Asset-Class Scorecard: Where Our Signals Worked Best This Quarter

Not All Markets Are Created Equal — And Our Data Proves It

The central question in systematic trading is rarely whether a strategy works in the abstract. It is whether it works here, in this asset class, under these conditions. A signal engine that generates alpha in currency pairs may bleed steadily in commodity futures. A mean-reversion model that thrives in liquid equities can be systematically destroyed by trend-driven, news-sensitive markets. This report does not argue that our signals are universally good or bad. It argues something more useful: that performance is not uniform across asset classes, and that understanding where a system struggles — and why — is the most actionable intelligence we can offer our readers.

Over the past 90 days, our proprietary backtest and live signal system generated 30 closed signals across three categories: forex, stocks, and commodities. The headline numbers are not flattering. An overall win rate of 33.3% and an average trade return of -0.74% are, by any honest measure, below the threshold of a net-positive system. We are publishing them anyway. Precision about failure is how improvement begins.

The Scorecard: Forex Leads a Weak Field

The 90-day data breaks cleanly into three tiers, and the separation between them is significant enough to warrant serious attention — even with the caveat, stated plainly here, that no single category reaches the sample size typically required for statistical confidence.

Forex is the relative standout. With 16 signals — more than half the total signal volume — the category posted a 43.75% win rate and an average return of just -0.04% per trade. That average return figure deserves a moment of careful reading. It is negative, but barely. In a system where losers across the other two categories are running at -1.05% and -2.19%, a forex average loss of four basis points is effectively a rounding-error deviation from breakeven. The win rate of 43.75% also sits meaningfully above the overall system average of 33.3%. Forex is not delivering profits, but it is where the system is most competitive with market noise.

The structural reasons are not difficult to identify. Currency pairs, particularly the majors, are among the most liquid and technically well-behaved instruments in global markets. Bid-ask spreads are tight. Price action is continuous, 24-hour, and less vulnerable to the gap risk that can catastrophically distort a signal's expected return at the moment of execution. Our signals — which appear to be directional in nature based on the win/loss framing — are more likely to find a clean entry and exit in forex than in a thinly traded commodity contract or a stock subject to overnight news flow. This is not a vindication of the signal logic. It is a structural compatibility argument.

Where the System Breaks Down: Stocks and Commodities

Stocks generated 8 signals over the period. The win rate was 25.0% — one in four trades — and the average return was -1.05%. This is a materially worse outcome than forex on both dimensions. A 25% win rate implies that the system is directionally wrong three times out of four in equities. At -1.05% average return, losses are not being offset by large wins; they are compounding into a drag that would erode a dedicated equity allocation meaningfully over time.

It is worth pausing on the sample size here. Eight signals in 90 days is a thin dataset. We cannot responsibly draw firm conclusions about the system's equity performance from eight data points. What we can say is that there is no positive signal in these numbers — no hint of emerging edge — and that the direction of evidence is consistently negative. That is enough to warrant caution, even if it is not enough to constitute proof of persistent failure.

Commodities is the sharpest underperformer. Six signals. A win rate of 16.67% — one winning trade out of six. An average return of -2.19% per trade. These numbers are, relative to the other categories, alarming. A system generating a sub-17% win rate is performing worse than a coin flip in one direction, consistently. And at -2.19% average return, the losses per trade are roughly double those in equities and more than fifty times the forex figure. If this system were running a live portfolio with equal allocation across categories, commodities would be doing the bulk of the damage.

Why might commodity signals underperform so sharply? Several structural hypotheses are worth considering. Commodity markets are sensitive to geopolitical disruption, weather events, and supply-chain developments that are fundamentally non-technical in nature. A momentum or mean-reversion signal built on price history will be systematically blindsided by a refinery outage or a crop forecast revision in a way that a currency signal — anchored more tightly to macro rate differentials and flows — will not. Additionally, many commodity instruments carry roll costs and contango dynamics that create a persistent headwind for long directional positions. These are not excuses. They are structural mismatches between signal design and market mechanics that are correctable if we choose to address them.

What This Data Does Not Tell Us — And Where We Could Be Wrong

Intellectual honesty requires naming the limitations of this analysis as explicitly as its findings.

