Market Research Glossary

Backwardation is a futures-market condition in which nearer delivery months trade above more deferred months, producing a downward-sloping segment of the futures term structure. It is the opposite of contango and is often called an inverted market. Backwardation is an observed relationship among contract prices at one point in time; it does not, by itself, mean that prices are certain to fall later.

Why the curve can invert

In physically deliverable, storable commodities, backwardation often appears when immediate availability is especially valuable. Holding inventory can keep a refinery, mill or production line operating, satisfy an unexpected order or reduce disruption risk. That operational benefit is commonly described as convenience yield. When inventories are tight and the value of possessing usable material becomes large relative to financing, storage and other carrying costs, nearby supply can command a premium over deferred delivery.

That mechanism is important, but it is not universal. Cash-settled or non-storable futures can also trade in backwardation even though physical inventory and convenience yield are not the relevant mechanism. A downward slope can reflect seasonality, expected changes in future supply or demand, contract-specific constraints, near-term uncertainty or other market structure. Backwardation can also exist in only part of a curve: one pair of delivery months may be inverted while later maturities have a different slope. Futures prices are current tradable prices for distinct delivery dates, not a pure forecast of future spot prices. The same backwardated shape can therefore arise from different economic conditions.

What the slope does and does not tell you

Backwardation matters directly for calendar spreads and for strategies that maintain futures exposure by rolling from an expiring contract into a later maturity. Selling a higher-priced nearby contract and buying a cheaper deferred contract does not, by itself, create an immediate profit from the price difference. The favorable return contribution commonly associated with backwardation arises if the newly held deferred contract subsequently rises relative to the underlying market as it ages and converges toward the nearby or spot price. If that pattern persists across repeated rolls, it can contribute positively to a long futures strategy. The effect is not guaranteed: outright prices can fall, the curve can flatten or reverse, and realized results depend on the contracts selected, the roll schedule and subsequent changes in the term structure.

Backwardation should also be distinguished from normal backwardation. The latter is a risk-premium theory in which a futures price is below the expected future spot price, which is not directly observable. Ordinary backwardation, by contrast, can be identified from contemporaneous market prices. An inverted curve may be consistent with tight nearby conditions or a high convenience yield, but it does not prove a particular future price path or prove normal-backwardation theory.

Basis is the price difference between a cash or spot market and a futures contract used to represent or hedge that exposure. In physical commodity markets, the common convention is basis = cash price − futures price: a positive basis means cash trades over futures, while a negative basis means cash trades under futures. The sign and construction are not universal across futures markets. Equity-index basis, for example, may be quoted as futures minus spot, while U.S. Treasury futures basis compares a cash security with a conversion-factor-adjusted futures price. A basis number therefore has little meaning unless the referenced cash instrument, futures contract and calculation convention are clear.

What makes the spread move

A futures hedge fixes the futures leg of a price relationship, not the eventual basis. For a local commodity market, basis can reflect freight, storage, handling, quality, deliverability and local supply-demand conditions. In financial futures, financing costs, income such as dividends, interest-rate relationships and contract-specific mechanics can dominate instead. Basis can strengthen or weaken even when cash and futures prices move in the same direction, because what matters is their relative movement.

Convergence is market-specific

For physically delivered futures, convergence links the expiring futures price to the cash value of commodities that can satisfy the contract under its delivery terms. That does not imply that every local cash quote must reach zero basis. A cash market at another location, for another grade or with different transport economics can retain a differential even as the relevant deliverable market converges. An apparent failure of local basis to reach zero may therefore reflect the fact that the cash price being compared is not the contract's deliverable economic equivalent.

Basis is an observation; basis risk is the uncertainty that this spread changes unfavorably between hedge initiation and completion. Research also needs a consistent construction. Switching the referenced futures month can change the measured basis even if the cash price is unchanged, so historical analysis should define the cash location and quality, futures month or roll rule, units and timestamp. Without those choices, two valid-looking basis series can describe materially different exposures.

A calendar spread combines a long position in one futures contract month with a short position in another contract month of the same futures product. A standard calendar spread is typically constructed one-to-one, although non-standard ratio versions can also exist. Its economic exposure is the price relationship between the two expiries: if both months move by roughly the same amount, the spread changes little; if one month strengthens relative to the other, the spread moves even when both futures rise or both fall.

