Understanding Option-Adjusted Spread (OAS)

The option-adjusted spread (OAS) measures the yield spread a fixed-income security offers over a risk-free benchmark (typically Treasuries) after adjusting for any embedded options—such as callable or mortgage-backed securities (MBS) features—that can alter the bond’s future cash flows. OAS helps investors estimate the compensation they receive for credit, liquidity and other risks once option-related uncertainties are removed.

Key takeaways
* OAS isolates the spread attributable to credit and other non-option risks by stripping out the value of embedded options.
* It produces a more accurate relative-value measure for securities with embedded options than yield-to-maturity alone.
* Estimating OAS requires modeling future interest-rate paths and option exercise (e.g., prepayment) behavior, often via Monte Carlo simulations.
* OAS differs from the Z-spread: the Z-spread is a static constant added to each point of the yield curve and does not reflect option value.

How OAS is determined
* Separate the security into two pieces: a vanilla bond and an embedded option.
* Model many possible future interest-rate scenarios (and, for MBS, borrower prepayment behavior).
* Value the bond’s cash flows under each scenario, net out the option value, and find the spread (over the risk-free curve) that equates the option-free bond’s present value to market price.
* The resulting OAS represents the spread attributable to non-option risks.

How options and volatility affect OAS
* Embedded options change expected cash flows:
* Call options allow issuers to redeem early, reducing future coupon income for investors.
* Put options allow holders to sell the bond back, limiting downside for investors.
* Two primary volatility sources for OAS calculations:
* Interest-rate volatility, which affects the likelihood of option exercise.
* Prepayment volatility (for MBS), driven by borrower behavior and economic incentives.
* Greater volatility typically increases the value of options and therefore alters the OAS necessary to match observed prices.
* Limitations: OAS depends on model assumptions and historical inputs; inaccurate assumptions about future rate dynamics or borrower behavior can misstate the spread.

OAS versus Z-spread
* Z-spread: the constant spread added to each point on the Treasury curve that makes a bond’s discounted cash flows equal to its market price. It is a static measure and does not account for embedded option value.
* OAS: adjusts the Z-spread by removing the value of embedded options, yielding a dynamic, model-dependent measure that reflects option-related risks (e.g., prepayment or call risk).
* Use Z-spread for option-free comparisons and OAS when embedded options materially affect cash flows.

Example: mortgage-backed securities (MBS)
* MBS include embedded prepayment options because homeowners can refinance or repay mortgages early.
* When rates fall, prepayment risk rises—borrowers refinance, shortening cash flows to MBS holders and reducing expected yield.
* OAS quantifies the spread investors require after accounting for expected prepayments. A larger OAS indicates higher compensation for risks (credit, liquidity, model risk) beyond option-related effects.

Practical uses and cautions
* Uses:
* Relative-value comparisons among securities with differing option features.
* Assessing compensation for non-option risks once option value is removed.
* Scenario and sensitivity analysis (rate shocks, different prepayment speeds).
* Cautions:
* OAS is model-dependent—results vary with interest-rate models, prepayment assumptions and volatility inputs.
* Illiquid markets and wide bid-ask spreads can distort implied OAS.
* Complement OAS analysis with credit research, liquidity assessment and stress testing.

Bottom line
OAS is a vital tool for valuing and comparing fixed-income securities that include embedded options. By adjusting spreads for option value and modeling interest-rate and prepayment variability, OAS provides a clearer view of the compensation for non-option risks. However, because it relies on models and assumptions, OAS should be used alongside other analyses and sensitivity checks.