The Government Curve vs the Swap Curve as Benchmark
Government bond yields and swap rates both claim to represent risk-free interest rates, but they diverge because of supply, liquidity, and credit differences — and which one a desk uses as its benchmark depends on what it's actually pricing.
Prerequisites: Swap Spreads, SOFR and Risk-Free Rate Benchmarks
Ask what "the risk-free rate" is at any given maturity, and there are two very different market-based answers available: the yield on a government bond of that maturity, or the fixed rate on an interest-rate swap of that maturity. Both are widely quoted, both are liquid, and both get called "risk-free" loosely in conversation — but they routinely trade at different levels, and the gap between them, the swap spread, moves for reasons that have nothing to do with either curve becoming more or less risky in any simple sense.
Government yields reflect a specific sovereign's actual credit and, crucially, the supply and demand for that government's bonds. Swap rates reflect the fixed leg of an interest-rate swap between bank counterparties, referencing an overnight benchmark rate. The two track each other closely but diverge with issuance patterns, regulatory balance-sheet costs, and flight-to-quality flows — which is exactly why each has its own dedicated use case.
Why the two curves aren't identical
A government bond is a direct obligation of the sovereign, funded by real cash paid upfront by the buyer — its yield reflects that sovereign's credit plus, importantly, how much investors are willing to pay for the safety and liquidity of holding that specific paper. A swap, by contrast, involves no upfront principal exchange at all — two counterparties simply agree to exchange a fixed rate for a floating reference rate on a notional amount, with credit risk mitigated through collateral posting (variation margin). The swap rate is set purely by expectations for the floating reference rate path plus a small risk premium, unaffected by government-specific issuance or safe-haven demand.
In words: the swap spread captures everything that makes swaps and government bonds price differently at the same maturity — funding costs, collateral rules, relative supply of each instrument, and demand for the specific safety of sovereign paper.
Worked example
The 10-year government bond yields 4.20%. The 10-year swap rate is 4.05%.
- Swap spread: , or -15 basis points — a negative swap spread, common in some markets when heavy government bond issuance pushes government yields up relative to swaps.
- Interpretation: this does not mean the government is viewed as riskier than the average bank counterparty in the swap market — it more likely reflects that there's simply a lot of government paper for the market to absorb, pushing its yield up (price down) relative to a swap contract that requires no funding of principal at all.
- A trading implication: a desk that wants exposure to interest-rate moves without directly funding a bond purchase, or without taking on repo and balance-sheet costs, may prefer the swap; a desk needing an actual safe asset to hold as collateral needs the government bond regardless of which one is "cheaper" on a yield basis.
What this means in practice
Derivatives desks use the swap curve (now typically built on SOFR or another risk-free reference rate) as the default discounting curve for pricing most derivatives, because most derivatives are themselves collateralized like swaps. Cash bond desks and anyone actually funding a real bond purchase use the government curve as their natural benchmark. Comparing a corporate bond's spread to government yields versus to swap rates is a genuinely different measurement, and both are quoted in practice — "spread to Treasuries" and "spread to swaps" — precisely because neither curve is uniformly the "correct" benchmark.
A negative swap spread is not evidence that a government is a worse credit risk than commercial banks. It usually reflects balance-sheet and regulatory costs of holding government bonds on dealer balance sheets, or the sheer amount of government debt the market has to absorb — supply and structural factors, not a literal credit-quality signal.
Further reading
- Hull, Options, Futures, and Other Derivatives (ch. 4)