Key Rate Duration: Construction and Application in ALM

Key rate duration quantifies the sensitivity of a fixed income security’s or portfolio’s value to a shift in the yield at a single specified maturity point on the yield curve, with all other points held constant. It is derived by revaluing the instrument after perturbing only the rate at that key maturity-e.g., 2-year, 5-year, or 10-year-while leaving the rest of the curve intact. This isolates non-parallel, curve-shaped risk that effective or modified duration, which assumes a parallel shift, cannot capture.

In ALM, key rate duration is used to build key rate duration gap measures-analogs to the standard duration gap-that reflect mismatches between asset and liability sensitivities at specific maturities. These measures support targeted hedging, cash flow matching, and risk attribution, especially where embedded options (e.g., prepayments, calls, lapses) create asymmetric sensitivities across the curve. As documented by the Society of Actuaries and J.P. Morgan Asset Management, key rate duration is a core component of interest rate risk attribution frameworks, alongside DV01 and cash flow sufficiency.

Key rate duration is constructed through a revaluation-based approach that requires a calibrated yield curve and a pricing model capable of handling non-parallel shifts. The steps are:

• Select a set of key maturities (e.g., 1y, 2y, 3y, 5y, 7y, 10y, 15y, 20y, 30y) that span the relevant part of the yield curve for the portfolio.

• For each key maturity k, shift the zero-coupon rate at k by +1 bp (or −1 bp), while holding all other zero rates unchanged.

• Revalue the security or portfolio under the shifted curve.

• Compute the key rate duration at k as:

  \[ \text{KRD}_k = -\frac{\Delta V / V_0}{\Delta y_k} \]

  where ΔV is the price change, V₀ is the original value, and Δy_k = 0.0001 (1 bp).

• Repeat for all key maturities to obtain a vector of key rate durations.

This procedure yields a key rate duration profile-a vector of sensitivities that, summed across all key maturities, approximates the effective duration (a simultaneous 1 bp shift at every point is a parallel shift). The profile’s value is that it also prices non-parallel moves, which a single duration number cannot.

In ALM, key rate duration enables institutions to:

• Construct key rate duration gap measures by subtracting liability key rate durations from asset key rate durations at each maturity, exposing mismatches in funding needs and cash flow timing.
• Design targeted hedging strategies-e.g., using interest rate swaps, basis swaps, or sector-specific bonds-to neutralize exposure at specific curve points rather than across the entire curve.
• Attribute changes in economic value or earnings to curve movements at specific maturities, supporting scenario analysis and stress testing.

For example, an insurer with long-duration liabilities (e.g., annuity guarantees) may find its assets less sensitive beyond the 10-year point due to callability or prepayment features. A key rate duration profile would reveal this mismatch at 15y and 20y, prompting a hedge using long-dated swaps or bullet bonds.

As emphasized in SOA and J.P. Morgan research, this granularity is essential for managing embedded optionality and aligning asset and liability sensitivities in volatile or steepening/flatening environments.

Key rate duration gap extends the classic duration gap framework by decomposing mismatch risk across maturities. For each key maturity k, the gap is:

where KRD_A,k and KRD_L,k are the asset and liability key rate durations at k. A non-zero gap implies sensitivity to a 1 bp shift at k, even if parallel duration gaps are neutral.

This decomposition supports risk attribution in ALM reporting. As noted by the SOA, key rate duration is a standard factor in interest rate risk attribution, alongside DV01 and cash flow sufficiency. It enables actuaries and risk managers to:

• Identify which curve segments drive economic value changes.
• Prioritize hedging actions (e.g., adjust duration at 5y vs. 30y).
• Monitor the impact of dynamic hedging strategies over time.

Key rate duration has important limitations that affect its reliability and implementation:

• Curve interpolation assumptions: The profile depends on how the yield curve is interpolated between key points; inconsistent interpolation can distort sensitivities.
• Non-linear instruments: For securities with path-dependent options (e.g., mortgage-backed securities), key rate duration may not fully capture convexity effects or rebalancing behavior.
• Computational cost: Calculating full key rate duration profiles requires repeated revaluations and a robust pricing engine-often necessitating integration with ALM or risk platforms (e.g., SAS, Moody’s ALM solutions).
• Interpretation complexity: A flat key rate duration profile does not guarantee immunity to curve shifts; it only neutralizes sensitivity to independent shifts at the chosen points.

In practice, institutions combine key rate duration with other tools-such as scenario-based cash flow testing and full revaluation under parallel and non-parallel curve shifts-to ensure robust ALM oversight.