Most retail strategies are one idea applied everywhere: a breakout system, a mean-reversion system, a trend system. One idea has one failure mode, and when the market moves into it, the whole account is in it. The desk's method is built the other way round: six sleeves, each a small rule set with its own regime filter, each given the same share of capital, each allowed to say "not now".
What a sleeve is
A sleeve is a self-contained rule set with four parts: the regime it is allowed to trade in, the setup it looks for, the invalidation that proves the idea wrong, and the sizing rule, which is the same for every sleeve: 1% of the account at risk per trade, with the position sized from the distance to the invalidation. A sleeve that is not in its regime does nothing. It does not reach for trades; it waits.
Why six, and why equal capital
Six is not a magic number. It is roughly the smallest set that covers the market's main states, trending up, trending down, ranging, expanding from a range, and the two transitional moments between them, without the sleeves overlapping so much that they are one idea wearing six hats. Equal capital is a discipline: it stops the desk from quietly overweighting whichever sleeve had the best recent run, which is how a diversified method turns back into a single bet.
Standing aside is a position
In most periods, some sleeves are active and some are standing aside. The route table on the method pages shows which, and why. A week in which four of six sleeves are idle is not a failed week; it is the method declining to trade conditions it was not built for. Over a full window this shows up as fewer trades and shallower drawdowns than a system that must always be in the market, and as a return that does not come from being clever in every regime but from not being wrong in most of them.
What the record shows
On the six-sleeve backtest, hypothetical and over the full window, the method compounds at 9.5% a year with a maximum drawdown of −6.3%. The worst calendar year, 2022, is −2.2%: a down year, on the same page as the up years, because a method that cannot show its down year has not shown its method. Those figures are labeled backtest and are kept separate from the paper and live records, which are shorter and their own.
What matters in those three numbers is their relationship. A return of that size with a drawdown of that size is what diversification across sleeves buys: not the highest return available, but a return that would have been possible to sit through.
How a signal comes out of it
When a sleeve is in its regime and its setup appears, the desk publishes a signal with the entry zone, the invalidation, the targets and the sizing rule, labeled, before the move. The signal names the sleeve it came from, so the record can be read sleeve by sleeve: which ones carry the result, which ones have been quiet, which ones were stopped. Every closed signal gets a written review, win or lose, and the review names what the rule did, not what the trader felt.
What the method does not do
It does not predict. It does not add size to a winning streak or cut size after a losing one. It does not override a sleeve because the week has been quiet. It does not blend the backtest into the live record to make the curve longer. Those are the ways a method stops being a method, and a reader who understands the six sleeves can see from the record whether the desk has kept to its own rules.
Backtested figures are hypothetical, do not represent actual trading, and have inherent limitations. Past performance of any kind does not predict future results. Nothing here is personalised investment advice.