First, sample size constrains all conclusions here. Thirty total signals across 90 days — 16 in forex, 8 in stocks, 6 in commodities — does not meet conventional thresholds for statistical significance in any category. A researcher requiring 95% confidence would need substantially more observations before drawing firm inferences about expected win rates. We are not in that position. These results are preliminary and should be read as directional, not definitive.

Second, we do not know the return distribution within each category. An average return of -2.19% in commodities is consistent with many different scenarios: six moderately bad trades, five small losses and one large loss, or some other configuration. If the commodity losses are driven by one or two outlier trades rather than consistent directional failure, the strategic implication changes. We do not have that granularity in the current data block, and we are not going to speculate about it.

Third, we do not know whether market regime explains the results. The 90-day window ending in mid-April 2026 is a specific slice of market history. Commodity markets may have been in a particularly trend-hostile or volatility-compressed regime during this period. Equity markets may have been experiencing the kind of intraday reversal conditions that punish directional signals regardless of quality. A different 90-day window might produce materially different rankings. We genuinely do not know yet whether the forex outperformance is structural or coincidental to this particular period.

Fourth, win rate alone is an incomplete measure of a system's merit. A strategy that wins 33% of the time but whose winners are three times larger than its losers is mathematically profitable. The average return numbers in this dataset suggest that is not occurring — the negative averages across all three categories imply that winners are not compensating for losers at any tier — but without median return, maximum drawdown, or a full distribution of outcomes, we cannot be fully certain. This is a known gap in the current reporting format, and we intend to address it in future editions.

What Traders Should Do With This Information

Analytical honesty without practical direction is just confessional writing. Here is what we think the data supports, at this stage, for traders using our signal feed.

• Weight forex signals more heavily, for now. The 16-signal, 43.75% win-rate, near-breakeven average return in forex is the closest thing to a functioning edge in this dataset. That is not a high bar — but it is a real one. Traders allocating risk across categories should be skewing exposure toward forex signals until the other categories demonstrate improvement. This is not a permanent recommendation. It is a current-data recommendation.
• Trade commodities signals at reduced size or paper-trade them. Six signals and a 16.67% win rate at -2.19% average return is not a foundation for full capital deployment. Until we have more signals and a clearer picture of whether the underperformance is structural or regime-specific, commodities exposure should be minimized. If you are already in commodity trades based on our signals, apply disciplined stop-loss management. Do not average into losing positions in the hope of mean reversion.
• Treat stock signals as under review. Eight signals is not enough to condemn the equity category, but 25.0% win rate offers no encouragement either. Equity traders should monitor the next 10 to 15 stock signals closely before drawing firm conclusions. If the win rate does not move toward 35-40% with additional data, that will become meaningful evidence of a systematic problem.
• Use this scorecard as a dynamic tool, not a fixed verdict. We will publish updated category-level data as signal volume grows. The rankings in this report reflect 90 days and 30 signals. They are the beginning of a performance map, not the final word on it. The most important discipline for traders reading this is to update their priors as new data arrives — not to lock in conclusions from a thin dataset.
• Consider position sizing adjustments based on category. Even within a fixed risk budget, allocating equal notional exposure to forex and commodities signals is not rational given the current return disparity. A simple approach: size forex signals at full unit risk, stock signals at half unit risk, and commodity signals at quarter unit risk, until category-level performance stabilizes. This is not a precise optimization. It is a common-sense response to unequal evidence.

Methodology Note

The data underlying this report covers 30 closed signals generated by Stocks365's proprietary backtest and live signal system over a 90-day trailing window ending April 18, 2026. Signals are distributed across three asset classes: forex (n=16), stocks (n=8), and commodities (n=6). Win rate is defined as the percentage of closed signals that produced a positive return. Average return is the arithmetic mean of all individual signal returns within each category, expressed as a percentage. No signals are excluded from the dataset. These sample sizes — particularly in stocks and commodities — are below conventional thresholds for statistically robust inference, and all findings in this report should be treated as preliminary and directional rather than conclusive. We do not have visibility into the return distribution, maximum drawdown, or Sharpe ratio at the category level from this dataset, which limits the depth of our analysis. Future reports will incorporate those metrics as signal volume increases. Our commitment is to publish the numbers we have, not the numbers we wish we had.