by AdvisorAnalyst.com Editorial Team
Graham Capital Management's May 2026 research note, "An Expanding Market Universe,"1 returns to a question every systematic allocator eventually confronts: how many markets are enough? Authored by Nastja Bethke, Senior Research Scientist, and Thomas Feng, CIO of Quantitative Strategies, the paper updates the firm's 2017 study by Tricker and Mitchell, which found that trend following performance plateaus once a portfolio reaches roughly sixty markets. The new note asks whether that ceiling still holds, and if so, which markets deserve serious consideration for inclusion today.
A Simple Test, A Familiar Constraint
The authors build a generic (20,120) moving average trend following model, designed to loosely mirror the SG Trend Indicator, and add markets in strict order of liquidity, defined as the product of dollar big point value, price change volatility, and volume. Running the backtest from 2000 using gross daily returns across a universe of 140 markets, they trace information ratio (IR) against portfolio size. The result echoes the earlier study: IR climbs sharply at first, flattens between twenty and fifty markets, rises again toward a universe of seventy to eighty markets, and then levels off. Beyond that point, additional markets add little because, as the authors put it, "more is not necessarily better."
Liquidity Thins Out Fast
Bethke and Feng flag two liquidity thresholds, 50% and 5% of the average liquidity found in the top fifty most liquid macro markets. Those thresholds arrive at approximately sixty and ninety five markets respectively, meaning that liquidity deteriorates quickly once a portfolio moves past its first fifty holdings. Price history compounds the problem. Table 1 identifies roughly a dozen markets, mostly commodities such as INE Crude Oil, Fuel Oil, TSR 20 Rubber, and Bonded Copper, whose histories only begin in 2018 or later. FTSE Taiwan is the lone equity index in this group. Five years ago, these markets simply would not have had enough backtested history to support inclusion in a systematic strategy.
Commodities Diversify, Financials Do Not
The most consequential finding concerns where new diversification actually comes from. When the authors isolate new global financial markets (equities, FX, fixed income), IR stays essentially flat as those markets are added. When they isolate new commodity markets instead, spanning China domestic, China international, and non-China contracts, IR rises meaningfully. The explanation offered is structural: commodity returns tend to be more idiosyncratic, while global financial futures and FX markets are already highly correlated with the most liquid instruments already inside the portfolio, so adding more of them contributes little new signal.
A Necessary Reality Check
Because the backtest imposes no market size bounds, the commodity driven IR gain in Figure 3 is, in the authors' words, "overly optimistic." Real world trading at scale (potentially several hundred million dollars or more per position) would require tighter contract bounds on illiquid names. To correct for this, the authors scale down each market's P&L by a factor tied to its liquidity. The muted result in Figure 4 still shows commodities improving IR, just less dramatically than the unconstrained version suggests, giving a more honest picture of what expansion is actually worth.
Five Key Takeaways for Advisors and Investors
- Adding markets to a trend following portfolio delivers diminishing, then negligible, returns beyond roughly seventy to eighty markets, regardless of how large the nominal universe grows.
- Liquidity, not just correlation, should govern which new markets earn inclusion; liquidity drops sharply after the first fifty to sixty names.
- Newer markets such as crypto futures and select industrial commodities may lack sufficient price history to validate model robustness, a limitation that fades over time but should not be ignored today.
- Commodity markets currently offer more genuine diversification potential than additional global financial or FX markets, which largely duplicate existing risk exposure.
- Backtested gains from expanding into illiquid markets should be discounted for real world trading constraints; unadjusted IR figures likely overstate the practical benefit.
Footnote:
1 Bethke, Nastja, and Thomas Feng. "An Expanding Market Universe: Inclusion Criteria for Diversified Systematic Portfolios as Illustrated on Generic Trend-Following." Graham Capital Management, May 2026.
