A relative-value checklist for Agency and SSA bonds

Independent research framework. All spread, duration, funding and transaction-cost inputs below are hypothetical, not market quotes.

A five-year agency bond yielding 40 basis points more than a Treasury offers a starting point for research. To judge its value, an analyst still needs to identify the issuer's obligations, check for embedded options, align the rate risk, and estimate the cost of owning and hedging it.

Agency and SSA markets cover a range of borrowers. SSA usually refers to sovereigns, supranationals and agencies, with market classifications sometimes including sub-sovereigns. A shared sector label does not establish a shared guarantee, liquidity profile or funding cost.

Begin with the legal claim

The Federal Home Loan Banks' Office of Finance states that consolidated obligations are not obligations of the United States and are not guaranteed by the United States. An analyst should therefore distinguish an agency or government-sponsored-enterprise label from an explicit sovereign guarantee.

For a supranational example, the World Bank's debt-products FAQ explains that IBRD debt instruments are not direct obligations of any government. The institution's capital commitments and financial policies support its credit standing. Those arrangements require their own analysis rather than treating an IBRD bond as another US Treasury.

For each security, the research record should identify the legal issuer, currency, seniority and any guarantee, then link the offering document. Issue size, settlement venue and available dealer liquidity also affect how readily an investor can enter or exit a position.

A call option changes the comparison

FHLBank bonds can be callable or non-callable. The Office of Finance's callable-bond documentation describes European, Bermudan and American call structures, with issue-specific call schedules and lockout periods.

The investor in a callable bond has sold an option to the issuer. A fall in rates can make refinancing attractive to the issuer, limiting the investor's price upside and creating reinvestment risk. A rise in rates can make a call less likely and extend the bond's expected life. Effective duration therefore changes with the rate and volatility assumptions.

A callable bond with an 80 bp yield pickup and a bullet with a 40 bp pickup cannot be ranked from those two numbers. Some of the callable's extra yield compensates the investor for the embedded option. Yield to maturity also assumes the bond remains outstanding until maturity, which may not describe the relevant scenario.

Use a spread measure that fits the question

These measures require consistent settlement, day-count and compounding conventions. For an option-free USD bullet, a Treasury curve spread can support a first relative-value screen. For a callable, a model-based comparison should include effective duration and sensitivity to volatility as well as OAS. An OAS estimate is conditional on those inputs; it is not a model-free valuation.

Two bonds with the same final maturity can also load on different points of the Treasury curve. A hedge can match overall DV01 and still retain exposure to a curve twist. Key-rate durations make that residual exposure more visible.

Work through a Treasury-hedged scenario

Assume a near-par USD bullet with 40 bp of annualized spread over its Treasury comparator. Assume matched Treasury rate exposure, spread duration 4.5, and a three-month horizon. Add 10 bp of extra annualized financing cost relative to the Treasury position and 6 bp of round-trip trading costs, measured against the bond's initial market value.

Ignoring roll and changes in rate-hedge sensitivity, the first-order relative return in basis points is:

Only 1.5 bp remains in the unchanged-spread case after the assumed financing and transaction costs. About one-third of a basis point of spread widening would consume that cushion under the linear model. The 40 bp headline pickup is an annualized yield spread; it is not a 40 bp three-month profit.

An actual package needs each leg's cash flows and financing balances. The hedge can require different market values on the two legs, and Treasury specialness can affect its economics. This example treats the extra funding cost as a net assumption and holds spread duration fixed; it does not simulate an executable trade.

Define what would make the spread worth owning

A wider spread could reflect a new-issue concession, weaker liquidity, a change in perceived credit quality, or demand for a particular maturity. Identifying the likely explanation helps determine whether convergence is plausible and how long it might take.

Before favoring a bond, I would want evidence that the spread compensates for those risks after the expected holding costs. The note should state a catalyst and horizon, then identify what would invalidate the view: an unfavorable credit development, persistent supply, a loss of liquidity, or a change in the issuer's expected call behavior.

Public issuer documents can establish structure and legal terms. They cannot provide executable bid-offer levels or position-specific financing. A defensible research conclusion should make that data boundary clear and specify the quotes needed to finish the comparison.

Reproduce the scenario

The scenario CSV exposes the assumptions and each resulting return. The Python script calculates them in basis points; the research package includes methodology and companion Treasury, TIPS and repo exhibits.

For the financing framework behind this comparison, see Treasury carry after the funding bill.