The QMAC Ladder
A moving-average scan from Robert T. H. Lee & Peter A. Tryde’s Timing Solutions for Swing Traders (Wiley, 2012) — what the setup is, how the author says to trade it, and what happened when we tested it over 21 years.
Read this before the rules
We did not just read this book — we tested it. Over 21 years and 259 monthly rebalances, the ladder did not beat the market after realistic trading costs. It is not an edge, and we do not trade it as one.
What it did do is control losses. Used as a filter on a trend book it cut the worst peak-to-trough fall from −51.3% to −30.9%, and that improvement held at every cost level we tested. That is the honest reason this page exists: the setup is a risk tool, not a money machine.
The one idea worth taking
Most of the book is standard technical analysis. One idea in it is original, and it is the only part we tested: the Queuing Theory of Moving Average Crossovers. In a genuine trend, price and its moving averages line up in strict rank order, shortest to longest. When they do not, Lee argues the move is borrowed time — “the respective rally or retracement most probably will be short-lived”.
Close > SMA50 > SMA90 > SMA200
Every term must sit above the next. This is a pass/fail gate, not a score — a stock that scores 8 out of 10 on this test is not a weaker buy, it is a reject.
The scan, and what it returned
Run on 29 September 2026 across every US symbol with usable history. Each tier removes names; we mark which rules are Lee’s and which thresholds we had to choose, because the book states a principle without giving a number.
The ladder removes 85% of liquid names in a single step. Everything after it trims; the ladder decides. That matters, because it means the other five tiers are convenience rather than the signal.
How the book says to trade it
These rules are Lee’s, not ours. He is explicit that this is a discretionary system: it produces candidates and context, not automatic orders.
“Look at the weekly, trade on the daily.”
Identify the trend on the weekly chart, time the entry on the daily, and never take a daily signal that fights the weekly direction. In Lee’s words: act in accordance with what has happened, not in anticipation of what will happen.
The 9/13 cross
The 9-day EMA crossing above the 13-day is the standard signal. Price closing above both is the stronger one, because price leads its own averages. Either is only valid while the ladder is intact and the weekly agrees — a trigger inside a broken ladder is exactly the case Lee tells you to skip.
1.3 x ATR(13), ratcheting
Stop at close minus 1.3 times the 13-day ATR, then held at the highest such value of the last 13 bars — so it trails a rising position and freezes when price falls. It never loosens. Lee publishes this formula and then writes, in the same paragraph, that he does not recommend trading it: the parameters are illustrative, not tuned.
Mirror of the entry
Exit when the 9-day crosses back below the 13-day, or price closes below both. Exit also when the ladder itself breaks — that is the entire thesis. The protective stop is the backstop for a move that gaps past your signal, not the plan.
Position sizing
Lee gives a range rather than a number: risk 1% to 5% of the portfolio per trade, and do not take the trade unless the potential reward is at least three times the risk.
Worked example — on a £100,000 account risking 1%, you can lose £1,000 on the trade. If the stop sits 4% below entry, the position is £25,000, because 4% of £25,000 is your £1,000. The stop distance sets the size; the size never sets the stop.
What we measured
2005 to 2026, 259 monthly rebalances, long-only, no leverage, costs charged on turnover. The benchmark is the equal-weighted eligible universe, which returned +457.5% over the same periods.
| Book | Cost | Total | CAGR | Max drawdown |
|---|---|---|---|---|
| trend | 30bps | 444.2% | 8.2% | -51.3% |
| qmac alone | 30bps | 308.1% | 6.7% | -35.5% |
| trend + qmac | 30bps | 374.2% | 7.5% | -30.9% |
| trend + qmac | 15bps | 573.2% | 9.2% | -30.2% |
| trend + qmac | 10bps | 656.5% | 9.8% | -29.9% |
| market | 457.5% |
It is not an edge. At realistic costs nothing here beats simply owning the universe. As a standalone strategy the ladder is worse than the trend signal it resembles on every measure except one.
That one measure is drawdown. Used as a filter the worst fall went from −51.3% to −30.9%, and it held at every cost level — so it is structural, not an artifact of the fee assumption. Return per unit of pain, measured as CAGR divided by worst drawdown, goes from 0.18 to 0.30.
Cost decides everything else. Filtering nearly doubles trading (turnover 0.52 to 0.91) because names cross in and out of the ladder. Below roughly 15bps that is worth paying; at 30bps it is not.
What we threw away
The financial astrology chapter
Lunar-phase entries and planetary aspects are not permissible for a Muslim investor, and the evidence offered is a quoted marketing line plus a 44-trade test on soybeans. Excluded entirely.
The volume ratio as a gate
On live data the 2-day versus 10-day volume comparison cut 374 candidates to 53, on a test that flips near-randomly day to day. Useful as a sort column; churn without a view as a filter.
The BIC colour indicator
The book describes what it does and never publishes the rule, so it cannot be reproduced or checked. We do not ship what we cannot verify.
Download the full guide (PDF, 5 pages)
The ladder, the scan funnel, entry, stop, exit, position sizing, the measured results, and the honest limits — in one document.
Honest limits
- It failed our test. Two of three pass conditions, twice — never all three. We publish no signals from it.
- No live evidence. Every figure above is a hypothetical backtest of a rule chosen after seeing the whole period.
- The book contains no backtest at all. Lee states plainly that his parameters are not optimised and his charts are illustrations chosen after the fact.
- Costs are assumed. Whether your fills land nearer 10bps or 30bps decides whether the filter helps or hurts, and only live trading answers that.
Educational research only. Nothing here is a recommendation to buy or sell any security, or advice on whether any strategy is suitable for you. Backtested results are hypothetical, carry the benefit of hindsight in the design of the rules, and are not an indicator of future results. Mezan does not sell trading signals. Source: Robert T. H. Lee and Peter A. Tryde, Timing Solutions for Swing Traders (John Wiley & Sons, 2012); Mezan research, measured 30 September 2026.