Why a persistence-filter model demanded confirmation on IREN and banked under half the run

IREN was one of the period's most explosive movers, and the interesting question is what a rule that insists on confirmation does with a run like that: the demand for persistence kept it out of plenty of noise, and also out of a large part of the climb. This is a backtest of one fixed rule set replayed over IREN's saved history, not a forecast or a recommendation, and the live sample is still too small to grade.

Refusing to react to a flash is what defines this model. A setup must endure, outlasting a confirmation window before the rule commits, and that wait is exactly how it strains out the brief signals a quicker system would chase. Only a condition still in force when the window closes earns an entry, long or short as the setup directs, here twenty-nine longs against twenty-three shorts. The filter divided its 52 trades down the middle at a 50.00% win rate, with the winners traveling far enough to reach +308.44% on a flat stake.

How the model is built, end to end.

The reading the model lives by is whether a condition holds, not merely whether it appears. A signal that flares for a moment and fades never clears the confirmation window, so the model ignores it; only a setup that stays in place across that window is treated as real. The length of that wait is the whole filter, the deliberate cost the model pays to avoid acting on noise. It is a patience mechanism, and on a stock as jumpy as IREN that patience throws out a great many false starts, at the price of entering late.

Confirmation is the gate, the position is what follows. Once a setup has persisted long enough the model opens in its direction with a stop and target set in advance, then lets the trade run on those rails. Because entries come only after the wait, the model is by design a step behind the very start of a move. After entry the trade ends in one of a few places, and on IREN the stop did the most work, closing more positions than the target or the timed exits, a sign that even confirmed setups failed often on this volatile name.

The illustration above walks the model across IREN in a single line: the raw setup, the confirmation window it must survive, and the long or short that clears it. The point of the design is patience, paying a delay to avoid acting on every flicker. The model's lineage belongs in the picture too. It began as one candidate among many in an automated search and reached deployment only by clearing backtest and walk-forward checks, the road from idea to working rules behind every number here.

At a glance

These figures come from a backtest of the model on IREN, scored against fixed acceptance gates, not from a live track record.

Walk-forward verification

A trade walked through

The walked example is a long held about eight days. The setup held through its confirmation window, the model bought near 20 dollars, and the move ran to its target near 28 dollars for +41.21%, the largest single winner in the book. It is the filter at its best, a confirmed setup that paid for the wait it imposed.

What the full trade record shows

Across its 52 IREN trades the model won 26 and lost 26. The exits leaned on the stop: 23 were stopped out, 20 reached the profit target, and 9 closed on a time exit.

A stop-led split on an even win rate fits a filter that enters late: even confirmed setups fail often enough on a volatile stock to knock out the most trades, while the survivors are carried to the target or out on time.

The biggest winner was a long that ran to its target for +41.21% over about eight days in August 2025 after the setup held through its confirmation window. The patient winners were the payoff for the wait, large moves the filter stayed with once it had committed.

The worst trade was a long stopped out for -26.53% over roughly three days in August 2024; that single loss is also the equity curve's deepest drawdown and the reason the drawdown gate is the one that bites hardest.

No single trade carries the record, and with IREN up more than sevenfold over the window, the book's real lesson is how much of that run the confirmation filter forfeited by insisting a signal prove itself before it would act.

How does this compare to just holding IREN

Over the same window the model was tested on, simply buying IREN and holding would have returned far more. Lining the two up shows whether the rule earned its keep or merely came along on an explosive tape. The honest read is that it came along, banking under half the run, and the tiles below put numbers on the distance.

How well does the model reproduce its tape?

Walk-forward verification checks whether the saved rule path reproduces the expected signal behavior on held-out timestamps it was not built on. It is a consistency and replay-integrity test, not proof of live profit. A clean reproduction means the deployed rules behave like the studied ones; it says nothing about whether IREN will keep producing setups that persist the way they did across the test window.

In live trading the model has been quiet so far, with too few signals to set beside the backtest. Until more live trades accumulate, the backtest is the only evidence available, and it should be read as exactly that, a study of how the rules behaved on saved history.

When this approach fails

The model's failures come from the cost of its own caution. By waiting for confirmation it enters after a move is underway, so when a confirmed setup fails it is often already extended, which is how the deepest trade lost 26.53% on a long stopped out in August 2024. The same wait also means the model misses the first and often largest leg of a move, and on a stock that septupled that lateness is exactly why so much of the run went uncaptured.

Three things are worth watching if this ever trades at size. The first is drawdown: the worst stretch in the backtest was about 26.53%, the deepest here and enough to break the drawdown gate. The second is entry lateness, the structural cost of demanding confirmation, which leaves the early part of a move on the table. The third is the gap between live and backtested behavior, the first sign IREN's setups are no longer persisting the way they did in the study.

Risk and honest limits

On this run the model's automated checks logged a caution rather than a clean pass. It is one more reason to read everything here as backtest evidence about the rules, not a verdict on the stock.

Lifecycle