Does the Regime Transfer?

The MDI on European and UK Indices

The question, asked plainly

Every input to the Master Danger Index is American: Federal Reserve net liquidity, the US high-yield spread, the US yield curve, VIX, MOVE, the dollar index, US sector ratios and US sentiment surveys. Two of its inputs are the dollar’s crosses against the euro and the franc. There is no European or British series in it, and the strategy it drives holds US-listed instruments.

Every performance figure Oneiros Vision publishes is measured on the S&P 500.

A Swiss, European or British manager’s book is not the S&P 500. It is the SMI, the EURO STOXX 50, the FTSE 100, global equity, reported in francs, euros or sterling. The fair question from that desk is not whether the signal works, but whether it says anything about their market — and the honest answer until this study was that nobody had measured it.

What was required in advance

Four criteria were fixed in writing before the first run, so that no result could be obtained by moving a threshold afterwards:

• T1 — monotonic forward drawdown. Forward drawdown worsens across CALM, CHOP, WARNING and CRITICAL.
• T2 — tail capture. The defensive regimes hold at least 60% of the index’s worst 1% of sessions, on at most 40% of sessions.
• T3 — not an FX artefact. Restating a dollar-quoted index in another currency does not reverse T1.
• T4 — power. At least 20 independent stress episodes.

An index passes only if T1 and T2 hold, T3 does not contradict, and T4 is met. A failure is recorded and published rather than dropped.

Results

Window: 2010 to August 2026, each index on its own trading calendar and in its own local currency.

Gradient is the CRITICAL forward drawdown divided by the CALM one: how much deeper the worst regime actually draws down, rather than merely whether it does. It is descriptive and is not one of the four criteria — those were fixed before the first run, and a fifth added afterwards would be the selection the fixed criteria exist to prevent. It is reported because a monotonic ordering can still be worth little if the worst regime is barely worse than the calmest.

The FTSE result has two halves

The index was chosen in writing before the study ran: the FTSE 100, because it is what a UK mandate is benchmarked against. The All-Share was ruled out in advance as a fallback, so that a failure could not have been converted into a pass by trying the other index afterwards.

It passed every criterion, and its tail capture is the highest in the study. 81.4% of the FTSE’s worst 1% of sessions fell inside a defensive regime, on 31.7% of sessions — better than the S&P 500 control’s 76.2%. On this evidence the regime identifies British stress at least as well as it identifies American stress.

Its drawdown gradient is the shallowest in the study. The FTSE’s CRITICAL forward drawdown is 1.42 times its CALM one, against 1.99 for the S&P control. The index is commodity- and staples-heavy and earns internationally; it does not sell off as violently, so the distance between a calm regime and the worst one is smaller in absolute terms.

Both facts are measured and both belong in a UK conversation. Read together they say the overlay’s value to a British book is concentrated in avoiding the worst sessions rather than in a large reduction of typical drawdown.

Where it did not pass, and why that is published

MSCI World did not meet the evidence threshold. Its table is as monotonic as the others and its defensive capture is 70.3%, but the instrument used to measure it lists from 2012, leaving 17 independent stress episodes where the rule requires 20.

That is not demonstrated, which is a different statement from does not work. The honest reading is that this study does not have enough independent stress in that series to make the claim. The remedy is a longer series chosen for its history — not a lower threshold, and not the same test re-run later on the hope that the count has grown. Selecting a threshold after seeing a result is precisely the error the fixed criteria exist to prevent.

Reading the control

The S&P 500 control reads 76.2% where the published tear sheet reads 78.6% for the same measurement. The difference is the instrument: the tear sheet measures SPY, a total-return ETF on the ETF’s own calendar, while this study measures the S&P 500 price index on its own. A control that reproduced the published figure exactly would indicate the study was reading the tear sheet’s own series rather than testing the same question on a different one.

What this supports, and what it does not

It supports: the regime published by Oneiros Vision has ranked forward drawdown on the SMI, the EURO STOXX 50 and the FTSE 100 over this window, in the same way and — on the tail-capture measure — to at least the same degree as it does on the S&P 500.

It does not support: that Oneiros Vision is a European or British model. No European or British series is an input to the signal, the strategy holds US-listed instruments, and the chassis is frozen. This is evidence that an American signal also ordered risk abroad — a statement about transfer, and nothing wider.

It is not a recommendation. Oneiros Vision measures; the allocation decision, its suitability for any client, and its implementation remain the manager’s.

Currency is a separate question, and a large one. These figures are the indices’ own local-currency returns. What a franc-based client earns from a dollar-denominated implementation is a different number — see the currency section of the tear sheet, where the same strategy’s drawdown restated in francs is more than twice the dollar figure, almost all of it exchange rate.

Reproducibility

The study runs from frozen inputs: the same month-end publication snapshot behind every published figure, and a separate digested file of index prices. Both carry SHA-256 manifests, and the study refuses to run if either fails its digest. It performs no network access and reads no live database.

Nothing in it touches the signal. The module that measures forward drawdowns is barred by an architecture test from being imported by anything that determines a regime, because a measurement that looks forward must never be visible to the thing being measured.