Why the two months do not cancel perfectly

Because the legs share the same product, much of their common outright price exposure offsets, but not all of it. Their relative price changes with the futures term structure. In storable commodities, that relationship can reflect inventory availability, storage and financing costs, seasonality, and delivery-specific constraints; in financial futures, different carry and forward-pricing mechanics may dominate. The spread should therefore not be read as a direct forecast of future spot prices.

Calendar spreads are also central to contract rolls. A trader who is long the expiring contract and wants to remain long can sell the nearby contract while buying a deferred one; a short position is rolled in the opposite direction. When an exchange lists the calendar spread as a single strategy order, the two legs can be executed together, avoiding the temporary outright exposure created by manually entering one leg before the other. Reduced directional exposure does not mean zero risk. Margin systems may recognize offsets between expiries, but the position retains basis risk because different contract months are not perfectly correlated.

The sign depends on the convention

A practical trap is assuming that “buying the spread” has a universal leg direction. It does not. Exchanges can use buy-near/sell-deferred conventions for some products and sell-near/buy-deferred conventions for others; the displayed spread can likewise be calculated as near minus deferred or deferred minus near. Across markets, a month-pair label alone is therefore not a universal statement of economic direction. This matters in charting and backtesting because the same relative relationship represented with the opposite subtraction convention produces a sign-inverted series. Before comparing spread histories, define the contract months, the long and short legs, any leg ratio, and the exact subtraction convention.

The Commitments of Traders (COT) Report is the U.S. Commodity Futures Trading Commission’s weekly publication showing how open interest in covered futures markets is distributed among reportable trader categories and nonreportable positions; combined formats also incorporate options-on-futures exposure on a futures-equivalent basis. The reports describe positions as of Tuesday’s close and are generally released on Friday at 3:30 p.m. Eastern Time, so the data are a delayed positioning snapshot rather than a real-time measure. A market is included when 20 or more traders hold positions at or above the CFTC’s applicable reporting levels.

The threshold has an important consequence that is easy to miss. Clearing members, futures commission merchants and foreign brokers report large-trader positions to the CFTC daily. If a trader reaches the reporting level in any single futures month or option expiration, the reporting firm reports that trader’s entire position in all futures and option expirations in that commodity, not merely the portion above the threshold. In the published COT data, Nonreportable Positions are derived as the difference between total open interest and aggregate Reportable Positions; the report therefore provides no trader count or category breakdown for that residual.

One label, several classification systems

“COT” does not refer to one universal trader taxonomy. The Legacy report divides reportable open interest into Commercial and Non-Commercial traders. The Disaggregated report, used for physical commodity markets, separates Producer/Merchant/Processor/User, Swap Dealers, Managed Money and Other Reportables. Traders in Financial Futures (TFF) uses Dealer/Intermediary, Asset Manager/Institutional, Leveraged Funds and Other Reportables for financial contracts. The Supplemental report adds an Index Trader classification for selected agricultural markets. These are report families; Futures Only and Futures and Options Combined are report formats, not separate classification systems.

That format distinction is economically meaningful. In a combined report, option positions are converted to futures-equivalent positions with delta factors supplied by exchanges before being added to futures positions. A combined COT number therefore is not simply a count of futures contracts plus raw option contracts. Changes in option deltas can alter futures-equivalent exposure even when the number of option contracts itself has not changed.

What the snapshot can—and cannot—tell you

COT data reveal aggregated positioning under a regulatory classification framework; they do not identify the motive behind every position or provide a directional forecast. The CFTC classifies traders rather than each individual trading activity, so category membership does not establish whether a particular position is hedging, speculative or serving another purpose. Similarly named categories in different report families are not interchangeable. Large net positions or extreme historical readings can be useful research inputs, but they do not by themselves establish that a market is bullish, bearish or due to reverse.

For systematic research, the timestamp matters as much as the number. Tuesday’s positions are generally not publicly available until Friday afternoon, and holidays or exceptional disruptions can shift the release schedule. A backtest that acts on Tuesday using that week’s published COT figures would therefore introduce look-ahead bias. The defensible research question is not only “what was the position?” but also “which report family and format produced it, and when could a market participant actually have known it?”

