Contango vs Backwardation: Why Commodity Returns Can Defy Spot Prices

What is the difference between contango and backwardation?

Contango means a futures contract trades above the spot price; backwardation means it trades below it. Traders also use these terms to describe an upward- or downward-sloping sequence of delivery-month prices. The important question is not whether the chart points up or down, but which prices, dates and investment exposure you are comparing. A futures curve is not a guaranteed forecast, and a favourable-looking curve is not a guaranteed return.

That distinction matters when an investor buys an oil fund, a trader rolls a commodity position, or a manufacturer decides whether to hold inventory. The cash price of the commodity can rise while a futures-based investment disappoints. Conversely, a backwardated market can reward a long position through convergence without preventing a loss when the whole price structure falls. This guide separates those mechanisms rather than treating a curve label as a trading signal.

Our original contribution is a controlled set of price-path comparisons: we hold the entry price and spot starting point constant, change the eventual settlement, and show exactly when convergence helps and when the outright move overwhelms it. Every price in those teaching examples is hypothetical. The historical section uses dated US Energy Information Administration evidence, not a live quote or an assumed current market level.

CME Group's educational explanation provides the starting definitions and the storage, financing and convenience-yield framework. The central lesson is simple: before taking commodity exposure, identify the contract you actually own, the way it expires, and the rules that replace it.

How do you read a futures curve without mixing up spot and delivery months?

A spot transaction concerns the commodity available now under a particular cash-market specification. A futures contract concerns a standardised delivery or settlement obligation at a later date. A curve places the prices of different delivery months alongside each other at the same observation time. It is a snapshot across maturities, not a time-series chart showing how one instrument moved over the last week.

Imagine a purely illustrative market with spot at $70, a nearer future at $72 and a deferred future at $76. Both futures are above spot, and the quoted maturities slope upward. Now imagine spot at $70, a nearer future at $68 and a deferred future at $65. Both futures are below spot, and the selected maturities slope downward. None of these invented prices describes oil, copper or any other instrument on publication day.

Illustrative observation	Contango example	Backwardation example
Spot price	$70	$70
Nearer futures price	$72	$68
Deferred futures price	$76	$65
Deferred minus nearer price	+$4	−$3

Real curves need not slope neatly in one direction. One nearby contract can trade above the next while more distant contracts rise again. Seasonal demand, refinery schedules, harvest timing, local inventory and delivery logistics can produce a hump or a dip. Calling the entire commodity “backwardated” without specifying the relevant section can hide the exact relationship that drives your exposure.

Write down the observation timestamp, exchange, commodity specification and contract months. Comparing yesterday's spot assessment with today's futures quote can create an apparent spread that partly reflects elapsed time. Comparing a cash assessment at one location with futures deliverable somewhere else also introduces a location basis. Those are different questions from the slope between matching contracts on the same exchange.

A useful habit is to replace “oil is in contango” with “this specified deferred WTI contract is above this nearer WTI contract at this timestamp.” The longer sentence forces precision. It makes the claim reproducible and prevents an article, a screenshot or an investment advertisement from substituting a broad label for a measurable comparison.

Why can a later futures contract cost more than the commodity today?

Holding a physical commodity can involve paying for storage, financing the inventory and insuring it. A deferred purchase lets the buyer avoid managing that inventory immediately. Under suitable conditions, those costs help explain why later delivery may command a higher futures price. CME identifies these carrying costs in its contango and backwardation course.

The intuition is easier to see from the perspective of a business rather than a chart. A company that purchases material now must fund it, secure storage and accept handling risks until it is needed. A company that arranges future delivery does not hold the same inventory during the intervening period. The difference between those choices has an economic value even if nobody expects the eventual cash price to rise.

Consider a deliberately simplified hypothetical calculation. Spot is $70, six months of financing cost $1.40 per unit, and storage plus insurance cost $1.10. If there is no assumed benefit from immediately holding inventory, adding those inputs gives an illustrative carrying-cost reference of $72.50. This is an arithmetic teaching model, not a quoted price, a universal valuation formula or a trading opportunity.

The reference depends on whether the commodity is storable, whether financing is actually available, and whether a trader can access the relevant storage and delivery arrangements. A textbook arbitrage assumes transactions that may be expensive or impossible for a retail investor. Transport restrictions, credit limits and differences in grade can prevent the clean comparison from translating into an executable trade.

