Combining Multiple Signals in a Portfolio

Individual signal benchmarks show how each market performs in isolation. They don't show what happens when you run more than one signal at the same time, or whether the results compound, offset, or produce something in between.

To find out, we ran backtests on three portfolio configurations covering January through May 2026. Each portfolio is built from independent strategy backtests using the same methodology as our individual benchmark reports: fixed-size trades every five minutes, no leverage, no optimization, all positions closed at end of day. The full strategy is described here. The short answer: in every configuration we tested, combining signals improved at least one dimension of risk-adjusted performance. Which dimension depends on the signals in the mix.

Portfolio A: Equity index signals

Composition: S&P 500 (33%), Nasdaq-100 (33%), Russell 2000 (34%)

Table 1: Portfolio A performance summary

Table 2: Portfolio A monthly returns (%)

Equal-weighting three equity index signals lands at Sharpe 0.969, below the best individual signal in this period (Russell 2000 at 1.080). What the combination does buy: lower portfolio volatility (4.12% vs 7.08% for Russell 2000 standalone) and a tighter max drawdown (-3.04% vs -3.19%). The three signals don't all move together month-to-month (January favored S&P 500, February favored Russell 2000, March and April were mixed), but they're driven by the same broad equity market order flow, which limits how much diversification the combination can achieve. Intra-asset combination reduces dispersion; it doesn't reliably lift Sharpe.

Portfolio B: Equity index + metals signals

Composition: S&P 500 (24%), Nasdaq-100 (24%), Russell 2000 (24%), Gold (20%), Silver (8%)

Table 3: Portfolio B performance summary

Table 4: Portfolio B monthly returns (%)

Adding Gold and Silver to the equity index blend changes the picture materially. Portfolio Sharpe moves from 0.969 to 1.654. Return moves from 1.76% to 3.23%. Max drawdown tightens to -2.61%. All three dimensions improve simultaneously.

The monthly table shows why. January was strong across the board. Silver posted 14.59%, pulling the portfolio up. February saw equities struggle while Gold was flat and Silver positive, providing a cushion. March was driven by Gold (3.11%) and Russell 2000, while S&P 500 and Nasdaq-100 were flat or negative. The metals signals and equity index signals aren't synchronizing month-to-month, which is what allows the combination to smooth the overall equity curve.

Silver's standalone volatility is high (26.16%) but its 8% weight bounds its portfolio contribution. At a larger weight it would dominate the portfolio rather than diversify it. The weighting here reflects a judgment about volatility contribution, not a prediction about which signal will outperform.

One caveat worth stating plainly: Silver's return is concentrated in time. Most of it came in January, and Silver was roughly flat across the following four months. With five months of data, these portfolio figures are directional, not a stable estimate. The structural point holds regardless of the single-month magnitude. Metals order flow decorrelates from equity index order flow, which is what lets the combination improve the risk-adjusted profile. The size of the lift in this particular window should be read with that concentration in mind.

Portfolio C: Metals signals

Composition: Gold (60%), Silver (40%)

Table 5: Portfolio C performance summary

Table 6: Portfolio C monthly returns (%)

The 60/40 Gold/Silver portfolio posts a Sharpe of 1.742, above either component. Return at 8.25% sits between Gold (4.05%) and Silver (14.55%), as expected from the weighting. Max drawdown at -4.54% is well below Silver's -7.03% standalone. The combination captures most of Silver's return upside while keeping the drawdown profile closer to Gold's.

April tells the story clearly: Gold was up 1.14% while Silver was down -2.23%. The positive offset from Gold held the portfolio's loss that month to just -0.15%. May was the exception. Both metals declined together and the blend tracked Gold's loss, so the 60/40 position offered no diversification benefit that month. Even within a single asset class, the two metals signals are responding to somewhat different order book dynamics. Gold and Silver don't trade identically at the microstructure level, even in months when they move in the same direction at the price level.

Three observations

Combining signals always helps something. Across all three configurations, the portfolio improved at least one of Sharpe, return, or max drawdown relative to every individual constituent. There was no case where adding a second signal made all three metrics worse.

Cross-asset combination does more. The step from Portfolio A (equity-only, Sharpe 0.969) to Portfolio B (equity + metals, Sharpe 1.654) is larger than anything achievable by rearranging equity index weights. The equity index signals share underlying market dynamics. They're all responding to the same broad order flow. Adding a structurally different asset class introduces signals that aren't driven by the same microstructure regime.

Weighting matters, but direction matters more. These portfolios use simple fixed weights chosen for volatility balance, not optimized in-sample. The point isn't the specific numbers. It's that even unsophisticated combinations produce a measurable risk-adjusted improvement when the signals carry low inter-asset correlation.

Where to go next

Individual signal benchmarks for S&P 500, Nasdaq-100, Russell 2000, Gold, and Silver are published at quantumsignals.ai/benchmarks. Each report includes monthly Sharpe, return, win rate, max drawdown, and Calmar.

If you're currently running one signal you might want to explore what a multi-signal portfolio would look like for your strategy.

All results are out-of-sample backtests using pre-trained signals under realistic transaction cost assumptions. Past performance does not guarantee future results. Full methodology available on request.