define

performance

controls

3

2

(^1) Perform
SPC
analysis
Just like in manufacturing, when monitoring of a process uncovers excessive variation
in a trading/investment system, a manager can either shut down the machine, or have it
monitored more closely until investigation uncovers the root cause of the excessive varia-
tion. (Shutting down a system means only closing trades. A good trader, like a good pilot,
can essentially land the plane safely after the autopilot is switched off. A human trader
adds value during disasters and abnormal environments.) The main difference between
manufacturing and systematic trading is that in manufacturing the monitoring is done
by sampling finished pieces (i.e., a random sample is used as the basis for accepting or
rejecting an entire lot), whereas in systematic trading/investment managers can perform a