Even a visible futures premium does not necessarily equal a producer's expected selling price. Forward curves bring together hedging demand, inventory economics, risk-bearing and market expectations. Some participants want to reduce business uncertainty rather than maximise investment returns. Treating the premium as a pure forecast ignores the reasons those participants are willing to transact.

What is convenience yield, and why can scarcity create backwardation?

Convenience yield is the implied economic benefit of holding the physical commodity rather than only a financial claim to later delivery. CME uses the practical example of keeping production running. A factory with usable inventory can continue operating when supply is disrupted; a future delivery contract may not solve an immediate shortage. That operational benefit can make the present commodity unusually valuable.

This is not automatically a cash dividend paid to the futures investor. It is an inventory benefit within the economic comparison. A retailer, refiner or manufacturer may value access to physical supply very differently from a fund that never intends to take delivery. The word “yield” can mislead readers if they assume it behaves like an interest payment arriving in their account.

In our hypothetical carrying-cost example, the $70 spot price plus $2.50 of financing, storage and insurance gave a $72.50 reference. If we assume an illustrative $4 benefit from immediate inventory access, subtracting it produces $68.50. The invented convenience benefit exceeds the invented carrying costs, so the reference moves below spot. Those inputs demonstrate the logic; they are not observable estimates for a current commodity.

Scarcity can be local and temporary. Inventory available at the wrong terminal, of the wrong grade or after the required production date may not solve an immediate problem. That is why a strong nearby premium does not prove there is a permanent global shortage. It can instead describe a particular bottleneck that later deliveries are expected to bypass.

Backwardation may therefore be consistent with tight current supply, but curve shape alone does not establish the cause. To assess a physical-market thesis, examine inventories, demand seasonality, delivery constraints and the specifications behind the contracts. Our copper outlook provides a related framework for separating supply conditions from outright price forecasts; its dated targets should not be treated as today's quotes.

Does contango predict rising prices or backwardation predict falling prices?

No reliable forecast follows from the label alone. A deferred contract trading above spot tells you what that contract costs now, not where spot must be when it expires. The future cash price can end above both today's spot and today's futures price, below both, or somewhere between them. Expectations matter, but they are not the only input in the curve.

This is a common source of contradictory headlines. An article may describe contango as “bullish because future prices are higher,” while a fund explanation describes it as “negative because rolling is expensive.” Both descriptions can miss the distinction between the level of future prices and the performance of an investment entered at those prices. An expensive entry is not the same thing as a forecasted gain.

Likewise, backwardation can accompany strong immediate demand while the eventual commodity price falls. A long investor who buys at a discount to today's spot still loses if settlement is below the purchase price. The discount provides a different starting point, not a safety net. A supply shock can reverse, demand can weaken, or the curve can change before a position is rolled.

There is also a technical distinction between an observed backwardated curve and “normal backwardation” in theories about futures risk premia. The latter concerns a futures price relative to an expected future spot price, which is not directly observable today. The two ideas can overlap, but they should not be treated as identical measurements. Our comparisons use observable spot and delivery-month prices, not an assumed consensus forecast.

The practical test is to ask what has to happen for your position to earn money. For an unleveraged price example, settlement must exceed a long entry price. For a rolling investment, each held contract's price path, collateral return and costs all matter. Saying “the curve is favourable” leaves too much of that return calculation unstated.

What actually happens when you roll a futures position?

Rolling means closing exposure in one delivery month and opening exposure in another. A continuously maintained commodity strategy generally cannot hold an expiring contract indefinitely. Its documentation specifies which maturities it uses and when it replaces them. The replacement changes the instrument being held even when the investor still thinks of the position as simply “long oil.”

Suppose a strategy closes a nearer contract quoted at $72 and opens a deferred contract quoted at $76. The price difference is $4, equal to 5.56% of the nearer quote after rounding. That calculation describes a spread between contracts. It does not mean the account immediately suffers a 5.56% realised loss merely because the two transactions occurred.

Futures are not normally purchased by paying the full quoted commodity value upfront in the same way as buying a share. They are subject to margin arrangements and gains or losses as prices change. The spread influences the exposure and the subsequent convergence experience, but a correct return statement needs the old entry, exit, new entry, position sizing and later prices. A two-price subtraction is not a complete account statement.

In an illustrative backwardated comparison, replacing a $68 nearer quote with a $65 deferred quote means the new delivery month has a $3 lower quoted price. That may create a helpful starting point if the deferred contract later rises toward an unchanged cash price. But it does not create an immediate free $3 payment or prevent the new contract from falling.

