All figures in this post are from paper trading — simulated trading without real capital — via Alpaca. Nothing here is investment advice.
Thursday 2 July, end of day: Apex Quantum +0.34% · S&P 500 −0.24% · NASDAQ 100 −1.48%. The simulated portfolio finished roughly 1.8 percentage points ahead of the NASDAQ, on a day when most things fell.
Any competitor would have written that in capital letters, buy button right below. We won't — not because the number is wrong, but because the conclusion would be.
One day proves nothing
Let's take the brag apart first: a single trading day is statistical noise, and one datapoint is not representative of long-term expected return — we wrote that in May, and it still holds. A portfolio with 30.8% in a single biotech name may well sit ahead of the NASDAQ when tech falls — and lag it the next time the index rips higher.
The last 24 hours in the simulated environment: +0.22% on the virtual balance. Pleasant. Proof of nothing.
What the day actually shows
The interesting thing about 2 July isn't the return figure, but that you can open the cockpit and see exactly what the engine did — and why. The decision log, as published, unedited:
"Hold 7 positions; top-up ASML (rising channel, RSI 58<65, small size) as 8th slot fill. No new external buys — strict RSI>65 filter and low RS on alternatives block adds. Maintain always-invested in uptrend names."
In plain language: hold all seven positions and use the eighth slot for a small ASML top-up — rising channel, RSI 58, still below the threshold. And most importantly: no new buys. The RSI filter and weak relative strength blocked every alternative — nothing qualified, so it bought nothing.
The filters that said no
Three rules did the work on Thursday (all figures still simulated):
- The RSI gate. The engine buys nothing new above RSI 65 — that's where the chase-the-train buys live, the ones that feel best and hit worst. ASML got through at 58 — even then only at small size.
- Relative strength. A stock rising less than the market isn't strong. VRTX is held as "High RS leader, RSI 56 ok, hold winner" — the largest position at 30.8% weight, up +7.50%.
- Fully invested in confirmed uptrends. SNOW (+11.24%) is held as "Rising channel leader, volume accumulation"; ABSI (+16.89%) is left to run as long as the trend is intact. Not selling a winner too early is the part humans struggle most to follow.
The losing side is in the log too
An engine that only shows you winners is lying. On Thursday the simulated portfolio also held HELP at −1.88% ("Priority-core, RSI 39, hold for rebound setup"), QBTS at −3.08%, ASML at −2.18%, and TSM. The portfolio sat −1.51% below its peak — within the −2.07% drawdown target, a target the engine steers by, not a guarantee against loss. Losses happen. The point is that every position in the red carries a logged reasoning you can read — and disagree with.
What you actually pay for
Apex Quantum doesn't sell returns and doesn't promise them — one good Thursday says nothing about the next. What we sell is technology: an engine that analyses, decides and executes by a fixed rulebook, without getting greedy or scared. Defined risk limits. And transparency: every decision is published with its reasoning — on green days and red ones. The customer pays for the engine, the discipline and the insight, not for a promise of results. The decision to use the tool — and the risk that comes with it — is yours.
Where we stand
Max still runs exclusively in paper trading via Alpaca. The licence application for AI-enabled advisory services (RAIEAS, Financial Services Commission, Mauritius) is under review — we are not licensed today. All trading in securities carries a risk of losing the entire invested amount, and historical and simulated results are no guarantee of future returns.
One good day against the NASDAQ proves nothing — we know that. But the log sits open in the cockpit, so you can audit every call yourself. That is the whole point.
— Andreas, founder of Apex Quantum