Normal Backwardation: Keynes’s Theory of Why Futures Prices Are Systematically Too Low

Most futures pricing discussions start with no-arbitrage cost-of-carry mechanics futures price equals spot price grown forward by financing and storage costs, minus convenience yield. That framework is clean, mathematically rigorous, and useful. But it’s also somewhat silent on a question that sits right at the intersection of derivatives theory and alternative investments: even in the absence of arbitrage opportunities, is there a systematic relationship between futures prices and expected future spot prices? And if there is, where does that relationship come from?

This is the question that normal backwardation theory answers and the answer, developed by John Maynard Keynes in the 1930s and later refined by John Hicks, remains one of the more intellectually interesting ideas in the commodities and futures curriculum, even if its empirical support turns out to be more complicated than the original theory suggested.

The Distinction That Matters: Backwardation vs. Normal Backwardation

Before going further, the terminology distinction deserves explicit attention because it’s genuinely easy to conflate these two related but different concepts.

Backwardation is a description of the current shape of the futures curve when the spot price is higher than the futures price for delivery in the future, the market is in backwardation. This is an observable, present-tense fact about the relationship between today’s spot price and today’s futures prices for future delivery dates.

Normal backwardation is something different. It’s a theory about the relationship between the futures price and the expected future spot price not today’s spot price, but where the spot price is expected to be on the delivery date. Specifically, normal backwardation says that futures prices are systematically set below the market’s expectation of where the spot price will be at delivery, with the gap between the two representing a risk premium earned by speculators who take long futures positions.

A market can be in backwardation (spot above futures) without being in normal backwardation (futures below expected future spot), and vice versa. The two concepts operate on different time dimensions — one compares spot to futures across delivery dates today, the other compares futures to expected spot at the same future point in time.

Keynes’s Original Argument

Keynes developed the normal backwardation theory in his Treatise on Money (1930), observing that commodity markets are populated by two fundamentally different types of participants with very different objectives.

On one side are commodity producers farmers, miners, oil companies, cattle ranchers. These participants face a natural short exposure: they own the commodity or are in the process of producing it, and they face the risk that prices will fall by the time they’re ready to sell. A wheat farmer who plants in spring and harvests in autumn faces months of price uncertainty. A copper miner committing capital to a multi-year mining project faces years of price exposure. These producers are natural hedgers who want to reduce that price risk by locking in a selling price through the futures market which means they want to take short futures positions.

On the other side are speculators participants with no natural commodity exposure who take positions purely based on their price views and their willingness to bear risk in exchange for expected return.

Keynes’s key insight was about the balance of power between these two groups. Producers, he argued, have an urgent, genuine need to hedge their price risk; it’s not optional, it’s a business necessity. Speculators, by contrast, need to be enticed into taking the other side of those hedging trades. They’re providing a service absorbing the price risk the producers desperately want to offload and they should expect to be compensated for providing that service.

The mechanism of compensation, in Keynes’s framework, is the futures price being set below the expected future spot price. If a speculator buys a futures contract today at a price F, and the expected future spot price is E(S), then if F < E(S), the speculator expects to profit by the amount E(S) − F when the contract expires and futures converge to spot. That expected profit is the risk premium the compensation the market pays speculators for absorbing producer hedging pressure.

The word “normal” in normal backwardation reflects Keynes’s view that this was the natural, expected state of futures markets whenever the hedging pressure from producers dominated the market which, in most commodity markets, he believed was the typical case.

The Risk Premium Framing: Connecting to Modern Asset Pricing

The normal backwardation theory has an important connection to broader asset pricing ideas that the CFA curriculum builds on, particularly at Level III.

In modern asset pricing frameworks, any risky asset is expected to earn a positive risk premium, a return above the risk-free rate in equilibrium, to compensate investors for bearing systematic risk. For futures markets, the analogous argument is: long futures positions bear price risk (the commodity price could fall), and if that risk is systematic correlated with broad economic conditions then speculators who bear it should earn a positive expected return.

The normal backwardation theory essentially predicts a positive expected return to long futures positions in commodity markets, driven by the systematic transfer of risk from natural short hedgers (producers) to natural long speculators. This connects directly to the alternative investments curriculum’s treatment of commodity futures total return, which decomposes into three components: the roll yield, the collateral yield, and the spot price change. Normal backwardation theory predicts that the roll yield component should be positive for long futures investors when hedging pressure from producers dominates precisely because futures prices are set below expected future spot prices, so rolling long positions generates a systematic expected gain.

The Contango Counterargument: What Happens When Consumers Hedge

Keynes’s theory had a natural counterpart that he acknowledged but somewhat underweighted: the possibility that consumers, not producers, might be the dominant hedgers in certain markets.

An airline needing to lock in jet fuel costs, a food manufacturer needing to secure wheat supply, a utility needing predictable natural gas costs these are all commodity consumers who face the opposite risk from producers. They’re naturally long the commodity (they need to buy it in the future) and want to hedge by taking long futures positions, locking in their future purchase price. If consumer hedging pressure dominates a market, the dynamic reverses: now consumers desperately want long futures positions, and speculators providing short positions need to be compensated. The compensation mechanism would be futures prices set above expected future spot prices normal contango, the mirror image of normal backwardation.

This consumer-hedging scenario creates what the broader theory calls normal contango (not to be confused with simple contango, which again is a curve shape concept). In practice, the balance between producer and consumer hedging pressure determines which direction the systematic bias runs and that balance can shift over time and across commodities.

