Bond Arbitrage Basics
Bond Arbitrage Basics
Bond arbitrage is a market-neutral strategy that profits from pricing discrepancies between two instruments that should theoretically trade at identical or predictable relative prices, earning returns with minimal directional risk.
Key takeaways
- On-the-run bonds (the most recently issued, most actively traded) trade at tight bid-ask spreads and are preferred by large institutional investors; off-the-run bonds (older issues of the same maturity) trade wider.
- The on-the-run/off-the-run spread is typically 0.5–2bp but can widen to 5–10bp during market stress, creating arbitrage opportunities.
- Basis trades compare futures (Treasury Note futures, for example) to the actual deliverable cash bonds and profit from pricing convergence at expiration.
- Relative-value arbitrage between two bonds (convertible bonds vs. stock plus straight bonds, or two tranches of the same deal) isolates mispricing.
- True arbitrage is rare and low-return by design; most "arbitrage" strategies carry execution risk, basis risk, or funding-rate uncertainty that reintroduces elements of directional bet.
On-the-run versus off-the-run dynamics
The U.S. Treasury issues new bonds regularly: fresh 2-year, 5-year, 10-year, and 30-year bonds enter the market at scheduled auctions. Each new issue is the "on-the-run" bond of its maturity. Once a newer bond is issued in the same maturity, the older bond becomes "off-the-run."
For example, in January 2024, the U.S. Treasury issued a new 10-year bond (the January 2024 on-the-run 10-year). Three months later in April 2024, a new 10-year was auctioned; the January bond was now off-the-run, and the April bond was on-the-run.
The on-the-run bond is the benchmark: the price point quoted in Bloomberg, the bond everyone watches. Large institutional investors (central banks, pension funds, hedge funds) trade on-the-run bonds to quickly establish or adjust Treasury exposure. A pension fund wanting 10-year duration exposure will buy the on-the-run 10-year, not hunt for an off-the-run issue.
The off-the-run bond is the same maturity but older, so it has a slightly shorter duration (as it has moved closer to maturity). If the on-the-run 10-year carries duration of 8.2 years, the off-the-run (now 9.95 years from issuance) has duration of 8.15 years. This small difference in duration should result in a slightly lower yield for the off-the-run bond, but the illiquidity discount often outweighs the duration benefit.
Consequently, off-the-run bonds often trade 0.5–2bp wider than on-the-run bonds of the same maturity. During normal conditions, a dealer will always have on-the-run inventory (sells to meet demand); off-the-run bonds are harder to borrow and less liquid. A bid-ask spread for an on-the-run 10-year is typically 1–2bp; an off-the-run 10-year might be 3–4bp. The effective cost of transacting in the off-the-run market is higher.
The on-the-run/off-the-run arbitrage
An arbitrageur exploits this spread discrepancy by shorting the on-the-run (selling what is readily available and liquid) and buying the off-the-run (buying the mispriced-cheap security). If the on-the-run 10-year is yielding 4.2% and the off-the-run 10-year (which should yield slightly less due to marginally shorter duration) is yielding 4.25%, the 5bp spread offers a trading opportunity.
Here's the execution:
- Borrow (via repo) the on-the-run 10-year bond (secure a short position); the repo rate is typically very low (0.5–1%) because on-the-run bonds are easy to borrow.
- Sell the borrowed bond (short sale).
- Simultaneously, buy the off-the-run 10-year bond in the cash market.
- The short is now 20k notional of on-the-run bonds; the long is 20k notional of off-the-run bonds.
The P&L of the trade:
- Long off-the-run yields 4.25%; long side accrues 425bp per year.
- Short on-the-run costs 4.2% carry; short side costs 420bp per year.
- Gross carry spread: 5bp per year on the long position.
- Minus repo cost (short on-the-run repo): ~75bp per year (at 0.75% repo).
- Net carry: 5bp — 0.75bp = 4.25bp per year.
The arbitrageur holds this trade for three months (until a new on-the-run issue is auctioned and the old on-the-run becomes off-the-run). At that point, both bonds become identical or nearly identical (they roll into the same-or-similar status), and the spread compresses to near-zero. The trade realizes the 4.25bp carry plus 2–3bp of spread convergence, netting 6–7bp of return on a three-month hold, or approximately 25–28bp annualized.
That return, while "risk-free" from a market-neutral perspective, is not without cost. The arbitrageur must maintain both positions (pay repo, mark-to-market, manage collateral), source the on-the-run bond to borrow (a small fee if the bond is in high demand), and execute the trades efficiently. Total transaction costs and funding costs might eat 10–15bp of the expected 6–7bp gain, leaving only 0–5bp net profit. This is why these trades are typically executed by dealers and large hedge funds with low funding costs, not retail investors.
