The Map Is Not the Trend: Why Chasing CTA Winners Across Asset Classes Fails

Here is a question that sounds reasonable on the surface. If commodities have been driving CTA performance since 2020 while fixed income has quietly faded, why not just overweight what is working? Quantica Capital tested exactly this logic across 26 years of data, through May 2026. The answer is not what most people expect.

The Shift Is Real. And It Is Dramatic.

Trend-following strategies, tracked through the SG Trend Index, delivered annualized returns of roughly 8.1% between January 2020 and May 2026, with a slightly negative correlation to global equity markets. Solid numbers. But what changed underneath those numbers is the real story.

Quantica's proprietary model shows that energy, metals, and agricultural markets may have accounted for the majority of CTA excess returns since 2020. That flips the 2010s script entirely, when fixed income led and commodities actually detracted. Quantica notes the current environment "increasingly resembles the commodity-driven regime of 2000-2009, when commodities were highly profitable for trend-followers while fixed income added little."

This is not a blip. It is a structural regime change. And it makes the tactical temptation obvious: lean into what is working.

A Clean Test of Performance Chasing

To see whether rotating toward stronger sectors actually pays off, Quantica built three portfolios. The Baseline Portfolio spreads risk equally across six asset groups: equities, fixed income, currencies, energy, metals, and agriculturals. The Trend-on-Trend Portfolio overweights sectors with stronger recent trend performance. The Contrarian Trend Portfolio does the opposite.

The underlying trend signal is identical across all three. Only the sector weights change. Rankings update weekly, with the top-ranked sector getting twice the baseline weight and the bottom-ranked sector getting zero.

By May 2026, the Trend-on-Trend Portfolio is "heavily overweight metals and agricultural commodities, mechanically responding to the strong trend-following profits generated by precious metals and by agricultural markets such as cocoa over recent years." Fixed income and currencies are underweighted. Equities have been only intermittently overweight, despite the long equity bull market, because "allocations are driven by realized trend-following profitability rather than by absolute asset returns."

The rotation looks sensible. The returns do not follow.

You Can See the Pattern. You Cannot Trade It.

There is something real in the data. Over lookback horizons of up to about one year, sectors with stronger recent trend performance do tend to keep outperforming. Correlations are positive and stable at roughly 0.05 across short lookback windows. That is modest, but it is there.

Beyond a year, the pattern flips. Sectors with multi-year runs of strong trend profitability tend to underperform afterward. So the short-term signal persists and the long-term signal mean-reverts.

The catch is that neither effect is strong enough to move the needle in a portfolio. Across every lookback horizon tested, "the median rolling three-year Sharpe ratio difference remains near zero across all specifications, while the distribution of realized outcomes spans both positive and negative values." The Trend-on-Trend approach "fails to generate a consistent improvement in risk-adjusted performance at any lookback horizon."

Quantica puts it plainly: "statistical predictability alone is insufficient to create an investable allocation signal."

The Signal Is Already Baked In

Here is the deeper reason why performance chasing fails. Short-horizon Trend-on-Trend allocations are not independent of the underlying trend signal. They are a nonlinear version of it.

When the profitability estimation window is similar in length to the horizon embedded in the underlying trend model, recent profit history largely reflects current trend strength. The overlay then amplifies strong signals and suppresses weak ones. No new information enters the process.

Quantica states it directly: "The Trend-on-Trend overlay, rather than introducing an independent source of information, largely reinforces exposures already embedded in the underlying trend signal and reduces exposure to weaker ones."

You are not adding a new lens. You are just turning up the contrast on the one you already have.

The Hidden Cost: Diversification

There is a second problem, and it is structural. The sectors with the best recent trend performance also tend to have higher pairwise correlations among their constituent markets. Chasing profitability means buying correlated risk.

Quantica finds that "the strongest trend-following environments tend to be characterized by a greater alignment of trend opportunities across markets," and that "reallocating capital toward these sectors therefore increases concentration and reduces diversification, offsetting much of the benefit from stronger trend regimes."

Strong trends do not come from many independent markets moving together by coincidence. They come from related markets moving in lockstep. That is highly profitable while it lasts. It is also fragile when it breaks.

What Actually Works

The future location of trend opportunities cannot be reliably forecast from past performance, beyond what the trend signal itself already tells you. Quantica's conclusion is unambiguous: "broad diversification across markets remains more robust than attempting to time the distribution of trend opportunities across sectors."

5 Key Takeaways for Advisors and Investors

1. Regime shifts in trend-following are real but not actionable. Commodities dominating CTA returns since 2020 does not mean overweighting them going forward improves outcomes. Leadership rotation is persistent enough to observe but not reliable enough to trade.
2. Short-horizon persistence exists but cannot be monetized. Trend-following profitability shows statistically significant persistence over horizons up to one year. Across all lookback windows tested, Sharpe ratio improvements versus the Baseline Portfolio remain near zero.
3. Performance chasing amplifies existing positions, not alpha. A Trend-on-Trend overlay largely acts as a nonlinear amplification of trends already in the portfolio. It adds concentration, not diversification or independent signal.
4. The strongest trend environments carry hidden correlation risk. High trend-following profitability is structurally linked to higher within-sector market correlation. Strong periods can reverse sharply when synchronized trends unwind at the same time.
5. Broad diversification remains the structurally sound approach. Quantica's evidence across 26 years and six asset classes confirms that equal-risk diversification across sectors outperforms performance-chased concentration on a consistent, risk-adjusted basis.

Footnotes:

1 Quantica Capital AG. "Chasing Trends or Chasing Performance? On the Challenge of Timing Trend Opportunities Across Asset Classes." Quantica Quarterly Insights, no. 26, June 2026