Regime scaling

This is the most surprising behaviour in the product. Read this before concluding something is broken.

The short version

When our risk rules classify the market as unfavourable for the strategy, AlphaHouse substantially reduces the size of your whole book, to roughly half. When conditions improve, full size is restored.

You did not do anything. Your leverage setting did not change. The system deliberately took risk off.

Why

The strategy is market-neutral, but its edge is not constant across market conditions: it earns in calm and rising markets and gets squeezed in falling, volatile ones. The backtests are unambiguous about this.

Reducing size in the unfavourable regime roughly halved the worst drawdown across the full history without giving up return, and the improvement held across a wide range of settings rather than at one lucky value, which is what separates a real effect from a fitted one.

The honest framing: this is a risk control that happens to also improve returns, not a prediction. It does not forecast the market. It observes the regime you are currently in and sizes accordingly.

How the switch behaves

The classification comes from a mechanical, predefined rule computed fresh each day from public market data. Two properties matter to you:

• It is deliberately slow to flip. The rule is built so that a market hovering near the boundary does not toggle your book between full and reduced size day after day. State changes happen on decisive moves.
• It fails toward full size. If the data needed to classify the regime is missing or invalid, the system runs at full exposure rather than silently shrinking your book. A data problem must not be mistaken for a market call, though it does mean an outage removes the protection rather than triggering it.

We do not publish the rule itself. It is part of what you are subscribing to, and the performance record is the record of the strategy with this behaviour in it.

What you will actually see

When the regime turns defensive, on the next run:

• Many orders, mostly reductions, on Trade history.
• Gross exposure and gross leverage on Portfolio dropping to roughly half.
• Your leverage setting on Systems unchanged.

That last point causes most of the confusion: your setting says one thing, your realised leverage says about half of it, and both are correct. The setting is an input; regime scaling is applied after it.

Can I turn it off?

No. It is part of the strategy’s definition, not a preference. An account running without it would not be running the strategy the published numbers describe.

Things this is not

• Not a stop-loss. It does not close losing positions; it resizes the whole book regardless of whether you are up or down.
• Not personalised. Every account on the system sees the same state on the same day.
• Not the only thing that affects your book size. Going over your plan’s leveraged capital cap has its own consequence — a plan change, with notice — but it does not scale your exposure and does not change your leverage setting. Regime scaling is the one that moves gross exposure. See Plan limits.