FINRA Regulatory Notice 10-51, published October 20, 2010, warns that futures-linked commodity products can perform differently from the underlying spot commodity, particularly over longer holding periods. The warning is relevant because a retail investor often observes a headline cash price while owning a product whose actual return follows a rolling futures strategy.

How can roll effects and outright price moves pull returns in different directions?

Here is the MarketsHQ controlled comparison. We assume an initial spot price of $70 and a hypothetical long futures entry of either $76 or $65. We then settle the contract at $60, $70 or $80. Every input is invented for instruction. There are no intervening rolls, no fees, no interest and no leverage in the percentage comparison; the result is a price change relative to quoted entry value, not a return on posted margin.

Hypothetical settlement	Spot move from $70	Long bought at $76	Long bought at $65
$60	−14.29%	−21.05%	−7.69%
$70	0.00%	−7.89%	+7.69%
$80	+14.29%	+5.26%	+23.08%

Each futures percentage is calculated as settlement minus entry, divided by entry, multiplied by 100. The spot column uses the same calculation with the $70 starting spot. We use positive starting prices so the usual percentage arithmetic remains interpretable. Contract quantities cancel in these price comparisons; dollar account profits would require a separately verified contract unit and actual position size.

The middle row isolates convergence under an unchanged spot assumption. The $76 long loses $6, or 7.89%, while the $65 long gains $5, or 7.69%. That is the clean scenario behind the intuition that contango can be a headwind and backwardation a tailwind for long exposure. It is conditional on the settlement being $70, not a claim that the market will deliver that outcome.

The top row prevents the most dangerous inference. The discounted $65 entry still loses 7.69% when settlement falls to $60. A favourable starting relationship cannot neutralise a sufficiently negative outright move. The bottom row shows the opposite: a $76 entry still earns 5.26% if settlement reaches $80, even though it began above the $70 spot reference.

Our decision rule is therefore not “buy backwardation and avoid contango.” It is “evaluate the outright thesis and the exposure mechanics separately, then combine them.” If a strategy depends on stable spot and persistent curve shape, state those assumptions. If it depends on an outright rally overwhelming convergence costs, quantify the required move rather than hiding it inside a bullish label.

Why is a negative roll effect not necessarily an instant loss?

A roll spread is a comparison at the switch. A realised trading profit or loss is a comparison between the entry and exit of a position. These measurements answer different questions. If a trader entered the old contract at $60 and closes it at $72, that contract has gained before the new $76 exposure is opened. Reporting only the $4 difference between old and new contracts erases the history that produced the account result.

Some investment presentations decompose returns into spot-related changes, roll yield and collateral return. Decomposition can be useful, but definitions differ across indices and research methods. Before comparing published roll-yield figures, inspect how each calculation treats contract selection, convergence, rebalancing and cash. An attribution label should explain a realised result rather than imply a separate source of guaranteed income.

The February 2014 Campbell research paper hosted by CME, Deconstructing Futures Returns, explores the role of roll yield in futures returns. Its broader relevance here is the need to distinguish the traded contract's experience from a cash-price narrative. We do not treat the paper's historical research as a promised outcome for a current investor.

A longer holding period can amplify repeated exposure to adverse convergence if similar conditions persist, but the losses cannot be calculated by blindly repeating today's spread every month. The curve may flatten, invert or move higher. The investment may change its chosen maturities. Fees and collateral conditions may change as well. A snapshot is evidence about current structure, not a complete backtest.

For the same reason, comparing the percentage change of a nearby continuous chart with a fund return can be misleading. A chart may splice contracts and adjust historical prices to avoid visible jumps. The displayed line can be useful for technical context while still not representing an executable buy-and-hold instrument. Always check the chart provider's series construction before using it as a performance benchmark.

How should you compare a commodity ETF with holding the commodity itself?

The first distinction is what the product owns. Some commodity exposures are physically backed; others use futures, swaps or other instruments. Shares in commodity producers are another category altogether: they represent businesses whose profitability also depends on operating costs, financing, management and other risks. A commodity label does not make these exposures interchangeable.

For a futures-linked product, read the benchmark methodology and prospectus before assuming a rise in spot will translate into the same percentage return. Identify the chosen maturities, roll window, collateral treatment, fees and any leverage. A strategy using a more distant contract can experience different curve dynamics from one repeatedly replacing the nearest month. Longer maturity is not inherently better; it changes the exposure.

Also distinguish an exchange-traded fund from an exchange-traded note. The legal structure and associated risks can differ, and the name of the underlying commodity does not answer those questions. FINRA's futures-linked securities notice is a useful starting point for understanding why product performance may diverge from spot and why investors need product-specific information.

