Beat the index.
Lose less doing it.
A systematic position book for iPhone. Seven names per sleeve. One rebalance every twenty-eight days. Price math only.
$100 in. $669 out.
Over the same window SPY turned into $545.
Both books normalized to $100 at the start of 2013. Monthly walk-forward, the same point-in-time Russell 1000 universe, the same fifteen basis points of round-trip cost. The lavender line is the book. The dashed line is SPY.
$669.62 vs SPY $545.63
Backtest results computed via walk-forward + combinatorial purged cross-validation on the point-in-time Russell 1000 universe. Hypothetical results do not reflect actual investment results and are not a guarantee of future returns. Live track record begins May 7, 2026.
Price data in. Seven names out.
The algorithm runs once a day, in four passes. Each pass is deterministic, point-in-time, and survivor-bias free. The code that ran on Tuesday in 2018 is the code that runs today.
Point-in-time Russell 1000
About 1,140 U.S. tickers per snapshot, sourced from monthly iShares IWB membership lists. Names need eighteen months of history before they can be ranked, so newly listed tickers are excluded rather than zero-filled.
Five signals, one composite
Trailing risk-adjusted momentum, residual momentum after Fama-French five-factor controls, post-earnings drift, industry momentum, and a quality screen. Each is normalized to a cross-sectional percentile.
Rolling statistics, three bands
A sixty-bar window over the index's volatility and trend classifies the market as risk-on, cautious, or defensive. That dial sizes the cash slice and the split between the two sleeves.
Seven slots, three caps
Top-down walk under each sleeve. Maximum five per GICS sector, three per industry group, plus an idiosyncratic vol cap. Whatever doesn't fit goes to cash via the BIL proxy.
Toggle the regime. Watch the book reshape.
Three regimes, three allocations. In risk-on, ten percent sits in cash and the sleeves run nearly equal. In defensive, more than half the book is parked. The regime classifier picks the bucket. The bucket picks the allocation.
The app.
One screen per card. The book is a position to read, not a list to scroll.
Why this category.
Active retail flow is growing. The cohort that wants algorithmic discipline without surrendering control is underserved on mobile. The published track record above is the moat.
The thesis, in three lines.
The active-retail sleeve is taking flow. Robinhood's assets under custody grew 88% to $193B in 2024. Schwab's grew low single digits. Mobile-native, active-leaning brokerages are eating share, but the only systematic products on those rails are robo-advisor passive (Wealthfront, Betterment) or AI-prompt portfolios (Public).
The cohort wants discipline without handing over the keys. Existing options trend toward two poles, full delegation to a robo-advisor or full discretion in a brokerage UI. The position-trader middle, weeks-to-months holds with a fixed cadence, has limited native mobile tooling.
The defensibility is a published, repeatable track record. A fourteen-year walk-forward backtest with CAGR, Sharpe, and max drawdown sits on the homepage above. Every month the cron runs adds another point to that curve.
| Product | Pricing | Published risk record |
|---|---|---|
| Composer | $24-32 / mo | No |
| Public.com | $10 / mo Premium | No |
| Robinhood Strategies | 0.25% AUM | No |
| Wealthfront | 0.25% AUM | No |
| Atom Finance | n/a | n/a |
| S7 | TestFlight | Above |
Get the book in your pocket.
S7 ships on iPhone via TestFlight. Limited beta seats while the rebalance cadence stabilizes.