Contango is a futures-market condition in which prices rise with delivery maturity, so a later-delivery contract trades above an earlier-delivery contract. The term is also commonly used when a futures contract trades above the current spot price. These comparisons are related but not identical: a futures contract can stand above spot even while part of the futures curve is flat or inverted. A market therefore need not be uniformly in contango across every maturity. Precise analysis should state which prices or contract months are being compared. Contango describes current price structure, not by itself a forecast that the underlying price will fall.

Carry can produce the slope without making a price forecast

For a storable commodity, the difference between spot and deferred futures can reflect carrying costs, including financing, storage and insurance, offset by the economic benefit of having inventory immediately available, commonly described as convenience yield. When net carrying costs dominate that benefit, futures above spot and an upward-sloping curve can be consistent with no-arbitrage pricing. Inventory conditions matter as well: abundant stocks and available storage often make contango easier to sustain, while scarcity can increase the value of immediate possession and push the curve toward backwardation.

That is why contango should not be read as a simple market forecast. Futures prices reflect carry economics and market expectations, and an upward-sloping curve can persist while both spot and futures prices rise, or while both fall. The same observed slope can also have more than one cause: changes in storage economics, financing, inventories or expectations can alter the curve. The curve shape is an observation; identifying its economic cause requires additional evidence.

The roll consequence is real, but it is not a bearish signal

Contango matters directly to traders or investors who maintain long exposure by repeatedly rolling futures. If the contract being sold is cheaper than the replacement contract, the position is rolled into a higher-priced maturity. Under the common roll-yield convention used for futures-linked products, that produces a negative roll-yield contribution and can drag performance relative to the commodity’s spot-price change. Persistent contango can therefore make a rolled futures strategy underperform spot. It does not guarantee a negative total return: the futures contract can appreciate before expiry, and the curve itself can flatten, steepen or reverse. Contango describes the price structure; the return outcome depends on how that structure evolves and on how the exposure is implemented.

Correlation measures how strongly two variables move together. In market research, the most common numerical measure is the Pearson correlation coefficient, r, which standardizes covariance by the two variables’ standard deviations. The result is dimensionless and ranges from −1 to +1: +1 indicates a perfect positive linear relationship, −1 a perfect negative linear relationship, and 0 zero linear correlation. That last case is easy to overread. A Pearson coefficient near zero can coexist with a strong nonlinear relationship.

One number can hide several relationships

A correlation coefficient is a property of the chosen data and sample, not a market mechanism. It does not establish causation, reveal which variable leads the other, or show that the relationship will persist. Two markets can move together because both are responding to a third factor. A full-sample estimate also averages across potentially different regimes. Rolling Correlation addresses that time variation by recalculating the statistic over a moving window. Cointegration asks a different question: whether a linear combination of non-stationary series is stationary, which concerns a long-run statistical relationship rather than the strength of contemporaneous co-movement.

The construction is part of the statistic

Daily-return correlation can differ from weekly-return correlation, and price levels, arithmetic returns and logarithmic returns need not tell the same story. With non-stationary time series, high association in levels can be spurious unless the time-series properties, including possible cointegration, are handled explicitly. Outliers matter too: ordinary Pearson correlation can be materially influenced by extreme observations. Sample dates, missing observations and continuous-futures construction can therefore change the estimate enough for two competent researchers to report different correlations for what appears to be the same market pair.

Correlation also removes scale, which creates a practical trap. Two assets can be almost perfectly positively correlated while one moves twice as much as the other. A one-for-one hedge would then leave material residual risk. In minimum-variance hedge sizing, relative volatility enters alongside correlation; high correlation alone does not determine the hedge ratio. The useful question is therefore not simply “What is the correlation?” but “Correlation of which variables, transformed how, over what sample, and for what decision?”

Cost of carry is the financing and holding economics that connect an asset’s spot price with its theoretical forward or futures price over a given horizon. The phrase does not have one universal accounting sign. In some futures terminology, it refers to the carrying charges incurred to own an asset, such as financing, storage and insurance. In other market practice, “carry” is quoted net of income or other benefits. The common principle is to compare the economics of owning the asset now with obtaining equivalent exposure for future delivery.

