Managed Futures ETF: How It Works, How Much to Own, and When It Fails

A managed futures ETF trades futures on stocks, bonds, currencies and commodities, long or short, by rule. Its cash sits in Treasury bills. The futures own whatever is trending, in either direction.

That is the entire product. No forecast, no analyst, no view on the economy. A rule reads price and sizes a position.

How does a managed futures ETF work?

Three pieces. Collateral, signals, sizing.

The collateral is short-term government paper. Paisley Nardini, then of Simplify, put it plainly: "the underlying collateral for managed futures is often treasury bills," so "T-bill collateral returns and yield is your starting point." The cash earns the short rate before the strategy does anything.

The signals read price across dozens of markets. Up over the lookback window means long. Down means short. Most funds are no more clever than that.

Sizing is where the leverage lives. A futures contract needs margin, not full payment. Nardini was direct: "the underlying notional exposure that we're achieving in this strategy is far greater""than the margin that we have to post as collateral." So the fund holds exposure to forty markets while its assets sit in bills.

Why does a managed futures ETF earn anything?

Most explanations start with the trend. Wrong end. Buying what already went up sounds like nonsense.

Jerry Prior of Mount Lucas answers from the other side. Producers and consumers of real things need price certainty. A farmer, a miner, an airline. They "face real risks to their margins" if the price moves before delivery. So they sell that risk, and somebody buys it. Prior's firm sits on the "other side of that risk transfer," and the practical method for "accepting price risk through time is through trend following."

Read that way the return is a premium, not a prediction. The hedger pays for certainty. You get paid for uncertainty.

What did a managed futures ETF hold in 2022?

2022 is the cleanest worked example available. Every major market trended, and they trended at once.

Start with the facts. The US 10-year Treasury yield opened 2022 at 1.51% and closed at 3.88%. WTI crude began the year near $76 a barrel and traded above $120 in early June. The dollar index started near 95.7 and peaked around 114 on 27 September 2022. The S&P 500 fell 19.4% over the calendar year and drew down 25.4% between 3 January and 12 October.

Now trace the rules. A rising yield means a falling bond price, so the system goes short bond futures and stays short. Crude is rising, so it goes long crude through the spring. The dollar is rising, so it goes long dollar. Equities roll over in January, so equity positions flip short.

Same rule, four unrelated contracts. That is where the diversification actually comes from.

What did a managed futures ETF hold in 2008?

2008 is the second worked example, and it teaches more. The same rules held close to the opposite book.

The facts first. The S&P 500 fell 38.5% in calendar 2008. It peaked at 1,565.15 on 9 October 2007 and bottomed at 676.53 on 9 March 2009. The US 10-year Treasury yield fell from about 4.0% at the end of 2007 to about 2.2% at the end of 2008. WTI crude settled above $145 a barrel in July 2008 and traded below $35 by late December. The dollar index bottomed near 71 in March 2008 and traded near 88 by November.

Now trace the rules. Equities are falling, so the system is short equity futures. Yields are falling, so bond prices are rising, so it goes long bond futures. Crude rises into July, so it is long. Crude collapses after July, so it flips short. The dollar turns up in the spring, so it goes long dollar.

Hold the two years side by side. In 2022 the rule was short bonds. In 2008 the rule was long bonds. Same code, opposite position.

That is the part advisors miss. The strategy does not hedge equities. It follows price. Sometimes those are the same trade. Often they are not.

What do the returns look like month to month?

Ugly, most of the time. Nobody says this clearly enough.

Jerry Parker of Chesapeake Capital has run this forty years. Across many markets, "five to 10% of the trades will be a big mega outlier." The rest is "taking small losses all the time" and letting the winners run, because he runs "predetermined stop losses" on every trade.

So it is wrong most of the time on purpose. A long line of cheap errors, paid for by a few very large wins.

Check it monthly and you will mostly see small red numbers. That is the machine running correctly. It is also why people sell in month eight.

How much of a portfolio should a managed futures ETF be?

Less than the statistics suggest. The people who sell it say so out loud.

Start with the industry's own number. Parker describes the standard pitch as "a five or 10% allocation to your stock and bond portfolio" for crisis alpha. That is the sleeve most models use.

Andrew Beer of Dynamic Beta investments argues for less. Over a long sample, he says, an investor would "have been unhappy with it for 20 years and deliriously happy with it for four years." His conclusion: "I think the appropriate allocation for people is much lower than that."