Our ETF versus index fund comparison explains why an investment wrapper and the exposure inside it are different decisions. Here the same principle is especially important: deciding that copper demand is attractive does not tell you whether a copper futures fund, physical holding or mining company is the right expression of that view.

Do not rank products on expense ratio alone. A lower headline fee may be outweighed by a different benchmark, maturity exposure or structure. Conversely, a poor return relative to spot does not by itself prove mismanagement when the disclosed mandate tracks something else. Compare performance against the intended benchmark, then judge whether that benchmark fits the investment question you actually wanted to answer.

What did the negative WTI price in April 2020 really show?

On April 20, 2020, the expiring May 2020 NYMEX WTI futures contract settled at −$37.63 per barrel. The US Energy Information Administration documents that figure in its May 27, 2020 analysis of North American crude prices. This was a specific futures settlement linked to delivery at Cushing, Oklahoma, not a universal price for every barrel of oil worldwide.

EIA's April 22, 2020 petroleum report describes the role of storage constraints and limited liquidity near expiration. A participant unable or unwilling to accept the physical obligation could have a strong incentive to exit. The episode illustrates that delivery rules and local logistical capacity are part of financial risk, not technical details that can safely be ignored.

The May 27 EIA analysis provides a striking contemporaneous contrast: Louisiana Light Sweet was assessed at +$13.37 per barrel that day. Those two observations concerned different market specifications. Their $51.00 arithmetic difference is not an immediately tradable arbitrage because location, contract obligations, transport and timing matter. It is evidence that the phrase “oil went negative” omits important information.

Negative prices also break many casual percentage comparisons. Moving from −$37.63 to zero is a $37.63-per-barrel change, but using a negative initial price in a conventional percentage-return formula produces an unintuitive result. For this episode we report the verified price and dollar difference rather than presenting a sensational percentage rebound.

We deliberately do not calculate a full-contract dollar loss here because that would require a separately verified contract specification and an actual entry price. We also do not borrow today's contract documentation to imply that an individual investor held the relevant 2020 contract. Correct arithmetic cannot repair a missing assumption about what was actually traded.

The lesson is not that all contango markets will repeat April 2020. It is that nearby futures can become dominated by delivery and storage constraints under extraordinary conditions. The curve is one window into those conditions; position size, expiry management and the ability to exit remain independent risks. A historical exceptional event should illuminate the mechanism without being marketed as the normal outcome.

How does storage pressure differ from a normal seasonal curve?

A seasonal premium can exist because the commodity is more valuable at one predictable time of year. Storage pressure concerns the ability to carry actual supply through time. Both can affect delivery-month relationships, but the underlying questions differ. A seasonal curve asks when demand or supply is expected to be stronger; a logistical squeeze asks whether inventory can physically be accommodated or delivered.

An investor should therefore inspect the relevant dates rather than extrapolate a single calendar spread across the whole year. A contract spanning a demand season may have different economics from a neighbouring pair outside it. Comparing equal-length intervals can help avoid treating a seasonal price jump as a uniform financing premium, although it does not remove every other source of variation.

Inventory data also need careful interpretation. Total reported stock is not necessarily freely available stock at the contract's delivery location. Ownership commitments, quality restrictions and transport access can limit what participants can use. A broad inventory headline might be consistent with tight deliverable supply, or a local surplus might coexist with stronger conditions elsewhere.

For a commodity thesis, combine the curve with primary physical-market evidence rather than forcing every observation into a bullish or bearish narrative. Our oil outlook is related reading on the distinction between supply-policy catalysts and price forecasts. Its historical and forecast observations retain their own dates; this explainer does not silently update them.

Can a calendar spread remove the risk of the outright commodity price?

A calendar spread takes opposing positions in different delivery months of the same commodity. It can focus the trade on the relative relationship between those contracts rather than leaving only a single outright position. But “relative” does not mean risk-free. The spread can move sharply when the nearby delivery situation changes, and the two legs may not offset every risk equally.

Consider an invented nearer quote of $72 and deferred quote of $76. Define the spread explicitly as deferred minus nearer: it begins at +$4. If those prices later become $74 and $77, the spread is +$3. The deferred-minus-nearer measure has narrowed by $1 even though both contracts rose. A trader's profit depends on which leg was bought, which was sold, quantities and trading costs.

