MBW Capital is Worsley Capital's systematic, algorithmic trading venture, currently in development — a proprietary engine designed to decompose market prices into their underlying frequencies to identify turning points early, and to trade them with discipline once live.
Everything that moves a market — flows, positioning, sentiment, news — ultimately expresses itself in one place: price. So that is what MBW measures. No forecasts, no stories, no committee views. A systematic process, applied the same way in good markets and bad.
Models are driven by price and price movement alone — not by claimed inefficiencies, fundamentals feeds or discretionary judgement. Price is the one input that never stops telling the truth.
Any single indicator produces false positives. Agreement across twelve independent frequency bands is required before a signal is actionable.
Entries, exits, profit-taking and stops are rule-based, defined before a position exists. No overrides, no averaging down, no hoping.
At the core of MBW is a proprietary signal engine built on years of research into the spectral behaviour of market prices. Rather than watching one smoothed line, the engine separates each asset's price history into twelve frequency bands — twelve independent views of the same market, each tuned to a different cycle length. In backtesting, the engine hunts for one thing: acceleration in price movement occurring in sync across the bands. When the spectrum aligns historically, a top or bottom tends to form, typically three days to three weeks before the turn completes. This research and backtesting is the foundation the live engine is currently being built and tested against, ahead of launch.
Price history per asset, used for research and backtesting, updated through the latest close.
Each series is separated and instrumented with the full sensor array.
Meaningful changes in the rate of price change — not just direction — are flagged within each band as they occur.
Signals firing in isolation are discarded; agreement across bands marks a candidate reversal in testing.
Every asset under research is scored by signal strength; the strongest candidate reversals are carried forward for review.
Within each of the twelve bands, the sensor array is designed to take continuous readings across three properties — not just whether a band is rising or falling, but how, and how convincingly.
Where a band sits within its own cycle. When multiple bands reach the same point of their respective cycles together, the market is designed to be read as being at a decision point.
The rate and steepness of change in price movement. A velocity spike against the prevailing trend, followed by a turn back through centre, is treated as an early footprint of a reversal.
How strongly the bands reinforce one another. High resonance means the spectrum agrees; low resonance means a move is designed to be read as noise, however dramatic it looks on a chart.
A conventional chart shows one price line — the sum of many overlapping cycles of different lengths, all added together and averaged into a single, noisy trace. A short-term swing, a medium-term trend, and a longer structural move can all be happening in the same instrument at once, and on a single smoothed line they are indistinguishable from each other. That is the problem the engine is built to solve.
Rather than smoothing price into one line, the engine decomposes it into twelve frequency bands — short, medium and long cycle-lengths considered independently, each retaining information the others discard.
Each of the twelve bands is measured by the full sensor array — over 600 readings per asset in total — tracking acceleration and deceleration within that specific cycle length, not just its direction.
A single band accelerating means little on its own — short cycles turn constantly. A candidate signal only exists, in testing, when multiple independent bands accelerate into alignment at the same time.
Because longer bands change more slowly, alignment across the spectrum has, in backtesting, tended to precede a turn in the underlying price rather than confirm it after the fact — the basis for the lead time referenced above.
Description of research methodology and design intent. MBW has not yet launched and is not currently trading live capital — nothing here is a claim about live signal accuracy.
MBW's flagship application, as designed, is sector rotation: anticipating imminent tops and bottoms across sector and index ETFs, and rotating exposure toward strength. The same engine is intended to drive bespoke long-only and long/short mandates once live. Sizing is systematic by design; monitoring, execution quality, data integrity and risk limits are intended to be supervised by the team, with signal logic never overridden on a view — the system decides, people supervise.
In backtesting against historical data, this systematic approach has performed ahead of broad market benchmarks such as the S&P 500, on both an absolute and a risk-adjusted basis. MBW is not yet trading live capital, so these results reflect research and design targets, not a live track record.
Strategies are designed to be expressed in ETFs, index products and stock baskets — reducing single-name risk and allowing for institutional scale without material market impact.
Not designed for high-frequency trading. Target holding periods run three to six months, with roughly twenty to thirty trades per portfolio per year.
Strategies are designed to stand on their own without leverage. Where a mandate calls for it, leverage would be applied deliberately — never to rescue a position.
Performance, risk, and compliance are designed to be reviewed on an ongoing basis, separate from the activity that generates returns.
Universe, base currency, risk tolerance, horizon, long-only or long/short — each portfolio is intended to be configured to the client. The discipline is fixed; the expression is tailored.
Illustrative only — based on research and backtesting, not live trading results. MBW has not yet launched and is not currently trading live capital. Past performance is not indicative of future results. Nothing on this page constitutes an offer, solicitation, or investment advice.
Once live, a mandate is intended to move through three stages, in order, with no step skipped and no discretionary override inserted between them. This is the operating pipeline the engine and the risk framework above are being built to support — described here as design intent, since MBW is not yet trading live capital.
Liquid ETFs, sector indices and stock baskets, set per mandate with base currency, horizon and risk tolerance — basket-level exposure keeps stock-specific risk low and capacity high.
The engine is designed to run the full universe and issue buy, sell, cover and exit signals under fixed rules — with take-profit and stop-loss levels attached before any order exists.
Signals are weighed against existing positions and mandate constraints to form the target book. Sizing is systematic by design; monitoring and reporting are intended to run continuously.
Description of a designed operating pipeline, presented for informational purposes. It is not a description of live trading activity. MBW has not yet launched and is not currently trading live capital, and nothing here constitutes an offer, solicitation, or investment advice.
MBW is currently in development. Rather than describe that as a single undifferentiated "not live yet," it is worth being specific about what development actually means here — a sequence of stages, each with its own bar to clear, moved through deliberately rather than rushed.
Developing and refining the spectral decomposition approach itself — which frequency bands matter, how they should be measured, and how signals across bands should be defined and combined.
Testing the resulting rules against historical price data across assets and market regimes, examining not just returns but drawdowns, false-signal rates, and sensitivity to assumptions.
Running the engine forward against live, real-time prices without committing capital — checking that signal behaviour observed in backtesting holds up outside of historical hindsight.
Trading real capital, sized conservatively at first and subject to the same risk oversight described elsewhere on this page. MBW has not yet reached this stage.
Description of a development process, presented for informational purposes. It is not a timeline, launch date, or forecast, and does not constitute an offer, solicitation, or investment advice. MBW has not yet launched and is not currently trading live capital.
No. MBW is currently in development. It has not yet launched and is not trading live capital. Every figure and chart on this page reflects research, design targets, and backtesting against historical data — not a live track record.
No. Nothing on this page constitutes an offer, solicitation, or investment advice. It describes a venture in development, for informational purposes only.
Entries, exits and position sizing are rule-based and defined before a position exists. The team's role, once live, is intended to be monitoring execution quality, data integrity and risk limits — not overriding signal logic on a discretionary view.
Sector and index ETFs are liquid, diversified, and well suited to a systematic process that needs clean, continuous price history — reducing single-name risk while allowing the same underlying engine to scale toward institutional size.
Via Worsley Capital directly — see Contact below. MBW does not currently publish a standalone signup or live results feed, consistent with not yet trading live capital.
MBW is being designed with professional and institutional investors in mind, not as a retail product. Once live, access is intended to be extended to a limited number of investors directly, consistent with the discretion the rest of the platform applies.
“The system decides; people supervise. Signal logic is never overridden on a view — that separation is the point.”
MBW Capital · Investment Doctrine