What is actually being carried?

For a storable physical commodity, ownership can require financing, storage, insurance and other inventory-related expenses. Physical inventory can also provide a convenience yield: the economic benefit of having the commodity available when it is needed. In a simplified representation, the theoretical deferred price reflects financing and storage costs offset by such holding benefits. A sufficiently high convenience yield can therefore outweigh carrying expenses and contribute to backwardation. Convenience yield is not itself another name for cost of carry; it is a distinct holding benefit that can offset costs in a net-carry framework.

Financial assets replace warehouse economics with different cash flows. For an equity index, financing the underlying shares pushes fair value upward while expected dividends work in the opposite direction. In FX, the relevant carry relationship comes from the interest rates of the two currencies, with the result also depending on the quotation convention. For Treasury securities, practitioners may compare coupon income with repo or other financing expense. The same no-arbitrage logic therefore survives across markets even though the inputs do not.

Why “positive carry” can mean the opposite of a positive cost

Sign convention is the practical trap. A cost-of-carry calculation may record financing and storage expenses as positive costs, so a larger net cost raises theoretical deferred value. Yet Treasury-futures practice commonly defines carry as coupon income minus financing cost, making positive carry a net holding benefit. Equity-index and FX materials likewise often describe positive carry in terms of income earned relative to financing. A carry figure is therefore incomplete information unless its components and sign convention are known.

Cost of carry is also distinct from roll yield. Cost of carry concerns the economics linking spot and deferred value at a given time; roll yield concerns the return effect associated with a futures position as contracts age and, where applicable, are replaced. Carry can help explain the shape of a futures curve, but the observed curve can also reflect convenience yield, inventory conditions and market frictions. It should not be read mechanically as a forecast of future spot prices or as evidence that one particular carry component caused the spread.

A futures contract is a standardized, exchange-traded derivative in which long and short positions take opposite obligations tied to a specified commodity, financial instrument, index or other defined reference for a future contract month. The exchange fixes the contract specifications in advance, while market participants choose whether to be long or short, how many contracts to trade and the price at which they transact. Depending on the contract, final settlement occurs through physical delivery or a cash payment.

Standardization changes who you are really trading with

Once a futures trade is accepted for clearing, the clearing organization becomes the central counterparty—buyer to the seller and seller to the buyer. The original buyer and seller are therefore not left with a continuing bilateral credit relationship. This clearing structure, combined with standardized contract terms, makes positions fungible enough to offset: a trader can normally close a long by selling the same contract, or close a short by buying it, without locating the original counterparty.

Margin is collateral, not the purchase price

The economic exposure of a futures position can be much larger than the cash posted to support it. Futures margin functions as a performance bond rather than a down payment, and open positions are marked to market using settlement prices. Gains and losses are credited or debited as prices move. That means notional exposure and margin are different quantities, and it also creates an important liquidity consequence: a position that later recovers can still require additional cash after adverse interim moves.

If a position remains open into a contract’s delivery or final-settlement process, the contract’s own rules determine what happens next. Cash-settled contracts are resolved financially; in physically deliverable contracts, open positions can incur obligations to make or take delivery during the delivery period. The relevant notice, last-trading, delivery and settlement dates are contract-specific and should not be collapsed into a generic notion of “expiration.” That is why traders need the exact contract specifications when carrying a position late in its life. A futures contract is best understood as a standardized, centrally cleared mechanism for transferring price risk through time—not as a prediction of where the spot price must be when the contract matures.

Futures term structure is the set of futures prices for the same underlying market across different contract months observed at the same point in time. The comparison is between contracts quoted using the same quotation convention and unit—for example, dollars per barrel, cents per bushel, index points or another market-specific quotation unit—while contract maturity is the dimension that changes. Often called the futures curve, it shows how the market prices delivery or settlement at different horizons. Curve slope is only one feature of that structure: the full curve can be upward sloping, downward sloping, flat, humped, seasonally patterned, or locally inverted. A calendar spread compares two maturities; the term structure is the broader cross-section from which those intermonth relationships are drawn.