The reason is behavioral, not statistical. Clients read line items. A sleeve that bleeds slowly for years gets sold before the four good years arrive. Beer calls it "a sizing issue."

One more rule from him. It has to be "a strategic allocation because you can't get the timing right." Nobody buys the insurance the month before the fire. So size it to what a client will hold through a quiet decade.

Why is the correlation to stocks low, and when does it stop being low?

Because the positions are not stock positions. They are direction bets across forty markets, and most of those markets are not equities.

Nardini gave the textbook version. Combine assets with "a correlation that's less than one" and portfolio volatility falls. Nothing controversial there.

The mechanism matters more than the number. Correlation is low because the signal is free to be short. A long-only diversifier still owns something that is falling. A trend system can own the fall itself.

Now the limit. That correlation is an average, and averages move. At a turn the system still holds the old trend. Equities bottom, the book is still short, and it loses money next to a recovering market.

So it is not a hedge. A hedge pays at the moment of stress. This seeks to pay across a long move, and it can be wrong at both ends of one.

How do managed futures ETFs differ from each other?

Less than the brochures suggest. Prior on his competitors: "Everybody's doing a trend following in some sort of way."

The first real separation is risk sizing. Most funds use volatility targeting, which adjusts position size to hold portfolio volatility roughly constant. It seeks to shallow out drawdowns. It can also trim participation in the violent moves the strategy exists to catch. Prior argued against it, then allowed that "it's not wrong."

The second is the market set and its weights. Prior moved his index to sector weights in 2005, giving "42 and a half percent to bonds, 32 and a half percent to currencies" after a commodity-heavy start. A bond-weighted book and a commodity-weighted book behave nothing alike.

Two funds, identical signals, different vol targets and different weights. Different products. Compare those lines first.

Managed futures ETF or trend following ETF?

Mostly the same thing in two labels. Managed futures is the older industry term for any rules-based futures program, and trend following is the signal nearly all of them run. Our piece on trend following ETFs walks through the signal itself.

The wrapper is what changed. Beer was blunt on the old structures: they "charge hedge fund fees. It's, you know, one in 20." One percent of assets and twenty percent of profits. Most of these funds now issue a 1099 rather than a K-1, but check the prospectus, because it differs by fund.

When does a managed futures ETF fail?

In chop. Prior volunteered it: there are "periods of time where trend following doesn't work." A market that reverses every few weeks hands the system a string of small losses and no outlier to pay for them.

It fails in fast reversals too. A V-shaped bottom flips the signal after the recovery already happened. March into April 2020 is the reference case. Short into the low, slow to turn.

It fails in a long quiet bull market. Prior again: "we work best when markets are more volatile." Calm grinding upside is the worst regime for it, and that describes a lot of years.

The reason to hold it anyway is narrow. Meb Faber of Cambria calls trend following "the premier diversifier to a traditional portfolio, particularly during the bad times." The claim is not that it beats stocks. It is that it can make money in long grinding declines.

Definitions

• Managed futures ETF. An exchange-traded fund that takes long and short positions in futures across asset classes using systematic rules.
• Trend following. A rule that goes long markets rising over a lookback window and short markets falling.
• Notional exposure. The face value the futures control, far larger than the margin posted.
• Volatility targeting. Adjusting position size so portfolio volatility stays near a fixed level.
• Collateral yield. Interest earned on the Treasury bills backing the futures positions.
• Crisis alpha. The industry term for gains earned while equities fall. It describes a long decline, not a single shock.
• Sector weighting. The share of risk a program assigns to each asset class rather than to each contract.

What to check before you buy one

• The vol target. The biggest single difference between funds running the same signals.
• The sector weights. A bond-weighted book behaves nothing like a commodity-weighted one.
• The collateral yield. When short rates are high, a real slice of the return is bill interest.
• The sleeve size. Five to ten percent is the industry default. Pick the number a client holds through a flat decade.
• The tax form. 1099 or K-1 changes the client conversation in April.

THOR runs adaptive, rules-based strategies that respond to price rather than forecast it. That is a different signal family from futures trend following, and it reads price rather than earnings or economics.

This is educational content and not investment advice. It is not a recommendation regarding any security. Investing involves risk, including possible loss of principal. Past performance is not indicative of future results.