Track record
4 strategies. 5 years. Judge for yourself.
We backtested four approaches across 249 ASX small caps from March 2021 to March 2026. Every result is here — the wins and the losses.
We backtested four approaches across 249 ASX small caps from March 2021 to March 2026. Every result is here — the wins and the losses.
Four strategies, same market data, different risk appetites. Click any card to see its full performance breakdown below.
The backtest above is a portfolio simulation. The same backtest file also runs statistical robustness checks that significantly qualify the headline numbers:
"likely_noise". The deflated Sharpe adjusts for multiple-strategy testing and overfitting; a value of 0 means the observed performance is not statistically distinguishable from chance after adjustment.In plain English: the +X% portfolio return is real as a simulation, but the strategy has not demonstrated statistically significant skill once we account for multiple testing, outlier concentration, and cost. This is exactly why we are building a live track record — it removes these biases.
We checked if our scores actually work — no cheating allowed
We went back through 5 years of real stock market data and pretended to run our scoring system each day. The important rule: on each day, we only used information that was actually available at the time. No peeking at tomorrow's prices. Then we checked — did the high-scoring stocks actually go up after 1 week, 2 weeks, and 1 month?
We left out annual report data on purpose. Those reports come out months late, so using them would be like reading tomorrow's newspaper — that's cheating.
Three simple steps, same process every time
When a stock scores high enough, we write down the price and start the clock. Everyone can see it — no hiding.
Each signal gets an automatic safety level. If the stock goes up, the safety level moves up too — like a rising floor that protects your gains.
Win, lose, or draw — we log it all. Hit the target? Great. Hit the safety net? Recorded. Ran out of time after 45 days? Also recorded. Nothing hidden.
Past performance is not indicative of future results. These metrics reflect historical algorithmic screening outcomes and do not guarantee future returns. Signals are data points, not recommendations.