Managed Futures and Trend Following: What the Evidence Says

Managed futures are investment programs run by professional managers, usually registered as commodity trading advisors (CTAs), that trade futures across many markets, most often using trend following: going long markets that have been rising and short markets that have been falling. The best-documented evidence shows trend following has delivered positive average returns over long periods and has often done well in prolonged stock market declines, such as 2008 and 2022, but it has also gone through long weak stretches, including much of the 2010s.

In short: A trend follower holds positions in dozens of futures markets (stock indexes, bonds, currencies, commodities), sized so each carries similar risk, and flips from long to short as trends change. Academic studies using data back to 1880 find trend signals have been profitable in every decade on a simulated basis, before and after estimated costs. Real fund indexes confirm strong years in 2008 and 2022, but they also show many flat or losing years, wide differences between managers, and returns well below the simulations once real fees and costs are counted. It is a strategy with a reasonable research case, not a reliable source of profits in any given year.

Who runs managed futures

In the US, a commodity trading advisor (CTA) advises on or manages futures trading for others, and a commodity pool operator (CPO) runs a pooled fund that trades futures. Both are generally required to register with the CFTC and be members of the National Futures Association (NFA), subject to exemptions. In Canada, advisers who manage futures portfolios are registered under provincial securities or commodity futures laws. The regulatory framework is covered in how futures markets are regulated.

Managed futures are available through private funds, separately managed accounts, and, in recent years, mutual funds and ETFs that follow trend-following strategies.

How trend following works mechanically

The signal

The simplest trend rule is time-series momentum: look at a market’s own past return over a set window (for example, 12 months). If it is positive, be long; if it is negative, be short. Real programs blend several windows (1, 3, and 12 months is a common research setup) and many other filters, but the core idea is the same.

This is different from cross-sectional momentum, which ranks assets against each other. Time-series momentum asks only whether each market is trending, so a trend follower can be short everything in a broad decline.

Diversification

Trend followers trade many unrelated markets so that a few strong trends can pay for the many small losses from false signals. The academic study discussed below used 67 markets: 29 commodities, 11 equity indices, 15 bond markets, and 12 currency pairs.

Volatility-based position sizing

Because markets differ wildly in how much they move, positions are usually sized to carry similar risk, not similar dollar value.

Worked example: sizing by volatility

Real contract specs: COMEX gold futures are 100 troy ounces, and CME Micro Gold futures are 10 troy ounces. Prices and volatility are illustrative.

A program wants each market to contribute about $10,000 of annual volatility to the portfolio.

Gold price: $2,000 per ounce (illustrative)
Standard gold contract notional: 100 oz x $2,000 = $200,000
Gold’s annual volatility: 15% (illustrative)
Annual dollar volatility per standard contract: 15% x $200,000 = $30,000
Contracts needed: $10,000 / $30,000 = 0.33 contracts

You cannot trade a third of a contract, so a smaller program might use Micro Gold:

Micro notional: 10 oz x $2,000 = $20,000
Dollar volatility per micro: 15% x $20,000 = $3,000
Micros needed: $10,000 / $3,000 = 3.33, rounded to 3 Micro Gold contracts

If gold’s volatility doubles to 30%, the same rule cuts the position roughly in half. This automatic de-risking is a big part of how trend followers manage risk, and also why they can be slow to re-enter after a shock. Smaller contracts and their risks are covered in micro futures explained.

Rolling

Futures expire, so a trend follower continuously rolls positions into later months. Returns therefore include roll yield, which depends on whether a market is in contango or backwardation. See rolling futures contracts.

What the evidence says

The long-run academic evidence

The most cited long study is Hurst, Ooi and Pedersen, “A Century of Evidence on Trend-Following Investing,” Journal of Portfolio Management, 2017. The authors (who worked at AQR Capital Management, a firm that runs trend-following funds) built a time-series momentum strategy on 67 markets from 1880 to 2016, scaled to a 10% annual volatility target.