This sign convention matters because market discussions sometimes quote the same pair in the opposite order. “The spread strengthened” is ambiguous without a definition. Writing both contract months and the subtraction order avoids accidental reversal of the trade idea. The commodity's outright move and the spread's move should be recorded separately in any analysis.

Liquidity can differ between legs, and expiry risk can become concentrated in the nearby contract. Execution slippage, margin demands and unexpected delivery constraints can still matter. Before treating a spread as a safer alternative, test the actual exposure under stressed price relationships. Reduced sensitivity to one broad factor is not the same thing as protection against every adverse market scenario.

How do leverage, collateral and position sizing change the picture?

The hypothetical return table above deliberately uses entry quoted value as its denominator. A leveraged futures account posts margin rather than the entire notional exposure, so its percentage gain or loss on posted capital can be very different. Margin is a performance bond, not a cap on losses. An investor must understand the product's actual funding arrangements before converting a price change into an account-return claim.

A strategy can be directionally right over a longer horizon and still fail if interim losses create cash demands it cannot meet. The ability to fund adverse variation matters alongside the final thesis. The curve does not tell you how much liquidity to reserve, how a broker handles stressed conditions or whether the position can be reduced without substantial slippage.

Collateral treatment also influences a funded commodity strategy's total return. Cash or eligible collateral may earn income, but the rate and implementation depend on the product and market conditions. We do not insert a current interest rate here because we have not established a particular product's holdings or yield. Comparing futures-only returns with total returns without identifying this component can distort conclusions.

Position sizing should therefore start with a tolerable adverse scenario, not with the apparent attractiveness of backwardation. A favourable $3 spread in a teaching example says nothing about the maximum loss on a real portfolio. Costs, leverage, liquidation rules and the investor's time horizon must be considered before any commodity thesis becomes a trade.

What should you check before buying or rolling commodity exposure?

Start by naming the exposure precisely. Record whether it is physical, futures-linked, a derivative with broker-specific adjustments or equity in a producer. Then identify the actual benchmark or contract. The word “oil” in an app is not enough. Two instruments with similar labels can have different delivery dates, settlement conventions and holding costs.

Next, define the horizon. A short-term move in one delivery month is not the same question as maintaining commodity exposure for a year. For a rolling strategy, inspect when contracts are replaced and whether the investor can reproduce the benchmark's published return. For a contract held near expiry, confirm the obligations and the intermediary's handling rules well before the last trading period.

Separate the outright thesis from the curve thesis on paper. Write the condition that would support a rise in the commodity, then write the expected contribution or headwind from the selected maturity relationship. Finally, write a scenario in which the two disagree. This makes it harder to mistake a strong physical story for an automatically attractive financial exposure.

Compare at least a stable-price case, a falling-price case and a rising-price case using consistent assumptions. Our controlled table is a starting method, not a recommendation. For a real product, replace the invented inputs with timestamped observations and incorporate its disclosed methodology, fees and collateral treatment. Do not convert an illustrative sensitivity into a forecast probability.

Check the economic calendar and primary supply releases relevant to the commodity. The MarketsHQ economic calendar can provide broader macro context, but no calendar replaces contract documentation or delivery rules. If you choose to explore an account, use the tracked account link only after checking the provider's instrument specification and risk disclosures.

How was this analysis prepared, and what are its limits?

This is an educational analysis prepared with AI assistance, source checking and reproducible arithmetic. It uses CME's definitions, FINRA's investor-product warning and dated EIA evidence for the historical case. The named byline is MarketsHQ's commodities-specialist attribution; this article does not claim a personal trade, interview, independently observed physical-market position or human review that has not been documented.

All teaching prices are explicitly hypothetical. Percentages are rounded to two decimal places and computed from the stated positive entry values. The carrying-cost calculation is a simplified illustration, not an estimated fair value for a current market. We make no claim about today's WTI, Brent, copper, gold or cryptocurrency price, and no historical value in the article is presented as a live observation.

The analysis does not backtest a particular fund or assert that a persistent curve would have produced a specified realised portfolio return. Product-specific outcomes require product-specific data. Taxes, financing, transaction costs, liquidity and implementation can change results. A calendar-spread illustration likewise does not establish how a real spread would behave under stress.

The useful takeaway is a decision framework: identify the instrument, define the maturity comparison, distinguish current curve shape from a forecast, and evaluate actual price-path risk before relying on a roll narrative. Contango and backwardation are valuable descriptions once those foundations are clear. Without them, the labels can make a complex investment look far simpler than it is. This article is educational information, not personalised investment advice.