Skip to main content
Track System optimization
11 / 23
Search the archiveLibrary
Library

1996issue C071-9

Evaluating a daily volume-strength and new-high new-low timing procedure

Daily NYSE advance-decline volume and daily 52-week new highs and new lows can be written as one mechanical timing procedure that updates inside the week. The same fitted Vrs-NH-NL rules looked stronger in the curve-fitted section than in the earlier window, and out-of-sample drawdowns were almost twice as large.

  • A weekly advance-decline and new-high/new-low market system can leave a multi-day gap between signal updates, during which adverse price movement can occur.
  • Daily NYSE advance-decline volume and daily NYSE 52-week new highs and new lows can be combined into one market-timing procedure that updates inside the week.
  • Volume strength (Vstr), new-high strength (NHstr), and new-low strength (NLstr) are built as separate percent-of-range readings, and the rolling 52-week new-high and new-low series should not be treated as continuous with the earlier calendar-year method.
  • The same fitted Vrs-NH-NL rules produced 23 trades with 70 percent profitable in the earlier window versus 21 trades with 95 percent profitable in the curve-fitted section, while out-of-sample drawdowns were almost twice as large.
Entries in this reading3 entries
Current trackSystem optimization131 readingsMarket breadth147 readingsMechanical trading system322 readings

The weekly update gap

A weekly advance-decline and new-high/new-low market system can leave a multi-day gap between signal updates. Adverse price movement can occur before the next weekly calculation.

Daily NYSE advance-decline volume and daily NYSE 52-week new highs and new lows can be combined into one market-timing procedure that updates inside the week.

A break in the new-high and new-low series

From January 1978 onward, NYSE new highs and new lows are computed on a rolling 52-week basis. That construction differs from the earlier calendar-year method and should not be treated as one continuous series.

Because only the 52-week new-high and new-low series begin in 1978, system development and testing that use those series are limited to advance-decline volume from 1 January 1978 forward.

Three separate strength readings

Volume strength (Vstr) is built from the advance-decline volume ratio (Vratio): advancing volume minus declining volume, divided by their sum. Vratio is smoothed with a 10-day exponential average. Today's reading is then scaled as a percent of its lookback range into an oscillator that runs from +50 to -50.

New-high strength (NHstr) and new-low strength (NLstr) are specified independently. The two series are treated as distinct investor classes with different horizons rather than as a single combined breadth count.

The fitted Vrs-NH-NL rules and the earlier window

On daily S&P 500 data from 2 January 1987 through 29 December 1995, the fitted Vrs-NH-NL rules used Vbuy=5, Vsell=-35, Vdays=10, NLbuy=-20, NLsell=30, Ldays=40, NHbuy=-10, NHsell=20, and Hdays=10. That later sample is the curve-fitted section, the window used to choose system thresholds and lookbacks, as distinct from the earlier holdout window.

The same parameter set applied to the earlier daily S&P 500 window from 3 January 1978 through 13 March 1987 produced 23 trades with 70 percent profitable, versus 21 trades with 95 percent profitable in the later fitted window.

Daily inputs versus weekly inputs

Using daily rather than weekly inputs gave little advantage beyond the ability to act during the week. The two systems' performances were described as almost the same, with only a slight edge to the daily version.

Holdout drawdowns as the comparison

Out-of-sample drawdowns were almost twice as large as those in the curve-fitted section. The earlier window included a 23 percent drawdown from the 12 June 1981 buy during the 1981-82 decline. Future drawdowns should be expected to exceed the fitted-window figures.

Editorial (TradersWeek): judge this mechanical procedure by whether out-of-sample drawdowns stay comparable to the curve-fitted section. The higher profitable-trade share in the window used to choose the thresholds is not the standard for accepting the rules.

S&P 500 daily path, April 1980–June 1982

Daily S&P 500 closes taken from Meyers' printed Figure 6B in the 1978–87 holdout window. After the 12 June 1981 buy at 133.49 the index fell into the low 110s, the 23 percent open drawdown he singled out, even though that long later booked an 8.97 percent gain.
Daily S&P 500 closes taken from Meyers' printed Figure 6B in the 1978–87 holdout window. After the 12 June 1981 buy at 133.49 the index fell into the low 110s, the 23 percent open drawdown he singled out, even though that long later booked an 8.97 percent gain.S&P 500 · Daily · 1980-04-01T00:00:00.000Z to 1982-06-30T00:00:00.000Z

Buy and sell prints use the exact Figure 4 entry prices; other closes are read from the printed daily curve to the nearest index point. The Vrs, NLstr and NHstr panes were left unread.

Educational research material, not investment advice. Historical source context does not establish present-day performance.
11 of 23 in the System optimization track
19971-9 pp.Next on System optimizationRule-based intraday system design with genetic searchThe mechanical trading system enters the S&P 500 on the next open and exits at that same session close only when two thresholded price-comparison rules are both true.
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 fieldMarket breadth135 readingsBy year1996121 readings
All 131 readings tagged System optimization
Also on System optimization5 readings
1984Evaluating mechanical systems beyond net profit1987Mechanical systems need more than pretty charts1988Two-sided optimization for volatility-scaled breakout rules1988Undisclosed rules block system robustness tests1989Constructing holdout-checked synthetic prices from chaos maps