Skip to main content
Track Walk-forward analysis
33 / 50
Library

2011issue C1151-55

Evaluate generated systems on holdouts, then add stops

Automated rule search invents a mechanical trading system. It does not prove that system. Freeze the same entry-exit package, score it on bars withheld from the hunt, and only then attach a stop-loss the search was not allowed to optimize away.

  • A mechanical procedure can be searched from a catalog of 23 indicators and 9 order types rather than from an unconstrained rule language.
  • Partition the series into a construction sample and a reserved sample before the hunt, then score the frozen package on both with the same metric list.
  • After the historical search window ended, later unseen bars were used as a forward check.
  • A fixed-dollar stop of 500 was attached only in that later phase.
Entries in this reading3 entries

Search invents a procedure

A mechanical trading system is a complete, repeatable set of entry, exit, and stay-out rules run as one procedure on historical bars. A mechanical procedure can be searched from a catalog of 23 indicators and 9 order types rather than from an unconstrained rule language.

Search objectives can be reduced to a small weighted set that raises profitability, trade count, and equity-path correlation while lowering rule complexity. Those build-metrics steer the hunt. They do not replace a later score on bars the hunt never used.

Partition the series first

Before the search starts, the price series can be partitioned into a construction sample and a reserved sample, and the reserved block can sit at either end of the series. The construction sample is the in-sample period used to search for or fit the rules. The reserved sample is the out-of-sample period. It is not used to invent the rules and is scored after the search is frozen.

Candidate procedures can be scored on both the construction sample and the reserved sample with the same metric list, including trade count, drawdown, correlation, and significance.

Judge the frozen package on withheld bars

Walk-forward analysis is a validation pattern that judges a frozen procedure on bars withheld from construction, then on later unseen bars. In the historical workflow, candidates were scored on the reserved sample after the search was frozen.

One demonstration used five-minute bars, a fixed session window, symmetric long and short logic, and a mandatory flatten at the session close. The demonstration entries compared two directional-movement readings of different lengths, and the timed exit fired after a fixed number of bars in the position.

After the historical search window ended, later unseen bars were used as a forward check, and a fixed-dollar stop of 500 was attached only in that later phase.

Closed-trade equity of a generated e-mini S&P day-trading system

Over 511 closed day trades the account rises from the $15,000 starting stake to about $38,500, staying close to a straight uptrend after an early several-thousand-dollar dip and later pullbacks. The path is read off the Adaptrade Builder plot for 24 May 2010–23 May 2011, which also printed net profit $23,505 and profit factor 1.49.
Over 511 closed day trades the account rises from the $15,000 starting stake to about $38,500, staying close to a straight uptrend after an early several-thousand-dollar dip and later pullbacks. The path is read off the Adaptrade Builder plot for 24 May 2010–23 May 2011, which also printed net profit $23,505 and profit factor 1.49.E-mini S&P 500 futures · 5-minute bars · 2010-05-24T00:00:00.000Z to 2011-05-23T00:00:00.000Z

The plotted backtest used no fixed dollar stop; a $500 stop was attached only on the later ten-trade holdout. The run used one e-mini contract, $25 round-trip costs, and the 07:05–13:45 CT session.

Add a stop after the hunt

A stop-loss is a precommitted exit that bounds loss or exposure before or during a position. In the demonstration, that bound was a fixed-dollar stop of 500, attached only after the historical search window ended.

A separate sample procedure expressed a trailing exit as a 3 percent retreat from the highest close over a 15-bar lookback. A trailing-stop follows a recent extreme and exits when price retreats by a set fraction or distance.

Educational research material, not investment advice. Historical source context does not establish present-day performance.
33 of 50 in the Walk-forward analysis track
201212-17 pp.Next on Walk-forward analysisWalk-forward analysis and out-of-sample tests for a mechanical trading systemA mechanical trading system was treated as unfinished until it was forced to trade periods withheld from development, first at the start of the sample and then at the end.
All readings on this track · 50 readings
  1. 1990Three-window walk-forward system evaluation
  2. 1990Building the construction layer of a mechanical trading system
  3. 1991Constructing walk-forward neural trading rules
  4. 1991Constructing neural trading systems from facts to walk-forward
  5. 1992Walk-forward evaluation of stop overlays on average crossovers
  6. 1992Audit mechanical system tests for fills and regimes
  7. 1993Walk-forward evaluation of monthly yield and real-rate forecasts
  8. 1993Constructing walk-forward forecasts with linear and moving-average baselines
  9. 1993Walk-forward hybrid rules for intermarket forecast stacks
  10. 1994Neural-net construction as a mechanical trading-system problem
  11. 1995Constructing an intermarket neural net trading system
  12. 1996Weekly market breadth as one procedure on an unused window
  13. 1996Walk-forward evaluation of gold-index bond-fund rules
  14. 1996Evaluating weekday-in-month filters for index day trades
  15. 1996Require both a trend filter and a cycle oscillator before entry
  16. 1997Walk-forward windows as a diagnostic of parameter instability
  17. 1997Walk-forward validation of a market-breadth timing rule
  18. 1997Sunspot spikes and walk-forward evaluation of an adaptive cycle rule
  19. 1997A walk-forward check for bond-breadth timing
  20. 1998Walk-forward audit of regression trend forecasts
  21. 1998Evaluating a cubic least-squares currency trend with walk-forward segments
  22. 1998Walk-forward evaluation of recursive yen trend signals
  23. 1999Personal system design under crowd psychology
  24. 1999Walk-forward evaluation of a polynomial price forecast
  25. 2000Walk-forward optimization of regression-slope-angle rules
  26. 2001Construct a winter seasonal window as one procedure
  27. 2001Inspectable rules when system write-ups dry up
  28. 2002Evaluating mechanical systems before position sizing
  29. 2003Walk-forward construction of rule-based market-position systems
  30. 2007Evaluating metal seasonal windows across regimes
  31. 2007Evaluating mechanical timing systems against hold baselines
  32. 2011Walk-forward reoptimization as a system design gate
  33. 2011Evaluate generated systems on holdouts, then add stops
  34. 2012Walk-forward analysis and out-of-sample tests for a mechanical trading system
  35. 2012Personality-first trading system design
  36. 2012Scorecard-first mechanical system construction
  37. 2012Constructing an advancer-decliner moving average for market breadth
  38. 2012Formula search as mechanical system construction
  39. 2013Identity-first system construction
  40. 2013Construct a swing system from bias rules to walk-forward
  41. 2014Evaluate mechanical stock systems with stops and walk-forward
  42. 2014Walk-forward velocity filters on noisy intraday trends
  43. 2015Event-predictability versus position-constrained rules
  44. 2015Constructing mechanical systems for walk-forward tests
  45. 2016When a tested system must be retired
  46. 2016Walk-forward metric filters and chance-level checks for selected inputs
  47. 2018Evaluate mechanical trading systems without catalog rankings
  48. 2019Phased stop construction from entry risk to trailing exit
  49. 2020Stockpiling simple ideas for mechanical system construction
  50. 2020A pretty first draft is not a walk-forward waiver
All 95 readings tagged Walk-forward analysis
Also on Walk-forward analysis5 readings