Skip to main content
Track Risk-reward ratio
3 / 27
Search the archiveLibrary
Library

1991issue C061-5

Ranking month-end holds with risk-reward and stops

An end-of-month seasonal-trading rule was specified as a complete buy-and-sell procedure. Four holding-period variants were ranked on the same two risk-reward-ratio measures, and a stop-loss was proposed only after large drawdowns appeared on the earlier entries.

  • Seasonal-trading at month-end is a complete entry, exit, and stay-out procedure confined to an end-of-month-window, not a story about why prices rise.
  • Four holding-period variants were ranked with average-winner-to-average-loser and average-winner-to-average-winning-drawdown before any stop-loss was added.
  • A stop-loss overlay was proposed only after large maximum drawdowns appeared on windows that entered three days before month-end.
  • The two-day-before to first-day-after hold was run as a regime-check in an up year and a down year, while the sample was still judged too short for a stable risk-reward reading.
Entries in this reading3 entries
Current trackRisk-reward ratio78 readingsSeasonal trading130 readingsStop-loss order297 readings

A complete calendar procedure

The archive specified seasonal-trading as a complete buy-and-sell rule confined to the last sessions of one month and the first sessions of the next. That band is the end-of-month-window: the only span in which the seasonal rule is allowed to act.

Outside that window the procedure stays out. Timing is driven by calendar location rather than by a security-specific forecast. Entry, exit, and abstention belong to one rule.

Four holding-period variants

Four holding-period variants were formed by pairing an entry two or three days before month-end with an exit on the first or second day of the following month. Each holding-period is the exact buy-day and sell-day pair that turns the calendar hypothesis into a single system that can be compared with the others.

The end-of-month-window stayed fixed. Only hold length changed.

Ranking on two risk-reward ratios

Those four variants were compared on a sample running from June 1988 through February 1991. Holding-period ranking used two risk-reward-ratio measures: average-winner-to-average-loser, and average-winner-to-average-winning-drawdown.

Average-winner-to-average-loser compares typical winning outcomes with typical losing outcomes. Average-winner-to-average-winning-drawdown pairs typical gains with typical peak-to-trough setbacks inside winning trades. The archive used those ratios to rank the holds before a later overlay was discussed.

A stop-loss after measured drawdowns

A stop-loss overlay was proposed after large maximum drawdowns appeared on the windows that entered three days before month-end. The stop-loss is a precommitted exit meant to bound loss or exposure after a seasonal hold is open, rather than leaving the calendar window unhedged.

Editorial reading: the stop-loss enters the procedure after drawdown size has been measured, not as the first design choice.

Regime-check and sample length

The two-day-before to first-day-after window was the variant carried into a year-by-year check of 1989 and 1990, described as an up year and a down year. That regime-check re-runs the same hold to test whether the calendar rule is only an artifact of one market state.

The sample was treated as statistically notable yet still too short for a stable reading of risk-reward characteristics, and a longer test was requested.

A hypothesized cause

Month-end clustering of purchases by corporate equity holders was offered as a hypothesized cause of the calendar pattern.

Editorial reading: the ranking, the stop-loss proposal, and the regime-check can be judged without accepting that cause.

Educational research material, not investment advice. Historical source context does not establish present-day performance.
3 of 27 in the Risk-reward ratio track
19941-8 pp.Next on Risk-reward ratioEvaluating percentage profit exits on trend-following averagesKeep the trend-following control as a hold-until-the-average-reverses path, then overlay profit-taking as a second, testable close.
All readings on this track · 27 readings
  1. 1987Precommit the stop and require a surplus payoff
  2. 1987When open losses overwrite precommitted exit rules
  3. 1991Ranking month-end holds with risk-reward and stops
  4. 1994Evaluating percentage profit exits on trend-following averages
  5. 2001Construct reward to risk before entry
  6. 2003A pre-trade readiness checklist for entry, abstention, and risk
  7. 2003Evaluating out-of-the-money strikes with saddle premiums and implied volatility
  8. 2004A critique of one-procedure trading decisions
  9. 2004Mechanical stops leak intent and misprice losses
  10. 2004Implied-volatility premium on rate indexes
  11. 2006Keep the stop that authorized the entry
  12. 2006Bound the loss with a risk-to-reward filter before entry
  13. 2007Audit the operator before abandoning a system
  14. 2007Failed-break slingshot reversal rules
  15. 2007A system-management template that sets the stop and the required reward before entry
  16. 2007Construct a pre-trade style envelope for entries and skips
  17. 2008Systems that skip the three decision gates
  18. 2010Four-gate veto before trend entries
  19. 2012Construct a trading plan as one written procedure
  20. 2012Forex stops and reward targets by quote increment
  21. 2013The average directional index as a lookback cluster and a risk-reward bound
  22. 2014A three-step worksheet for VIX futures tick value and low-vol risk-reward
  23. 2014Binary option premia, fees, and pre-expiry risk-reward
  24. 2015Selling far out-of-the-money commodity option premium as a forgiving timing case
  25. 2016Prepaid option spreads that bound range-day risk and reward
  26. 2016Pre-trade risk-reward versus instinctive exits
  27. 2020Bounding a short-dated option spread with a stop, a gap remainder, and a credit exit
By fieldSeasonality and regimes180 readingsBy year1991166 readings
All 78 readings tagged Risk-reward ratio
Also on Risk-reward ratio5 readings
1987Precommit the stop and require a surplus payoff1987When open losses overwrite precommitted exit rules2000Budget the loss before entry with stops and position limits2004Mechanical stops leak intent and misprice losses2004A critique of one-procedure trading decisions