Futures roll mechanics: how and when to roll futures contracts, and what it costs
Maintaining long-term exposure through futures is not the same as holding a stock. Every few weeks, the front-month contract approaches expiry and must be replaced with the next one. That transition—the roll—is not administrative noise. In a market with persistent contango, roll costs compound month after month into a drag that can turn a profitable spot-price bet into a loss. In backwardation, the roll earns money. The difference between the two states is the single largest source of return variation across commodity futures strategies.
What rolling is
Rolling means simultaneously closing an expiring futures contract and opening a position in the next expiry—typically the following month's contract, though longer-dated contracts are used in some systematic strategies. The result is continuous market exposure without ever taking delivery of the underlying commodity or financial instrument.
The decision of when to roll is called the roll window. Most institutional traders and systematic strategies roll during a window of five to ten days before the contract's first notice day or expiry date. Rolling too close to expiry risks encountering thin liquidity, wider bid-ask spreads, and the possibility—in physically settled contracts—of receiving a delivery notice. Rolling too early means moving to a less liquid contract before it has become the market's primary focus.
How roll cost is calculated
The roll cost (or roll yield, when positive) equals the price of the new contract minus the price of the expiring contract. When the market is in contango—where further-dated contracts trade at a premium to nearer ones—the roll is a net cost. You sell the expiring contract at the lower price and buy the next one at the higher price. When the market is in backwardation—where further-dated contracts trade at a discount—the roll earns money. You sell the expiring contract at the higher price and buy the next one at the lower price.
The magnitude matters. Suppose a crude oil position is being rolled: the expiring contract trades at USD 80 per barrel, and the next-month contract trades at USD 82. Rolling one NYMEX crude oil contract (1,000 barrels) costs USD 2,000. In a period of persistent contango—as crude oil frequently experiences outside supply shocks—that cost recurs every month. Over twelve months, a long position loses USD 24,000 per contract in roll costs alone, independent of any change in spot price. Over five years at a consistent USD 2 monthly roll cost, the cumulative drag on a single contract is USD 120,000.
For equity index futures—the S&P 500 E-mini (ES), the Euro Stoxx 50, and most major index contracts—the roll is structurally a cost in most interest-rate environments. Equity index futures trade at a premium to spot because they embed the cost of carry: the risk-free rate minus the dividend yield. When interest rates exceed the dividend yield, equity futures are in contango and rolling costs money.
What the evidence shows
Erb & Harvey (2006), The Strategic and Tactical Value of Commodity Futures, Financial Analysts Journal, found that the roll yield component of commodity futures returns has historically been one of the largest contributors to total return variation across contracts. In markets with persistent contango, such as crude oil and natural gas at various periods, the roll yield has been substantially negative—meaning investors in those contracts lost money from rolling even during periods when spot prices were rising.
The 2009–2011 period is the canonical example. WTI crude oil spot prices recovered strongly from post-crisis lows, yet long-only commodity ETFs based on front-month crude oil futures returned far less because contango in the futures curve eroded gains at each monthly roll. The United States Oil Fund (USO) gained approximately 17% over a period when the front-month crude oil spot price rose more than 60%—the difference was almost entirely roll drag.
Limitations
Roll timing is not standardised. Different systematic strategies, different data providers, and different institutional desks roll on different days within the expiry window. Two apparently identical strategies can produce different backtested returns purely from differing roll dates. The practical impact is greatest in commodities with steep term structures, where rolling one week earlier or later can mean the difference between catching peak contango or a flatter part of the curve.
Bid-ask spreads also widen as a contract approaches expiry and liquidity migrates to the next contract. Large positions face market impact on the roll—executing a significant sell in the expiring contract while buying the next can move both markets, particularly in less liquid commodity contracts. This execution cost is separate from, and in addition to, the roll yield cost.
Futures roll mechanics in pfolio
pfolio's continuous futures chain builder embeds the roll cost assumption directly into each constructed price series. The roll date parameter—the number of days before expiry at which the series transitions to the next contract—is configurable, allowing users to model different rolling strategies and observe their effect on backtested returns. Because the roll cost is embedded in the continuous series, any performance metric calculated from that series reflects what an investor following that roll schedule would actually have earned, roll costs included. Users can compare a five-day and a ten-day roll schedule side by side to quantify the difference for any specific instrument and period.
Related articles
Disclaimer
Get started now
