Basis trades: positive and negative basis in bonds, CDS and futures

Advanced11 min read · updated 2 September 2026 · 13 sources

A basis trade bets that two prices for almost the same risk will converge: a bond and its credit derivative, a government bond and its future, an index and its futures contract. The gap between them is called the basis. Because it usually amounts to only a fraction of a percentage point, such trades are run with high leverage, and that is exactly what made them a problem in 2008 and in March 2020. Here you will learn how positive and negative bases arise, how to calculate the profit and the implied repo rate, and which risks remain.

What is meant by the basis

According to the law of one price, two routes to the same cash flow should cost the same. The basis measures how far apart they are. Three forms are common:

• CDS-bond basis: the premium of a credit default swap minus the spread of a bond from the same issuer (CDS-bond basis).
• Cash-futures basis in government bonds: the cash price of a deliverable bond minus the futures price times the conversion factor.
• Basis in index and commodity futures: the spot price minus the futures price; it reflects interest, dividends, storage costs and the benefit of holding the physical commodity (cost of carry).

A basis trade buys the relatively cheap side and sells the expensive one. It does not bet on interest rates or credit quality but on the gap, and it needs financing to hold the position until the gap closes.

CDS-bond basis: formula and direction

If you buy a bond and at the same time protect yourself against the issuer's default with a CDS, you roughly hold a risk-free bond. The bond's spread over the risk-free rate should therefore be about equal to the CDS premium. The spread used is the Z-spread over the swap curve or the asset swap spread, where the bond's interest rate risk is swapped into a floating reference rate.

• Negative basis, the CDS is cheaper than the spread: the negative basis trade buys the bond, finances it through a repo and buys CDS protection. It earns roughly the size of the basis as long as financing stays cheap.
• Positive basis, the CDS is more expensive: the positive basis trade sells CDS protection and short-sells the bond, which it borrows through a reverse repo. This is harder to implement because corporate bonds are often scarce and expensive to borrow.

Worked example: a negative basis trade

Why the basis is not zero

• Financing: the bond has to be paid for, the CDS hardly at all. If repo rates or haircuts rise, the negative basis trade becomes less attractive and the basis becomes more negative.
• Cheapest-to-deliver option: in a credit event, the cheapest deliverable bond of the issuer determines what the protection seller pays (cheapest to deliver). This makes the CDS more valuable and pushes the basis up.
• Counterparty risk: if the protection seller defaults, the protection is gone. This lowers the value of the CDS and pushes the basis down.
• Liquidity and borrowing costs: if the bond is hard to trade, buyers demand more spread; if it is expensive to borrow, the positive basis trade becomes expensive.
• Balance sheet costs: repo financing takes up bank balance sheets. When balance sheet becomes scarce, for instance at quarter ends, such gaps widen, much as with covered interest parity (Du et al. 2018).
• Price far from par: the CDS protects the face value. If the bond trades well below 100, the CDS covers more than the purchase price, and the spreads are no longer directly comparable.

Bai and Collin-Dufresne (2019) confirm this in the cross-section: the deviations are larger for bonds with more frictions, i.e. low trading liquidity, higher financing costs, more counterparty risk and poorer quality as collateral. Surprisingly, the basis was more negative when the bond's lending fee was high: arbitrageurs apparently shied away from the negative basis trade when many were betting on falling bond prices.

Treasury basis trade: cash versus futures

The best-known basis trade takes place in the US Treasury market. A fund buys a deliverable bond, sells the matching future and finances the purchase through repo. On delivery, the buyer of the future pays the invoice amount: the futures price times the conversion factor plus accrued interest. Of all deliverable bonds, the seller delivers the one that is cheapest for them (cheapest to deliver). Whether the trade pays off is shown by the implied repo rate: the return from buying today and delivering into the future (Hull 2021).

March 2020 and the regulators' concerns

Using supervisory data, Barth and Kahn (2021) of the US Treasury's Office of Financial Research show that the trade became popular with hedge funds after 2016: at times it accounted for up to half of their US Treasury positions and around a quarter of dealers' repo lending. In March 2020, the market's illiquidity during the pandemic turned many risks into reality, and hedge funds sold large parts of their positions. In their assessment, the unwinding was more a consequence than a main cause of the stress; the Federal Reserve's swift intervention probably prevented the trade from making the situation worse.

