How do you construct a custom fixed-income benchmark by blending standard indices?

Custom benchmark construction involves blending two or more standard indices (or sub-indices) in proportions that replicate the risk characteristics of the portfolio mandate. This approach is common when no single published index adequately represents the manager's opportunity set.\n\nConstruction Process:\n\n1. Define the mandate's target risk exposures (duration, credit quality, sector weights, currency)\n2. Identify candidate sub-indices that cover the relevant segments\n3. Solve for blend weights that match the target characteristics\n4. Document the methodology and rebalancing rules\n5. Backtest to verify the blend behaves as expected\n\nWorked Example:\n\nStonebridge Endowment needs a benchmark for a core-plus fixed-income mandate with:\n- Target duration: 5.5 years\n- Investment-grade bias with 15% high-yield allocation\n- 10% emerging market debt exposure\n\n| Sub-Index | Duration | Weight | Duration Contribution |\n|---|---|---|---|\n| Bloomberg US Aggregate | 6.2 | 60% | 3.72 |\n| ICE BofA US High Yield | 3.8 | 15% | 0.57 |\n| Bloomberg EM USD Aggregate | 6.8 | 10% | 0.68 |\n| Bloomberg 1-3 Yr Treasury | 1.9 | 15% | 0.29 |\n| Blended Benchmark | | 100% | 5.26 |\n\nThe blended duration of 5.26 is close to the 5.5 target. Adjusting the 1-3 Year Treasury allocation down to 10% and increasing the Aggregate to 65% would bring the duration closer:\n\nRevised: 65% x 6.2 + 15% x 3.8 + 10% x 6.8 + 10% x 1.9 = 4.03 + 0.57 + 0.68 + 0.19 = 5.47 (much closer to 5.5).\n\nRebalancing Considerations:\n\nThe blend weights should be rebalanced at a defined frequency (monthly or quarterly). Drift between rebalancing dates creates tracking differences. Calendar rebalancing is simpler and more transparent than threshold-based approaches.\n\nCommon Pitfalls:\n\n- Overlap: Some sub-indices share constituents (e.g., investment-grade corporates appear in both the Aggregate and stand-alone credit indices), leading to double-counting\n- Stale pricing: Sub-indices with illiquid bonds may have stale prices, creating artificial tracking error\n- Survivorship bias: Backtested blends look cleaner than live performance because index composition data is revised retroactively\n- Complexity: More components increase transparency cost and make it harder for stakeholders to understand the benchmark\n\nDocumentation Requirements:\n\nThe Investment Policy Statement should specify the exact sub-indices, blend weights, rebalancing frequency, and the rationale for each choice. This ensures the benchmark remains specified in advance and unambiguous.\n\nPractice custom benchmark construction in our CFA Fixed Income question bank.