The Empirical Reality: More Complicated Than the Theory

Normal backwardation theory makes a testable prediction: long futures positions should earn positive expected returns over time, because futures prices are systematically set below expected future spot prices. Testing this prediction empirically has generated decades of inconclusive and sometimes contradictory evidence.

Some evidence does support the existence of a positive commodity futures risk premium over long horizons. Studies by Gary Gorton and K. Geert Rouwenhorst in the mid-2000s found that a diversified portfolio of long commodity futures positions delivered equity-like returns over multi-decade periods, with low correlation to both equities and bonds broadly consistent with the normal backwardation story. This research was enormously influential in driving institutional interest in commodities as an asset class in the 2000s.

Other evidence is less supportive. The commodity futures risk premium, where it exists, varies enormously across individual commodities and time periods. Some commodities show persistent positive roll yield consistent with normal backwardation; others show persistent negative roll yield consistent with contango. The same commodity can shift between these regimes as the balance of hedging pressure changes. And the 2000s commodity boom followed by the 2010s performance drought raised questions about whether the historically estimated premium was durable or an artifact of a specific historical period.

The most nuanced current view in the academic literature and in the CFA curriculum is that normal backwardation describes a plausible mechanism for a commodity futures risk premium, but that this premium isn’t universal, isn’t constant, and isn’t large enough or reliable enough to be the dominant driver of commodity futures returns in any given year or even decade.

What This Means for the Roll Yield in Practice

The connection between normal backwardation and roll yield is the most practically relevant dimension for the CFA alternatives curriculum, and it’s worth spelling out precisely.

Roll yield, as covered elsewhere in the curriculum, is the return contribution from rolling expiring futures contracts into longer-dated ones. When a market is in backwardation, the longer-dated contract costs less than the expiring near-month contract, so rolling generates a positive return: sell the near-month at a higher price, buy the longer-dated at a lower price. When a market is in contango, the opposite is true: rolling generates a negative return.

Normal backwardation theory provides the theoretical justification for why commodity markets might persistently tend toward backwardation and therefore toward positive roll yields for long futures investors: it’s because producers’ hedging pressure systematically depresses futures prices below expected future spot, which is precisely the condition that creates backwardation (spot above futures) when the expected future spot is roughly equal to today’s spot.

But the empirical observation that individual commodities and even broad commodity indices have spent significant periods in contango, generating negative roll yields makes clear that this tendency is not universal or guaranteed. The theory describes a mechanism; actual market conditions depend on the balance of forces in any specific commodity market at any specific time.

Implications for Portfolio Construction

For institutional investors allocating to commodity futures the context in which this theory most directly matters for CFA candidates normal backwardation theory has several portfolio implications.

If the theory holds, the expected return to a long commodity futures position comes partly from the risk premium (futures set below expected future spot) and partly from spot price movements. A long commodity futures position in a normally backwardated market generates a positive expected return even if spot prices don’t change, simply from the convergence of futures to spot as delivery approaches. This positive roll return is separate from any speculation on spot price direction.

This also implies that the portfolio benefit of commodity futures isn’t simply exposure to commodity price movements; it includes this potentially positive carry component that’s driven by hedging market structure rather than spot price forecasts. Whether this carry is reliably positive depends on whether normal backwardation actually prevails in the markets being invested in, which requires monitoring the actual shape and evolution of the futures curve rather than assuming the theory holds automatically.

Exam Perspective: What to Lock In

For CFA Alternative Investments, several points are worth anchoring. Normal backwardation is a theory about the relationship between futures prices and expected future spot prices. Futures are systematically set below expected future spot, generating a positive expected return for long futures investors. This is distinct from backwardation, which describes the current shape of the futures curve (spot above near-term futures prices). The theoretical mechanism is producer hedging pressure: producers need to offload price risk by taking short futures positions, and speculators absorbing that risk earn a premium in the form of futures prices below expected future spot. When consumer hedging pressure dominates instead, normal contango can result (futures above expected future spot). The connection to roll yield is direct: persistent normal backwardation produces persistent positive roll yield for long futures investors. Empirical evidence for the theory is supportive but not definitive. The commodity futures risk premium exists in some markets and periods but is far from universal or stable. And for portfolio construction, commodity futures’ return potential includes both spot price exposure and the roll return, with the sign and magnitude of the roll return depending on actual market structure rather than theoretical prediction.

Final Thoughts

Normal backwardation theory is one of those ideas that’s most valuable not as a precise forecasting tool but as a conceptual framework, a way of understanding why commodity futures markets might systematically price contracts at levels that reward patient, long-term speculative capital. Keynes was describing something genuinely real about the structure of commodity markets: producers need to hedge, hedging requires counterparties, and counterparties don’t provide risk absorption services for free.

Where the theory gets complicated is in the assumption that producer hedging pressure always dominates and in the observation that financial markets are far more complex than the simple producer-versus-speculator dichotomy Keynes was working with in the 1930s. Modern commodity futures markets involve commodity index funds, ETFs, swap dealers, sovereign wealth funds, and algorithmic traders, all of whose positions affect the supply and demand for futures contracts in ways that don’t map cleanly onto Keynes’s original framework. The theory remains a useful lens; it just isn’t, on its own, a reliable predictor of roll returns in any specific market at any specific time.