DAY 32: Indicators to Look at in Verification – Winning Percentage Alone Isn’t Enough

On DAY 31 we learned about the importance of forward testing and concrete ways to proceed.
Today, DAY 32, focuses on indicators to check when evaluating verification results. One common mistake beginners make is,focusing too much on win rate.
However, a high win rate alone does not guarantee continued success. There are other important evaluation items, solet's learn how to judge the quality of the logic and methods comprehensively.

1. Why is "win rate alone" dangerous?

1. A poor risk-reward ratio makes it easy to lose

   • Example: even with a 90% win rate, if one loss is -100,000 yen and a win is +10,000 yen, you are likely to end up in the red overall.
   • Statistically, even with a high win rate, if the logic permits large losses, the risk of ruin is high.

2. Cannot grasp drawdown (funds decline during a losing streak)

   • High win rate does not necessarily mean few losses in a row.
   • Even with 85% win rate, occasional streaks can deplete a large portion of funds (martingale-type EAs are typical).

3. Subject to market conditions

   • If a high-win-rate logic is only effective in trending markets, it may not work at all in range-bound markets.
   • Win rate alone makes it hard to see when and in what environment that win rate occurs.

2. Indicators to pay attention to besides win rate

(1) Drawdown

• Definition：

  • A value that shows how much the account equity has fallen from its peak.
  • Example: Maximum drawdown 30% → funds temporarily fell by 30%.

• Importance：

  • Large drawdown often makes it mentally intolerable.
  • If funds drop to half or less, recovering expects a large return.
  • Even if ultimately positive, a period of sharp declines increases bankruptcy risk.

(2) PF (Profit Factor)

• Definition：
  • Calculated as "Total Profit ÷ Total Loss".
  • Example: total profit 1,000,000 yen and total loss 500,000 yen → PF is 2.0.
• Guidelines：
  • PF > 1.0 indicates at least a positive result.
  • PF > 1.3–1.5 gives a relatively stable impression.
  • PF ≥ 2.0 suggests a highly excellent logic, but also beware of sample size.

(3) Number of consecutive losses & Win rate × Risk-Reward Ratio

• Consecutive losses：

  • Example: if the maximum losing streak is 8, be prepared for that level of drawdown.
  • In discretionary trading, consider whether you can endure 8 consecutive losses and adjust position sizing accordingly.

• Win rate × Risk-Reward：

  • Expected value = win rate × average win − (1 − win rate) × average loss
  • Even with a 50% win rate, if average win is twice the average loss, it is positive in the long run.
  • Conversely, with an 80% win rate, if losses when you lose are large, it becomes negative.

(4) Average Return Rate (Average Profit/Loss Rate)

• Definition：
  • Average profit in pips or amount per trade.
  • Shows tendencies like whether there are large stop-outs or if profits are not growing because of gradual take-profits.

**3. An example where "low win rate can still win": Trend-following type】

1. Example：

   • Win rate 30%, average +100 pips per win, -30 pips per loss.
   • Expected value: 0.3 × 100 − 0.7 × 30 = 30 − 21 = +9 pips
   • Overall positive. If you can ride a strongly trending market, it can cover the losing streaks.

2. Merits：

   • Even during drawdown periods, the strength is that “when a big trend is captured, you can turn positive quickly.”

3. Demerits：

   • During longer losing streaks, mental strain is high.
   • In range-bound markets, consecutive wins are hard.

**4. An example where "high win rate but you lose": Martingale type】

1. Example：
   • Win rate 90%, gradually accumulate small profits, butone loss can bring a large loss.
   • Many cases lead to debt or substantial margin reduction.
2. Characteristics：
   • In normal times it looks attractive because wins continue, buthistorical testing cannot handle “black swan” level markets risk.
3. Mental load：
   • You cannot predict when a huge loss will occur, so it tends to be unstable.

5. How should we evaluate? A comprehensive checklist

1. Win rate：as a rough guideline for reference.However, do not judge it alone.
2. Risk-reward ratio：1:1 or higher is the minimum. Higher like 1:2 tends to raise expected value.
3. Drawdown：whether you can tolerate the decline.
4. PF (Profit Factor)：PF > 1 is assumed; 1.3–1.5 or higher tends to be stable.
5. Consecutive losses：Can your psychology tolerate it? How long can losing streaks last?
6. Average gain/loss rate：a measure of how much gains are eroded by a single loss.
7. Performance in different market environments（during trends, during ranges, etc.）: identify weak markets and consider ON/OFF or position sizing adjustments.

6. Summary & next preview

Summary

1. “High win rate does not guarantee winning”; it is important to verify multiple indicators such as risk-reward and drawdown comprehensively..
2. Combining PF, consecutive losses, and average gain/loss rate helps clarify the strengths and weaknesses of the logic or method.
3. Evaluating trading methods: even with a 70% win rate, large drawdown can be dangerous, and a 30% win rate can be stable if risk-reward is high.
4. In real operation, always monitor these evaluation axes and adjust or modify position sizing according to market environment.

Next time (DAY 33) theme: How to modify rules based on verification results

• Tomorrow, we will cover the注意点 for modifying rules based on multiple verification indicators.
• If you try to reduce drawdown by adding too many indicators, you risk falling into the “optimization trap.”
• We will skillfully avoid that and gradually introduce techniques to improve “win rate & risk-reward,” so please look forward to it!

If you’re interested in automated trading, please also check the link below.
https://www.gogojungle.co.jp/users/147322/products

If this was helpful, we would be grateful if you press “Read more.”
Thank you.