The figures here come from a backtest of Shishin over 1.258 US trading days (2021-05-12 → 2026-05-14). The rotation sleeve described is currently in paper-observation — computed and logged daily against the live book, but not yet placing live orders. Past performance does not predict future results. Shishin is a research publisher; this is research output, not personalised investment advice.
The most expensive position in most portfolios is the one nobody talks about: cash. Not the cash held dry on purpose — the cash that sits idle because the strategy simply hasn’t found anywhere better to put it.
The cost of doing nothing
Cash is not neutral. Over the long run, US cash and equivalents have returned on the order of 3–4% a year while US equities have returned roughly 9–10%. That gap — five or six percentage points, compounded — is the silent tax practitioners call cash drag. The rule of thumb is blunt: every 10% of a portfolio left uninvested forfeits roughly 10% of the return it could have earned, and because the shortfall compounds, decades of a few idle percent can cost six figures on a six-figure account.
A regime-aware, long-bias system has a structural cash problem. By design it sits out when conditions are wrong: across the Shishin backtest the system was fully in cash on a large fraction of sessions, and even when deployed it rarely uses every available dollar — a position cap, conviction sizing, and the gaps between entries all leave a residual balance. That residual is honest risk control. It is also, every single day, earning nothing.
A momentum sleeve, not a money-market
The conventional fix for idle cash is a money-market fund: collect the short rate and move on. Shishin’s answer is more ambitious — park the idle balance in a small, disciplined ETF rotation sleeve. Momentum rotation is a well-documented technique: ranking a basket by trailing returns and holding the leaders has shown a persistent edge in the sector- and asset-class-rotation literature, on the order of a few percent a year over buy-and-hold.
The sleeve’s rules are deliberately plain. Each month it ranks a broad universe of liquid, distinct-exposure ETFs by twelve-month momentum, keeps only those above their own long-term trend, and holds the three strongest — inverse-volatility weighted, so the calmer leader gets more capital than the jumpy one. It carries a volatility-scaled stop on each holding. And it has an off-switch: when the broad market is below its two-hundred-day average, the sleeve steps entirely to cash. The off-switch is the whole point — it makes the sleeve defensive by construction, out of the market in exactly the windows that hurt a long-bias book most.
The trinity result
Added to the backtested book as an idle-cash overlay, the sleeve did something rare. It improved all three of the metrics that usually trade off against each other — return, risk-adjusted return, and drawdown — at once. The five-year backtest moved from $6.78M without the sleeve to $10.81M with it: a lift of $4.03M (+59%) on the same capital. The Sharpe ratio rose +0.17 to 1.87, and the maximum drawdown got 2.4 points shallower, not deeper.
The shallower drawdown is the counter-intuitive part, and it is the mechanism worth understanding. A naive overlay that simply bought more equities with idle cash would have added risk. The sleeve does the opposite, because of the off-switch: it holds its momentum basket in good tape and sits in cash through bear phases, so its losing windows do not line up with the stock book’s. Two return streams that draw down at different times combine into a smoother curve than either alone. Idle cash stopped being a drag and became a diversifier.
One detail mattered more than expected: concentration. Holding the three strongest names beat holding the five strongest at the book level — three is concentrated enough to ride the actual leaders, broad enough to stay invested through a stop-out, and the full book absorbs the extra single-name volatility that made the tighter sleeve look riskier on its own. Judged alone, the concentrated sleeve had a lower Sharpe; judged inside the book, it was the better choice. The standalone number was the wrong yardstick.
Why it isn’t free money
Three honest caveats. First, these are backtest figures: the sleeve is in paper-observation, not yet trading live, and live execution will diverge for the usual reasons — fills, slippage, and the normal gap between a model and its sample. Second, momentum rotation has a known weakness: it lags during sharp, choppy reversals off a bottom, when a twelve-month ranking is still pointing at yesterday’s leaders. Third, the off-switch is a trailing signal — it protects against sustained downtrends, not single-day shocks, and it re-enters after a recovery is underway, not at the low.
None of that changes the shape of the finding. Cash that earns nothing is a cost paid every day whether anyone notices it or not. A disciplined, self-hedging momentum sleeve is one way to stop paying it — and, in the backtest, to be paid instead.