The Propulsion Room: Thrust, Drag, Gravity, and Steering
in the Net-Return Accounting of Eleven Measured Strategies
In one line: Of 11 strategies I measured, only a third made money before costs, and most lost to index funds.
Abstract. A portfolio’s net return is decomposed into four forces by analogy with powered flight: thrust (gross edge before costs), drag (the friction of trading), gravity (the volatility drag of compounding), and steering (the cost of each rebalance). Using backtest telemetry from 11 strategies with complete cost accounting, about a third produced positive gross thrust. Among those, the median cost drag was 42% of gross thrust, and the median breakeven cost, or stall speed, was about 12 bps per unit of turnover, against a model assumption of 5 bps and realistic institutional costs of 10 to 20 bps. Median annual turnover was approximately 10×, reaching 19× at the extreme. One strategy earned a positive average return yet compounded to a loss, turning $100k into $93k over twenty years, because its variance load exceeded its arithmetic lift. The governing conclusion is that gross edge net of drag and volatility drag must exceed a passive benchmark, at realistic friction and out of sample, before an active strategy is worth flying.
1. Introduction
Powered flight balances four forces, and a portfolio’s net return admits the same decomposition. Thrust is the gross edge before costs; drag is the friction of trading; gravity is the volatility drag of compounding; and steering is the cost of each course change. The decomposition matters because a positive gross edge is neither necessary nor sufficient for a positive net result, and drag ends more strategies than weak signals do. This note maps the pipeline’s backtest telemetry onto the four forces and measures each. Table 1 states the correspondence.
| Flight force | Portfolio force | Where it is measured here |
|---|---|---|
| Thrust | Gross edge, before costs | backtest gross return |
| Drag | Friction: costs × turnover, slippage, borrow | 5bps per unit turnover + 25bps/yr borrow |
| Gravity | Volatility drag: growth ≈ return − σ²/2 | the $100k→$93k plaques |
| Steering (control inputs) | Trading: every course change burns fuel | annual turnover × cost per trade |
2. Measured forces
2.1 Cost drag
Of the strategies with full cost telemetry, about a third produced positive gross thrust. Among those, the median friction ratio, defined as cost drag as a share of gross thrust, was 42%. Nearly half of the gross edge was consumed by trading before any return reached the account. The median stall speed, the per-unit-turnover cost level at which net return reaches zero, was about 12 bps. The cost model assumes a generous 5 bps per unit of turnover; realistic institutional all-in costs for less liquid names run 10 to 20 bps. A strategy whose breakeven is 12 bps is viable only under the model’s cost assumption, not under realistic friction.
2.2 Steering and turnover
Each course correction, or rebalance, pays the friction toll, so steering authority must be budgeted like fuel. The telemetry median is approximately 10× annual turnover; the highest-turnover strategies reach approximately 19×. The governing constraint is the signal half-life check: a strategy whose signal half-life is shorter than its rebalance period pays turnover to chase noise rather than to track a persistent edge. This check is a stage-4 standard in the pipeline.
2.3 Volatility drag
Volatility drag is the gravity of compounding: geometric growth is approximately the arithmetic return minus half the variance. One strategy in the sample had a positive average return yet compounded to a loss, because its variance load exceeded its arithmetic lift; a positive average turned $100k into $93k over twenty years. Leverage scales thrust and volatility drag together, which is the mechanism behind structural failures under load such as LTCM, documented separately in the Case Studies.
3. Benchmark and discussion
A passive index fund is the limiting case: near-zero thrust of its own, single-digit-bps drag, near-zero turnover, and a high ratio of return to cost, so compounding does the flying. Most active strategies carry large gross thrust but also large drag and turnover. The standing finding across this sample is that most of the active strategies tested did not exceed the passive benchmark once their own trading costs were charged. The bar for an active strategy is therefore explicit: gross thrust net of drag and volatility drag must exceed the passive benchmark, at realistic friction, measured out of sample.