Their headline findings:

Positive average returns in every decade since 1880.
Good performance in 8 of the 10 largest drawdowns for a 60/40 US stock and bond portfolio.
Low correlation to stocks and bonds over the full period.

The numbers depend heavily on costs and fees. From the paper’s Exhibit 1, annualised returns above the risk-free rate were:

Period	Before costs and fees	After estimated costs	After costs and a 2-and-20 fee
1880-2016	18.0%	11.0%	7.3%
2000-2009	11.6%	9.9%	6.3%
2010-2016	7.6%	6.2%	3.3%

Three honest caveats belong next to these figures. First, they are simulations, not a fund’s actual returns. Second, the authors’ estimated transaction costs are, in their words, made “with a significant amount of uncertainty.” Third, the most recent period shown, 2010-2016, was the weakest.

The paper also notes that the strategy “may not perform as consistently in bear markets that occur very rapidly, such as the 1987 stock market crash,” because it cannot build positions fast enough, and that its own worst drawdowns reached about 25% and lasted extended periods.

Earlier work by Moskowitz, Ooi and Pedersen (“Time Series Momentum,” Journal of Financial Economics, 2012) found the effect across 58 liquid futures and forward markets from 1985 onwards, which the 2017 paper extended back in time.

Real index data: 2008 and 2022

Real fund indexes are the better test, because they include actual fees, costs, and manager decisions.

Year	Barclay CTA Index	S&P 500 total return
2008	+14.09%	about -37%
2022	+7.13%	about -18%

The Barclay CTA Index, published by BarclayHedge, is an equally weighted average of several hundred CTA programs with at least four years of history. The SG CTA Index, which tracks a smaller group of the largest managers open to new investment, rose about 20.1% in 2022, which Societe Generale reported as the index’s best year since its calculation began in 2000.

Both years featured long, persistent trends. In 2008, stock indexes fell for months while government bond prices rose. In 2022, stocks and bonds fell together while many commodities and the US dollar rose, which is a hard environment for traditional portfolios and a favourable one for trend signals.

Real index data: the lean years

The same Barclay CTA Index shows how uneven the strategy is:

Year	Barclay CTA Index
2011	-3.09%
2012	-1.70%
2013	-1.42%
2014	+7.61%
2015	-1.50%
2016	-1.23%
2017	+0.70%
2018	-3.17%

That is six losing years out of eight, during a period when stocks performed well. Investors who bought managed futures after 2008 waited a long time for the next strong year. The gap between the Barclay index (+7.13%) and the SG CTA Index (about +20%) in 2022 also shows that “managed futures” is not one thing: results vary enormously by manager, speed of signal, and market mix.

What can go wrong

Long flat periods. Choppy, trendless markets generate repeated small losses from false signals.
Sharp reversals. A trend can turn faster than the signal reacts, as in a crash like 1987 or a V-shaped rebound.
Fees. The paper’s own numbers show a 2-and-20 fee structure removing several percentage points a year. Many products charge less today, but costs still matter.
Backtest optimism. Simulated strategies are chosen with hindsight. Real programs usually underperform their backtests.
Crowding and correlation. The paper finds the strategy did best when markets were less correlated, and that correlations rose meaningfully from late 2008 to mid-2014.
Leverage. Futures positions are leveraged, and a fund’s volatility target can be exceeded in sudden moves.
Not a hedge you can count on. Doing well in 8 of 10 large drawdowns is not the same as doing well in the next one.

The short version

Managed futures are futures programs run by CTAs, most of which follow trends by going long rising markets and short falling ones across dozens of futures, with positions sized by volatility. Long-run simulations back to 1880 show positive returns in every decade, though much smaller after costs and fees. Real indexes confirm strong years in 2008 and 2022, when stocks fell, but also a long run of flat or losing years in the 2010s and big differences between managers. The evidence supports trend following as a diversifier with real drawbacks, not as a dependable profit engine.