TailorBirdCapital runs proprietary algorithmic strategies continuously refined through machine evolution — not human guesswork. Every strategy deployed is the survivor of thousands of battles. Only the elite make it to your capital.
Strategy Performance — Simulation & Live
Backtested equity curve on 12 months of real market data, overlaid with live paper-trading positions opened today. The simulation ran 32 qualifying trades. Live deployment began this session.
⚠ Simulation results are backtested on historical data and do not guarantee future performance. Live positions are paper trades on IBKR DUP444417. No real capital is at risk in paper trading mode.
The Origin Story
Every strategy in our portfolio has survived a brutal journey. Nothing is deployed because someone thought it sounded good. Everything is deployed because it proved itself — repeatedly, under pressure, against data it has never seen.
Thousands of strategy candidates are born simultaneously — each with a unique combination of rules, thresholds, timing parameters, and market conditions they respond to. No human selects these. The machine throws everything at the wall.
Every candidate is immediately put to work against hundreds of bars of real market history. Entry, exit, position sizing, stop losses — all simulated as if it were live. Strategies that can't produce a minimum number of qualifying trades are eliminated immediately.
Survivors are scored across five dimensions simultaneously: risk-adjusted return, consistency, capital efficiency, recovery speed, and trade frequency. A strategy that wins big but loses big is penalized. Only strategies with genuine, balanced edge move forward.
The top performers are selected as parents. Their genetic material — their rules, parameters, and structure — is combined in new ways, with controlled mutations introduced. The next generation inherits what worked, discards what didn't, and explores slight variations. This cycle repeats indefinitely.
Strategies that consistently rank at the top face a final test: performance on data they have never encountered. This is the difference between a strategy that memorized history and one that genuinely understands the market's structure. Only those that pass enter the vault.
Vault-certified strategies are deployed to live paper trading on real exchange infrastructure. Their live performance is tracked against their backtested profile. Strategies that underperform their predictions are automatically retired. The portfolio is self-cleaning.
The Gauntlet
Before a single dollar of capital is committed, every strategy must clear five sequential filters. Fail any one of them — eliminated. This is why our deployed strategies look nothing like typical algorithmic trading outputs.
A strategy must generate a statistically meaningful number of trades. Too few trades means the results are luck, not skill. Strategies that don't trade enough are discarded regardless of their win rate.
Peak-to-trough drawdown must stay within a hard limit. No strategy is allowed to "swing for the fences." Capital preservation is the first mandate — returns are secondary to survival.
The five-dimensional fitness score must exceed a dynamic threshold that rises as the population improves. A strategy that was excellent last week may be average today. The bar keeps rising.
The strategy is retested on market data outside its training window. Its fitness score on unseen data must be within an acceptable ratio of its training score — proving the edge is real, not overfit.
Once deployed, every strategy is continuously monitored against its predicted profile. If live performance degrades below a threshold relative to backtest expectations, it is automatically removed and replaced.
The Architecture
Six principles that distinguish machine-evolved strategies from conventional algorithmic trading.
The system never stops running. New strategy variants are bred, tested, and scored around the clock across multiple asset classes. The portfolio upgrades itself automatically as better strategies are discovered.
Capital is deployed across equity futures, precious metals, energy, crypto, and fixed income — independently signaled. Not correlated diversification. Structurally independent diversification.
Strategies include embedded market environment detection. They go dormant when conditions don't support their edge — protecting capital in choppy, directionless markets without human intervention.
Position size is derived from each strategy's historical performance profile using a conservative sizing methodology. Size scales with proven edge — not intuition, not fear, not greed.
Strategies that underperform their predicted profile are automatically retired and replaced. The portfolio has no emotional attachment to any single strategy — only to performance.
Orders execute on signal. No hesitation, no second-guessing, no overrides. The system doesn't panic in drawdowns or get greedy in winning streaks. Discipline is structural, not aspirational.
By The Numbers — Current Best Strategy
All figures reflect backtested performance of the current top-ranked vault strategy. Market: ES/S&P 500 · Timeframe: 4-hour bars · Direction: Long & Short.
Return Projections
Modeled from backtested strategy returns with real-world conservative assumptions applied. The base case uses a fraction of the backtested annualized return — because we plan for reality, not best-case. Adjust any input to model your own scenario.
Scenario Analysis — $10,000 Starting Capital
| Scenario | Monthly Return | Year 1 | Year 2 | Year 3 | Year 5 | 3-Year Return |
|---|
Early Access
TailorBirdCapital is currently in a select early-access phase. Reach out to discuss allocation sizing, onboarding, and reporting access.
Early investors receive priority allocation and direct performance reporting.