Federal Reserve economists reach similar findings (Kruttli et al. 2021): hedge funds' gross positions in US Treasuries doubled from 2018 to February 2020 to $2.4 trillion, mainly through relative value arbitrage with repo financing. In March 2020, the average fund lost 7% and cut its positions by almost 20%; funds focused on the cash-futures basis came under the greatest margin pressure. Economists at the BIS (Bank for International Settlements) saw fire sales by investors trying to exploit small price differences with leverage, and warned that large holdings by leveraged investors could impair market functioning and make central bank intervention necessary (Schrimpf et al. 2020). In 2023, Fed economists asked whether the basis trade was back. Their answer: probably, at least in part (Barth et al. 2023).

Cash-and-carry in index and commodity futures

For equity index futures, the fair price follows from the spot price, the interest rate and expected dividends; for commodities, storage costs are added, minus the benefit of holding the commodity itself. If the future trades above its fair value, cash-and-carry pays off: buy the underlying, sell the future. If it trades below, the reverse applies, provided the underlying can be sold short. A future above the spot price is said to be in contango, one below it in backwardation.

Advanced: delivery options, margins, spreads

Delivery options: the seller of a bond future chooses which bond to deliver and, within the delivery month, when. These options belong to the seller; that is why the net basis is normally positive and the implied repo rate is below the repo rate. A negative net basis, as in the example, means that the future is more expensive than carry and delivery options justify.

Margins as a price: Gârleanu and Pedersen (2011) explain bases with margin requirements. Capital-constrained investors demand a higher return for securities with high margins; in funding crises the constraints become binding, and price gaps open up between securities with the same cash flows but different margins. Their size depends on the shadow cost of capital, which can be measured by the interest rate spread between collateralised and uncollateralised loans. In simplified form:

Asset swap spread or Z-spread: the asset swap spread depends on the bond's price, while the Z-spread shifts the entire zero-coupon curve. Near par the two are close together; at prices far from 100, the choice can change the basis noticeably. Default: the CDS pays the face value minus the recovery rate. If you bought the bond above par, you lose the premium in a default.

What research shows

Capital moves slowly. Duffie (2010) describes how prices react sharply to supply or demand shocks and then recover only slowly, because capital needs time to move to the opportunities. In the financial crisis, particular obstacles hindered the formation of capital and distorted prices; afterwards, the distortions receded.

In 2008 the basis turned deeply negative. After the collapse of Lehman Brothers, the CDS-bond basis became strongly negative. Choi and Shachar (2013) find that the main reason was hedge funds unwinding the basis trades they had built up before the crisis; dealers actually increased their bond inventories but were unable to close the gap. Mitchell and Pulvino (2012) explain why: the threatened failure of prime brokers abruptly cut hedge funds' leverage, seemingly long-term financing became short-term, and arbitrageurs could no longer hold identical or similar securities at similar prices. Gârleanu and Pedersen (2011) test their model on precisely this basis.

Arbitrage has limits. This fits with Shleifer and Vishny (1997): anyone arbitraging with other people's money may have to give up precisely when prices are furthest from fair value. Rising margins can turn this into a spiral (Brunnermeier and Pedersen 2009).

Risks and limitations

• Leverage: the basis is small; the return comes from leverage, and so does the loss.
• Margin calls and haircuts: rising repo haircuts and higher futures margins force additional payments precisely under stress (margin call).
• The basis can widen: until convergence, there is a risk of mark-to-market losses that can exceed the equity.
• Counterparty: repo lenders, CDS sellers and prime brokers can default or withdraw financing.
• Crowded trade: many funds in the same trade sell at the same time under stress.
• Contract details: the credit event, deliverable bonds and conversion factors determine the outcome. The examples are heavily simplified and are not investment advice.

How Portfolio Lab calculates

Portfolio Lab has no tool for basis trades: the app knows neither CDS nor repo financing nor delivery options. A related tool is the Carry backtester. Its bond carry selects maturities of US Treasuries or German federal securities (Bundeswertpapiere) by coupon plus roll-down assuming an unchanged curve, optionally duration-neutral long and short between two maturities (carry methodology). How the default risk of issuers and counterparties enters prices is explained in CVA and issuer risk; price, yield and duration are worked through in bond maths.

This article explains a method; it is not investment or tax advice. Worked examples are simplified and rounded. All terms in the glossary.