What shapes prices across maturities

In storable commodity markets, differences between nearby and deferred prices can reflect financing, storage, insurance and the economic value of having inventory available now. Tight inventories can raise the convenience value of physical supply and support backwardation, while abundant inventories and carrying costs can support contango. Other futures markets have different carry mechanisms: interest-rate differentials matter in currency futures, for example, while seasonal supply-and-demand patterns can dominate parts of energy or agricultural curves. This is why contango and backwardation are useful labels but incomplete descriptions of the curve. A market can contain several distinct slopes or seasonal peaks within the same term structure.

A curve is not a forecast strip

The most important interpretation boundary is that a futures curve is not a strip of future spot prices or necessarily an unbiased forecast of them. Each point is a tradable price agreed today for a particular contract maturity. Hedging demand, carry, inventory conditions, arbitrage constraints and liquidity can all affect that price, and farther-dated contracts may contain less precise information when trading is thin. The curve can contain expectations about the future without being numerically identical to expected future spot prices.

Term-structure research also depends on construction choices. Prices should be compared on a consistent timestamp and quotation convention, and the contract set must be defined. A history built from “front month,” “second month” and similar relative labels changes the underlying delivery months whenever the market rolls. Consequently, part of an apparent historical change can come from contract replacement rather than a price jump in the same maturity. For reproducible analysis, named contract months or a clearly specified constant-maturity method should be distinguished from rolling ordinal contracts.

Managed Money is one of four CFTC classifications of reportable traders in the Disaggregated Commitments of Traders (COT) report. For this report, the CFTC defines a money manager as a registered commodity trading advisor (CTA), a registered commodity pool operator (CPO), or an unregistered fund identified by the CFTC; the explanatory notes also state that so-called hedge funds are included whether or not they are registered. Managed Money is therefore a regulatory reporting category, not a generic label for every speculative, hedge-fund, or professionally managed futures position.

The Disaggregated COT framework covers agriculture, petroleum and products, natural gas and products, electricity, metals, and other physical contracts. It is published in futures-only and futures-and-options-combined formats. Financial contracts reported under Traders in Financial Futures (TFF) use a different classification framework that includes Leveraged Funds, so Managed Money and Leveraged Funds should not be treated as interchangeable categories. COT reports generally describe Tuesday open interest and are released Friday at 3:30 p.m. Eastern Time, with holiday-related schedule exceptions.

The category classifies the trader, not the trade

Trader classification is based on the predominant business purpose reported on CFTC Form 40 and reviewed by CFTC staff for reasonableness. Traders may report business purpose by commodity and can therefore have different COT classifications in different commodities. The CFTC also states that it does not know the specific reason for each reported position. A Managed Money label consequently does not prove that every position held by traders in the category is a directional speculative bet.

This distinction matters when interpreting weekly changes. The CFTC notes that reported category totals can change because a trader reports a different primary business function, a new reportable trader enters the data, or an existing trader leaves the market. A large one-week change in Managed Money positioning therefore need not represent only buying or selling by an unchanged population of traders.

Managed Money open interest is reported as long, short, and spreading. Spreading is a computed amount equal to offsetting long and short positions held by a trader; any residual exposure is assigned to the long or short column, and inter-market spreads are not included in that computation. In the futures-and-options-combined report, option open interest and option positions are converted to a futures-equivalent basis using delta factors supplied by exchanges. A simple net position can be useful, but it compresses information contained in the gross long, gross short, and spreading structure.

Historical continuity has a caveat

The CFTC’s Disaggregated COT explanatory notes identify an important limitation in the historical data. Because the agency did not maintain a history of large-trader classifications, historical positions back to 2006 were classified using more recent trader classifications. The CFTC calls this a “backcasting” approach and states that its accuracy diminishes further back in time as trader classifications change.

For research, Managed Money is best treated as an aggregate snapshot of reportable traders assigned to a particular CFTC category at a point in time, not as a permanent roster of identical participants or a direct measure of speculative intent. Extreme net positioning, rapid weekly changes, or a large spreading component can be informative inputs, but none is a self-contained bullish, bearish, or reversal signal.