Skip to main content
Track System optimization
4 / 23
Search the archiveLibrary
Library

1988issue C041-10

Simple systems over optimized complexity

The archive treats computers as verification tools for already chosen rules and describes extra rules and harvested price patterns as a move away from generality. Editorial: a trading procedure should stay general enough to survive unseen markets; complexity that only explains the past is a design failure, not sophistication.

  • Computers verify already chosen rules; they are not engines that invent a trading system by searching combinations.
  • Adding rules to patch historical failures, or letting a computer harvest ever-longer price patterns, moves a system toward a paper result that is unlikely to hold in unseen conditions.
  • Simpler procedures can look worse on historical tests while offering a more reliable correspondence between past and future behavior than spectacular in-sample results.
  • Slow moving-average trend systems are preferred because they are said to track more persistent market events than short-term trends.
Entries in this reading3 entries
Current trackSystem optimization131 readingsTrend following168 readingsMoving average597 readings

Verification, not invention

Computers are treated as verification tools for already chosen rules, not as engines that invent a trading system by searching combinations. System optimization, in this workflow, checks a procedure the operator has already specified.

The revised systems text is described as expanding coverage of optimization and of many systems per category while shortening background mathematics now handled by programs.

Generality versus paper results

Repeatedly adding rules to patch specific historical failures is described as moving a system away from generality and toward a paper result that is unlikely to hold in unseen conditions.

Letting a computer harvest ever-longer price patterns is characterized as extreme overfitting: the procedure can always invent a new pattern and therefore never appears wrong.

Simpler procedures are said to look worse on historical tests while offering a more reliable correspondence between past and future behavior than spectacular in-sample results.

Moving average speed and persistence

A prior practice of retuning moving-average speed to keep a system aligned with changing markets is reported as having failed in use.

Slow moving-average trend systems are preferred because they are said to track more persistent market events than short-term trends, which are judged too inconsistent to rely on.

Trend following entries and risk control

A trend-following entry is described as traditionally receiving a poor fill. Scaling into more than one contract is presented as a way to tolerate mistimed first entries without abandoning the longer-term position.

Risk control is framed as knowing how the system behaves. A rigid stop is advised when expectations are unclear. A trend change or an unusual reaction to news may replace a fixed stop when the operator understands the risks.

Combining techniques from different areas is offered as a design option, such as entering on a moving average and exiting with a chart support or resistance level or another independent exit method.

Educational research material, not investment advice. Historical source context does not establish present-day performance.
4 of 23 in the System optimization track
19881-1 pp.Next on System optimizationCycle rule inputs until the parameter search stopsA multi-parameter search can begin with a user-specified rule-input set, score that starting combination on a historical quote file, and vary the first parameter until a preferred value is located.
All readings on this track · 23 readings
  1. 1984Evaluating mechanical systems beyond net profit
  2. 1987Mechanical systems need more than pretty charts
  3. 1987Evaluating silver moving averages and RSI by scoring rule
  4. 1988Simple systems over optimized complexity
  5. 1988Cycle rule inputs until the parameter search stops
  6. 1990Split-sample tests for stochastic parameter grids
  7. 1993Four-subsystem trading system construction
  8. 1993Randomized price paths for stock system parameters
  9. 1994Mechanical timing with out-of-sample and filter checks
  10. 1995Evaluating asymmetric percent-reversal trading rules
  11. 1996Evaluating a daily volume-strength and new-high new-low timing procedure
  12. 1997Rule-based intraday system design with genetic search
  13. 1997Evolving three-rule systems with genetic templates
  14. 1997Audit the holdout and the exposure before the fitted entries
  15. 2001Inner weights are part of the candlestick signal
  16. 2003Mechanical short-term rules for news-sensitive rallies
  17. 2004Band construction as a testable trading procedure
  18. 2006Layered day-session automation from brackets to rules
  19. 2006Why mechanical rules must be retuned per security
  20. 2006A linear-regression next-bar forecast needs a searchable cutoff
  21. 2015Construct if-then rules, then backtest and optimize
  22. 2015Mechanical system design as a critique of discretion
  23. 2025Joint channel design with a trend filter and stop-loss orders
By fieldSystem design203 readingsBy year1988116 readings
All 131 readings tagged System optimization
Also on System optimization5 readings
1984Evaluating mechanical systems beyond net profit1987Evaluating silver moving averages and RSI by scoring rule1989Binary search limits on multi-peak average grids1989Four-span MACD lookbacks as perishable parameters1989Parameter neighborhoods that survive a shift