Method

What 243 trades can and cannot prove

The main trend strategy was positive on average in its unseen test and is still inconclusive. Here is why the confidence interval matters more than the average.

21 Oct 2026 · 2 min read · 390 words · updated 22 Oct 2026

What the record says

Rules: Long only, on completed daily candles, one position per coin at a time.
Market filter: Bitcoin's daily close must be above its 100-day average before any entry is allowed.
Entry: A coin closes above the highest close of its prior 20 days.
Exit: The daily close falls below its prior 10-day low, or the market filter turns off.
Fill assumption: The next day's open after the 00:00 UTC daily close, never the price shown at the close.
Costs: 30 bps, or 0.30%, per round trip.
Sample: 243 trades from January 2025 to September 2026.
Win rate: 33.7% of the trades won.
Average: +1.5% net per trade after costs.
Profit factor: 1.32.
Winners and losers: The average winner was about +19% and the average loser about −7.3%, which is the shape trend following is supposed to have.
Uncertainty: The 95% confidence interval for the average return was −2.9% to +6.3%. It includes zero.
Independence: There were only 33 independent entry-weeks, so the trades are far from independent samples.
Development bias: Development leaned heavily on the 2020–2021 bull market, and Bitcoin was in an uptrend for most of the test window.
Portfolio view: As a portfolio the same trades made +12.8% with a −16.9% maximum drawdown. Holding the same coins made +40.5% with a −54.2% drawdown.
Concentration: Without ZEC the portfolio was about flat, so one coin carried the result.
Execution sensitivity: Repricing every trade at fills 1 and 4 hours late barely moved the result.
Conclusion: The honest label is inconclusive, not proven. Live paper tracking from 1 October 2026 is what adds evidence now.

Where the live record stands

This note covers a frozen or retired result, and no live engine reports on it.

How to check it

Every figure above is taken from the frozen test outputs and the live paper record.
The changelog dates every version and its result, and the strategy catalog lists the rules as they were frozen.
Positions described here are paper positions; no order is ever placed.

Evidence behind this note

Rules: Long only, on completed daily candles, one position per coin at a time.
Market filter: Bitcoin's daily close must be above its 100-day average before any entry is allowed.
Entry: A coin closes above the highest close of its prior 20 days.
Exit: The daily close falls below its prior 10-day low, or the market filter turns off.
Fill assumption: The next day's open after the 00:00 UTC daily close, never the price shown at the close.
Costs: 30 bps, or 0.30%, per round trip.
Sample: 243 trades from January 2025 to September 2026.
Win rate: 33.7% of the trades won.
Average: +1.5% net per trade after costs.
Profit factor: 1.32.
Winners and losers: The average winner was about +19% and the average loser about −7.3%, which is the shape trend following is supposed to have.
Uncertainty: The 95% confidence interval for the average return was −2.9% to +6.3%. It includes zero.
Independence: There were only 33 independent entry-weeks, so the trades are far from independent samples.
Development bias: Development leaned heavily on the 2020–2021 bull market, and Bitcoin was in an uptrend for most of the test window.
Portfolio view: As a portfolio the same trades made +12.8% with a −16.9% maximum drawdown. Holding the same coins made +40.5% with a −54.2% drawdown.
Concentration: Without ZEC the portfolio was about flat, so one coin carried the result.
Execution sensitivity: Repricing every trade at fills 1 and 4 hours late barely moved the result.
Conclusion: The honest label is inconclusive, not proven. Live paper tracking from 1 October 2026 is what adds evidence now.

Where it lives in the record

More notes

How this note was written

It is generated from a fixed fact sheet built from the frozen test outputs and the live paper record, then checked so that every figure in the text appears in that sheet. Fact sheet 5e280178592b746e.

Paper research only. Not investment advice, and not a signal. Read the commitments for what stays public and the signal log for every signal as it was published.

What the record says

Rules: Long only, on completed daily candles, one position per coin at a time.

Market filter: Bitcoin's daily close must be above its 100-day average before any entry is allowed.

Entry: A coin closes above the highest close of its prior 20 days.

Exit: The daily close falls below its prior 10-day low, or the market filter turns off.

Fill assumption: The next day's open after the 00:00 UTC daily close, never the price shown at the close.

Costs: 30 bps, or 0.30%, per round trip.

Sample: 243 trades from January 2025 to September 2026.

Win rate: 33.7% of the trades won.

Average: +1.5% net per trade after costs.

Profit factor: 1.32.

Winners and losers: The average winner was about +19% and the average loser about −7.3%, which is the shape trend following is supposed to have.

Uncertainty: The 95% confidence interval for the average return was −2.9% to +6.3%. It includes zero.

Independence: There were only 33 independent entry-weeks, so the trades are far from independent samples.

Development bias: Development leaned heavily on the 2020–2021 bull market, and Bitcoin was in an uptrend for most of the test window.

Portfolio view: As a portfolio the same trades made +12.8% with a −16.9% maximum drawdown. Holding the same coins made +40.5% with a −54.2% drawdown.

Concentration: Without ZEC the portfolio was about flat, so one coin carried the result.

Execution sensitivity: Repricing every trade at fills 1 and 4 hours late barely moved the result.

Conclusion: The honest label is inconclusive, not proven. Live paper tracking from 1 October 2026 is what adds evidence now.

Evidence behind this note

Rules

Long only, on completed daily candles, one position per coin at a time.

Market filter

Bitcoin's daily close must be above its 100-day average before any entry is allowed.

Entry

A coin closes above the highest close of its prior 20 days.

Exit

The daily close falls below its prior 10-day low, or the market filter turns off.

Fill assumption

The next day's open after the 00:00 UTC daily close, never the price shown at the close.

Costs

30 bps, or 0.30%, per round trip.

Sample

243 trades from January 2025 to September 2026.

Win rate

33.7% of the trades won.

Average

+1.5% net per trade after costs.

Profit factor

1.32.

Winners and losers

The average winner was about +19% and the average loser about −7.3%, which is the shape trend following is supposed to have.

Uncertainty

The 95% confidence interval for the average return was −2.9% to +6.3%. It includes zero.

Independence

There were only 33 independent entry-weeks, so the trades are far from independent samples.

Development bias

Development leaned heavily on the 2020–2021 bull market, and Bitcoin was in an uptrend for most of the test window.

Portfolio view

As a portfolio the same trades made +12.8% with a −16.9% maximum drawdown. Holding the same coins made +40.5% with a −54.2% drawdown.

Concentration

Without ZEC the portfolio was about flat, so one coin carried the result.

Execution sensitivity

Repricing every trade at fills 1 and 4 hours late barely moved the result.

Conclusion

The honest label is inconclusive, not proven. Live paper tracking from 1 October 2026 is what adds evidence now.

How this note was written

Paper research only. Not investment advice, and not a signal. Read the commitments for what stays public and the signal log for every signal as it was published.