Treasury futures basis arbitrage
The basis is the difference between a Treasury futures contract price and the price of the cheapest-to-deliver (CTD) bond in the futures contract. For the 10-year Treasury Note futures contract, the exchange specifies which bonds are eligible for delivery; the futures seller will deliver whichever is economically cheapest.
Example: The December 2024 10-year Treasury Note futures contract is priced at 109.50. The CTD bond (the bond the futures seller will likely deliver) is the January 2024 on-the-run 10-year, trading at 109.48 in the cash market. The basis is +2 ticks (where one tick is 1/32 = 3.125bp).
An arbitrageur can:
- Buy the futures contract (long the futures).
- Short the CTD bond in the cash market.
- Wait for expiration (when futures must be settled by delivery or cash settlement).
At delivery, the futures converges to the cash price (they become identical). If the futures were overpriced (basis positive), the arbitrageur profits on the long futures position and breaks even (or profits from carry and financing) on the short cash position.
In practice, the basis fluctuates due to repo rates, supply, and demand. If the CTD bond is in high demand from major investors (e.g., central banks, insurance companies), it becomes "special" in the repo market, and repo rates for it decline below general repo rates. This makes the CTD cheap to finance if you are short it, widening the basis and benefiting the arbitrageur. The trade is nearly risk-free if held to expiration.
The profitability is modest: 2–5bp per month of hold depending on repo rates and basis dynamics. But leverage makes it worthwhile: a 4× levered position on a 3bp monthly basis trade yields 12bp monthly return, or about 150bp annualized—sufficient to justify the infrastructure and funding costs.
Convertible bond arbitrage
A convertible bond is a hybrid: a straight bond with an embedded equity call option. It should theoretically trade at a price equal to the straight bond value plus the value of the equity call.
Straight bond value = present value of coupons and principal discounted at the bond's credit spread. Call option value = the value of the option to convert into a fixed number of shares (set at issuance). Convertible price = straight bond value + call option value.
If the convertible is trading below this theoretical fair value, an arbitrageur can:
- Buy the convertible bond.
- Short the underlying stock (hedge the equity upside).
- Buy a call option on the stock (to cap the hedging loss if the stock rallies sharply).
The position isolates the straight-bond component. If the convertible reprices to fair value, the trade profits from the reversion. The short stock hedge protects against large equity moves (it offsets the embedded call's losses if the stock rallies), leaving only the bond component and the cost of the hedging option.
In January 2024, a tech company (fictitious: TechCorp) issued a convertible bond yielding 3% with a call option on its stock struck at 50 per share (stock trading at 45). The fair value of the straight bond (without the call) was estimated at 98, and the call option at 5; fair convertible price should be 103. The convertible was trading at 100. An arbitrageur would buy the convertible at 100, short 100 TechCorp shares at 45 per share, and buy a call at 50. The net cost: buy convertible (100) + call (5) — short sale proceeds (4,500 / 45 = 100 share equivalents) = net outflow is the cost of hedging. The trade profit if the convertible reprices to 103 and the stock does not move.
Convertible arbitrage is more complex than Treasury basis trades because it requires hedging, gamma risk (if the stock volatility changes), and credit event risk (the issuer's credit deteriorates). But it is still a high-precision strategy exploiting isolated mispricing, not a directional bet.
Financing-rate risk and the illusion of risk-free returns
The major hidden risk in all arbitrage strategies is the financing rate (the repo rate on the short or leveraged position). If an arbitrageur is short the on-the-run Treasury bond with an on-the-run repo rate of 0.75%, and the repo rate is guaranteed to stay at 0.75% for six months, the arbitrage is nearly risk-free.
But repo rates are not guaranteed; they are market prices. During the September 2019 repo crisis, the federal funds effective rate spiked from 2% to 10% intraday; Treasury repo rates moved in tandem. An arbitrageur short on-the-run bonds at an assumed 0.75% repo rate suddenly faced 5% or 10% repo costs, turning a profitable trade into a loss. Arbitrageurs who were leveraged and could not fund the mark-to-market losses were forced to liquidate.
Similarly, if the shorted security becomes "special" in the repo market (high demand), its repo rate can turn negative, meaning the short position is paid to maintain the short. This changes arbitrage dynamics but also introduces the risk of the special status reversing (the security becomes unpopular, special repo rates normalize), eliminating the edge.
Next
Arbitrage strategies, while intellectually elegant and seemingly risk-free, require deep expertise, low funding costs, and constant monitoring. Most investors are not equipped to compete in arbitrage; they are better served by focusing on fundamental value and relative value. The next article pivots to international bond markets, where fundamental opportunities abound in developed sovereign debt and emerging-market debt.