Rundown + Tear Sheet of the Systematic Long Only Strategy

Invite link to discord + rundown of strategy for paid members below:

Hey guys! Rolled out the strategy today and wanted to do a full write up of the logic of it & build out the discord channel to get the alerts in there as well. I’m live testing this with $10K and if we find it’s performing well, may add additional capital to try and scale it a little bit but if it truly has edge trading live, then we should compound nicely.

---

Strategy Overview

1. Summary

A fully systematic, long-only mean-reversion strategy for liquid US equities. It buys quality, liquid names when they become statistically oversold and begin to turn back up, holds a diversified book, and exits each position once the move reverts or a fixed time limit is reached. Everything is 100% rules-based — no discretion, no news, no narratives, no forecasting. The edge is a repeatable statistical pattern harvested broadly across hundreds of names, not a handful of high-conviction bets.

2. The edge (high level)

Markets overshoot to the downside short-term and tend to snap back. The strategy systematically identifies names that have been pushed statistically stretched below their recent trend and are showing the first signs of turning, takes a position, and harvests the reversion. It is a breadth strategy: each individual trade carries a small, positive expectancy; run across many concurrent positions with fast turnover, the law of large numbers compounds that thin edge into the return stream. It is deliberately not a momentum or stock-picking strategy.

3. Universe & selection

Broad, liquid US equity universe (~1,000 names) filtered purely on characteristics — liquidity, price level, and sufficient daily range. Names too illiquid, too cheap, or too quiet to trade cleanly are screened out.

• Sector-agnostic. Signals fire wherever names get oversold; there is no sector tilt or rotation.

• No performance-based selection. The universe is never narrowed to “names that did well.” This was validated the hard way — walk-forward testing showed that picking the historically best names hurts this strategy (it overfits and collapses out-of-sample). The universe is defined by characteristics only.

4. Portfolio construction

• Diversified and equal-weight — every position is sized to the same target weight; no single name dominates.

• ~20 concurrent positions max, capped per name, so idiosyncratic blowups are contained (a single name is a small fraction of the book).

• Fast turnover — average holding period is short (positions exit quickly when the bounce works), so capital recycles many times a year and compounds the per-trade edge.

• The book scales into developing setups to improve the average entry.

5. Risk management

Three layers, deliberately no per-trade price stops (testing showed tight stops chop winners and a wide “catastrophe” stop actually deepened drawdowns by clustering losses in crashes):

1. Per-trade time cap — a position that hasn’t reverted within a set window is closed, so non-reverters can’t bleed indefinitely.

2. Portfolio diversification — ~20 uncorrelated-ish positions average out single-name noise; sizing bounds any one disaster.

3. Volatility-targeting overlay — total book exposure is scaled to hold risk near a steady target, using modest margin leverage in calm markets and automatically de-risking when volatility spikes. Because large losses cluster in high-volatility regimes, the book is naturally smallest exactly when crashes hit — which roughly halved the worst historical drawdown while preserving the return.

6. Execution

• Enters at the open (market-on-open) and exits at the close (market-on-close), trading the deepest-liquidity auctions of the day — no after-hours.

• Liquid names only, so spreads are tight; all results are stated net of realistic transaction costs.

7. Risk profile

High-octane long-equity exposure: it co-moves with the market and does so with higher volatility and fatter tails. Its return distribution is many small wins plus occasional sharp losses, smoothed at the book level by diversification — except in crashes, when correlations spike and the diversification temporarily fails. That single failure mode (fast, broad selloffs) is the entire risk story, and it’s what the volatility overlay is designed to manage.

8. Methodology & integrity

• Costs are mandatory — no gross/headline numbers; everything net of realistic slippage + commissions + margin costs.

• No look-ahead — signals act on completed data; entries fill the next open.

• Walk-forward validation for any selection claim — no in-sample-picked universes.

• Broad, not lottery-ticket, edge — the top 1% of winners account for only ~16% of gross profit, so no single moonshot carries the record.

9. Current status

The strategy is running as a live forward-test — daily buy/sell signals and a running equity curve published in real time, building a genuinely out-of-sample track record from here forward.

Full Tear Sheet + Backtest Results + Discord Link