When the method stops fitting

A method stops fitting when the life it was built for has changed, or when you have evidence, gathered the slow way, that the text no longer does what it was for. This chapter separates those two retirements from the ordinary urge to quit during a losing streak.

A losing streak is not a verdict

If the test said six losing months happen, and you are in month four, you are inside the method, not past it. Quitting here is the behaviour the previous chapter warned about: taking the decline and leaving before the recovery that justified the decline. The review during a streak is an obedience review. Are you still following the text? If you are, the streak is the cost. If you are not, restore obedience before you judge the text.

Evidence that the text itself is tired

Collect it in a window you named before you were unhappy. For example: ‘I will review the method only after thirty live trades, or after twelve months, whichever is later, and I will retire it only if the live worst-decline is far beyond the tested one and the overrides are not the cause.’ That sentence prevents a bad fortnight from becoming a redesign. When the date arrives, compare like with like. Live trades against the frozen text, costs included, overrides separated. If the live sample is simply small, the honest answer is ‘not enough to know’, which is an uncomfortable answer and often the correct one.

A life change is a different door

The method can be fine and still no longer be yours. You will need the money inside a year, because of a house or a parent. You no longer have a weekly hour to review it, and the method requires a weekly close. Your surplus has shrunk, so the size rule produces quantities too small to bother, or so large in your new life that the old rupee limit is reckless. These are not failures of the 30-week average. They are changes in the constraints the method was written under. Rewrite the constraints first. Sometimes the new constraints leave no room for the method, and the correct replacement is a broad fund and a calendar, not a more complicated screen.

How to retire it without a dramatic last trade

1. Stop new entries on the day you decide. Do not take ‘one more, because it is setting up’.
2. Manage the positions that are already open under the old exits, or close them at the next scheduled review. Write which of those two you are doing, and do not blend them trade by trade.
3. Keep the log. The next method is not entitled to pretend the old one never happened.
4. Wait a month before the replacement method is allowed to take a live trade. A replacement written on the evening you quit is usually the old method with the painful rule deleted.

What to remember

• A streak that the test anticipated is not a reason to quit.
• Judge the text on a window you scheduled before the disappointment.
• Separate ‘I cannot follow it’ from ‘it no longer fits my life’ from ‘the results left the tested range’.
• If the need for the money has acquired a date, the method is the wrong home for that money regardless of its record.
• Retire entries first, and do not write the successor on the same evening.