Skip to content
GetProfitable
Search

Versioning and the quarterly review

Lesson 19 · about 9 min

Every version of the plan is a document with a number and a date range, kept forever. The quarterly review is the one moment when the big questions are allowed: is the setup still a setup, is the market still the market, and is the edge still there. This lesson covers the bookkeeping and the red flags.

Version numbers

The plan carries a version and a valid-from date at the top. The rule for bumping it:

  • Minor (v1.1 to v1.2): one exit, filter or breaker rule changed, per the test protocol.
  • Major (v1.x to v2.0): a setup criterion, market, session or risk number changed. Statistics restart; v2.0 trades are not pooled with v1.x trades for setup evaluation.

A changes file lives next to the plan. One line per version:

v1.0  1 Jul   Initial. MES ON-range breakout, 30-min time stop.
v1.1  1 Aug   Filter: ON range 0.3% -> 0.4%. Shadow-tested 34 trades, +0.5R.
v1.2  15 Sep  Time stop 30 -> 60 min. Tested 50 trades, +0.42R. KEEP.
v1.3  1 Oct   Trail: swing-low -> 10 EMA close. Tested 30, -0.1R. REVERTED 1 Nov.
v1.2  1 Nov   Back to v1.2 rules.

The file is what stops you retesting v1.3 in March. It is also what you read at the quarterly review to see whether the plan has been drifting in one direction, which is itself information.

Every journal row carries the version, so any statistic can be cut by version. Cumulative numbers are always "since the current major version".

The quarterly review

Two hours, once a quarter, market closed, with the last three monthly sheets, the changes file, and the journal open. It asks six questions, in order.

1. Was the plan followed? Plan-follow score at 3 on at least 85% of trades, across the quarter. If not, the quarter's results do not test the plan and the rest of the review is about execution. Stop here, write execution actions, and come back next quarter.

2. Is the edge inside its expected range? Cumulative average R on the current major version against the expected numbers from the journal's front page. Inside: fine. Below: go to question 5. Above: be suspicious of the sample size or the cost accounting before being pleased.

3. Are the A-grade trades carrying it? Average R, A-grade versus B and C, over the quarter. If B and C trades are still being taken at more than 15% of the total, the execution action is to stop taking them, and the setup evaluation uses A-grade trades only.

4. Is the filter earning its keep? Count of filter-fail days versus trading days; average R on days that barely passed versus days that passed easily. A filter that excludes 70% of days for a 0.1R difference is costing sample for nothing.

5. Is the edge gone? See the red flags below.

6. What is the one major change for next quarter, if any? At most one. Written as a test protocol, with its version number, starting on a date.

Key idea: The quarterly review asks whether the plan was followed before it asks whether the plan works. A plan followed 70% of the time has not been tested, and its results say nothing about the setup.

Red flags that the edge is gone

None of these is proof on its own. Two together are a reason to halve size and test. Three are a reason to stop trading the setup live and shadow it.

Flag What it usually means
Cumulative average R on A-grade trades negative over 60+ trades The setup is not paying, at plan-follow 85%+
Average winner shrinking quarter over quarter while losers stay at −1R Follow-through is fading; the move the setup catches is smaller
Win rate stable but average R falling Same; winners getting smaller, often a volatility regime change
Filter-pass days falling sharply The market's character changed and the setup's conditions are rare
Drawdown exceeds twice the worst drawdown of the previous two quarters, at plan-follow 85%+ Outside the plan's own history; treat as regime change
Costs per trade rising as a share of average R Spreads or slippage widened; a thin edge has been eaten

What is not a red flag: a losing month. A six-loss streak. A quarter at +0.05R after two at +0.25R, on 40 trades. All of those are inside the noise of a 0.2R plan, and Module 2's table says so.

When the edge is gone

The response is not to abandon everything. It is:

  1. Halve 1R, per the risk plan's drawdown rules.
  2. Keep trading the setup at half size for one more quarter, logging as before, so the data continues.
  3. Shadow-test the strongest candidate from the list, if there is one.
  4. If the following quarter confirms, retire the setup to the changes file with the date and the numbers, and start a v2.0 with a different setup on the same market, or the same setup on a different market, never both at once.

A retired setup is not deleted. Markets cycle, and a setup that stopped working in a low-volatility year may be the first one to revisit when volatility returns. The changes file and the journal are the record that lets you do that with evidence instead of nostalgia.

Try it: Create your changes file with a single line for v1.0 and today's date. Then put the quarterly review on the calendar, two hours, on the first weekend after the quarter ends, and write the six questions at the top of a blank page in the back of the journal so they are waiting.

Recap

  • Minor versions change one exit, filter or breaker rule. Major versions change setup, market, session or risk, and restart the statistics.
  • A one-line-per-version changes file records every keep and every revert, so nothing is retested by accident.
  • The quarterly review asks six questions in order, starting with "was the plan followed?", and allows at most one major change.
  • Red flags: negative A-grade R over 60+ trades, shrinking winners, falling filter-pass days, drawdown outside history, rising costs. Two together mean halve size; three mean shadow.
  • A retired setup goes to the changes file with its numbers, not to the bin. Markets cycle.

See it drawn

Original diagrams for the ideas on this page. Illustrative, not real market data.

The spread of outcomes behind an expectancyA histogram of forty trades: a tall block of small losses on the left, a low spread of larger wins on the right, and a line marking the average outcome.NUMBER OF TRADES051024 LOSSES, AVG −$20016 WINS, AVG +$600EXPECTANCY +$120−$400−$200$0+$200+$400+$600+$800PROFIT OR LOSS PER TRADEexpectancy = (40% × $600) − (60% × $200) = +$120 per trade
Expectancy: the average trade. Forty trades sorted by outcome: 24 small losses and 16 larger wins. Weighting each side by how often it happens gives the average result per trade, marked here by the dashed line at +$120.
An equity curve and its drawdownAn account balance rising over a year, falling from a peak to a trough, then climbing back to the old peak.ACCOUNT EQUITY$20k$12k$8k024681012TIME (MONTHS)PEAK $16,000TROUGH $12,000DRAWDOWN−25%RECOVERY
Equity curve and drawdown. An account balance plotted month by month. The fall from the $16,000 peak to the $12,000 trough is a 25% drawdown, and the shaded area lasts until the balance climbs back to the old peak.
A range beside a trendOne chart swinging between a flat floor and ceiling, another stepping upwards inside a pair of sloping lines.Range-boundresistancesupportprice bounces between two levelsTrendingthe trend channelhigher highs and higher lowsA range has two flat edges; a trend has two sloping ones.
Range versus trend. On the left price keeps bouncing between the same floor and ceiling, which is a range. On the right each high and each low is higher than the last, inside a pair of sloping lines called a channel.