All signal, no timing.

The momentum core is textbook 12-1

Total return from t−252 to t−21 trading days, with that 21-day gap deliberately excluded to dodge the short-term reversal effect. On top of that: close above the 200-day SMA, and positive six-month relative strength versus the home index (FTSE All-Share for the UK, S&P 500 for the US). A name has to be trending on an absolute basis and beating its benchmark — three momentum conditions, all binary pass/fail.

Then it ranks cross-sectionally: survivors get a percentile score on 12-1 return, averaged 50/50 with a percentile score on cheapness, and the top names take the twelve slots. This is rotational RS — not "is this a good chart", but "is this among the strongest of what's left after the filters".

The fundamental gates are the universe definition

Positive free cash flow, growing EPS, controlled leverage, cheapest 30% on EV/EBIT. Functionally this is a very aggressive pre-filter on the momentum universe — it strips out exactly the high-beta, story-driven, no-earnings names that generate the fattest momentum runs. You get slower, uglier trends in cheaper stocks. Fewer 300% winners, fewer −80% unwinds.

Now the parts you'll object to

Resolution is monthly. Full stop. Signals are computed once, on the first Saturday, off the prior Friday close. Execution is Monday at market. Between rebalances the system does not look at price at all — no daily check, no alerts, no intraday anything. A name can gap 40% on the Tuesday and nothing happens for four weeks.

There are no stops. Not wide ones, not trailing ones, none. The 200-DMA evaluated monthly is the stop, which means the real stop distance is however far below the 200-DMA the name has travelled by the time the next screen runs. On a fast break that's a materially worse exit than any stop you'd place. The stated reasoning is position size: at roughly £85–165 a slot depending on phase, a 15% stop protects £13–25 while adding spread cost, whipsaw exits, and — the actual concern — a daily reason to look at the screen.

No entry timing whatsoever. No pullback entry, no breakout confirmation, no support and resistance, no volume analysis, no structure. If a name qualifies on Saturday it's bought Monday at market, whether it's extended three ATRs above the moving average or sitting on it. Positions are equal-weight in cash terms — not volatility-adjusted, not ATR-sized. A quiet defensive and a wild small cap get the same money.

And no chart inputs at all, anywhere. No Ichimoku, no oscillators, no patterns, no divergence — an explicit constitutional exclusion in this project, not an oversight. The only price inputs that exist are the 200-day SMA and two lookback returns.

Where it should behave well — and where it will hurt

Sustained trending markets, and — more interestingly — bear markets. When breadth rolls over, the momentum gates fail across the board, holdings exit at the monthly screen, and the slots simply stay in cash. Nothing is relaxed to stay invested. The system de-risks structurally rather than by decision.

The pain is chop. Monthly rebalance with a rank-25 buffer means names oscillate in and out around the boundary, and every round trip pays spread plus 0.15% FX each way on the US side. That whipsaw drag in neutral regimes is pre-registered as an expected cost, not a surprise.