Methodology — how Witan Way reads market regimes

Witan Way reads markets with rules and explains what it sees. This page is the canonical account of how those readings are produced — the philosophy behind them, what the rules actually read, how the models are built, what they cannot do, how often they speak, who governs the practice, and how the record is kept. It describes logic, not advice.

Philosophy

A systematic strategy does not know the future. It defines how you intend to behave when the future becomes uncomfortable. No strategy removes risk, and no strategy works in every environment. The point is not perfection — it is consistency under pressure.

Markets should be read before they are judged. A signal is not automatically an instruction. A rule chosen in calm is worth protecting under stress, and constraints are a form of honesty. Witan Way publishes readings, not recommendations: we describe what the rules see, and you keep the decision.

What the rules read

Every reading is computed from prices and market data for public reference indices — never a named fund, share class, or product. Across four pillars — equities, gold, commodities, and bitcoin — the rules read a small set of public dimensions:

Trend — is the pillar above or below its own longer-run path?
Momentum — has recent strength been building or fading?
Coherence — do trend and momentum agree, or is the picture mixed?
Volatility and breadth — how turbulent is the environment, and how broadly are constituents participating?
Drawdown from the recent high — how far a pillar sits below its own peak.

These dimensions combine into a per-pillar market regime — a descriptive state such as constructive, weakening, or transitional. It is a reading of the present, not a forecast. For the academic strategies, the published thresholds that turn these dimensions into a state are disclosed; for the proprietary Witan Way Core, the logic is explained here while its exact parameters stay private.

How the models are built

Witan Way exposes several strategies. The academic strategies — among them All-Weather, Dual Momentum, risk-parity and trend-following designs — follow a published academic lineage: their logic and windows are disclosed. The proprietary Witan Way Core is explained here at the level of its architecture, with its exact parameters kept private.

Witan Way Core reads four assets — US equities, gold, broad commodities and bitcoin — chosen for distinct roles: core performance, a macro hedge, an inflation-decorrelation hedge, and an asymmetric growth sleeve. Each asset runs two independent sub-strategies — eight in total. For each asset, the two are combined into a simple 50/50 average, producing a discrete exposure of 0%, 50%, or 100% per asset. A portfolio-level exposure ceiling keeps the whole book within a fixed limit. Keeping the weights discrete keeps the system legible and the turnover modest — this is tactical asset allocation done with a small number of honest states.

The backtests that motivate the design model real frictions: a cost per position change, a financing cost on any leverage, and a one-day lag between a signal and its execution (the close after the signal, not the close that produced it). Some costs — intraday slippage, market impact, taxes — are not modelled, so historical results should be read as an upper bound, not an expectation.

The design deliberately does not expose dozens of knobs to tune. Unlimited customization tends to become hindsight with a user interface; the richness lives in the explained logic, not in thresholds you set.

Limits we own

No strategy removes risk, and every strategy can lose money. Each family owns a specific trade-off:

Trend and momentum strategies lag at sharp reversals — a window that avoids noise also arrives late when a regime breaks hard.
All-Weather and risk-parity designs give up upside in strong, broad rallies and do not model tail co-movements, when correlations converge in a crisis.
Multi-strategy designs like Witan Way Core trade some upside for fewer and shallower drawdowns — absolute return is not the headline.
The four-asset universe is small by choice — focus over breadth, not a limitation we plan to remove.

And, as above, modelled costs are a floor, not a ceiling: real-world slippage, market impact and taxes are not included, so live results will differ from a backtest. Diversification softens asset-specific risk; it does not remove market-wide risk.

Cadence

Witan Way publishes on a weekly rhythm. The Charter — the market-regime reading — is published every Sunday. It is built for a weekly cadence, not constant inspection: quiet weeks are the norm, not a failure, and when the rule does not change, the Lab does not invent urgency.

Between Sundays, a signal is surfaced only when something material changes — a genuine shift in trend, momentum, or coherence, not daily noise. Rare signals are a feature, not a gap. Benchmarks used for comparison are rebalanced on a fixed quarterly schedule.

Governance

Witan Way is a council of rules, not a guru. The method is the product; the author is deliberately kept out of view, so that what you inspect is a process, not a personality. The council has five members — trend, risk, regime, discipline, and limits — and the work is meant to be inspectable: published rules, disclosed thresholds for the academic strategies, known strategy families, clear boundaries, and no personalised recommendations.

The editorial practice is bound by a signed code of conduct. Among its commitments:

No recommendation — readings describe regimes; they never tell you what to buy.
Indices only — regimes are read on reference indices, never on a named product.
A timestamped record — readings are not revised after the fact; the practice is not rewritten.
No claim of gain — the past behaviour of a strategy is no indication of its future behaviour.

A deterministic editorial check runs over every reading before a human sees it, screening for promissory language and for any private parameter — so what is published stays descriptive and leak-free.

History and versions

This methodology carries a version and a last-updated date, shown at the top of the page; material changes are recorded in the changelog below. The living, dated record of the readings themselves is the Charter archive: every weekly edition keeps a permanent, dated URL, and readings are never edited after publication, so the track record is public and verifiable.

Signals are computed on the closing price of the day and stamped in; the same code, the same prices and the same parameters always reproduce the same signal. That reproducibility is what makes the archive an honest record rather than a story told after the fact.

Changelog

v2026.07 — First public edition of this methodology page.

Educational, not advisory. Witan Way Core is a pedagogical illustration: its logic is explained in full while its exact parameters remain private. Descriptive market- regime readings are not investment advice, recommendations, or forecasts, and past performance does not guarantee future results.