Mean-Variance Construction of a Corporate Bond Portfolio

Summary

The document considers whether an actively constructed corporate bond portfolio can target a yield premium over a benchmark while still seeking to outperform it. The proposed starting point is to collect historical returns for corporate bonds and form a diversified portfolio with exposure to the benchmark calibrated through beta. It suggests using mean-variance optimization to select bond weights under a return target relative to the benchmark.

The answer notes that the return differential could come from either alpha or greater beta exposure, and argues that a diversified portfolio’s excess return would likely be driven mainly by beta risk. It does not provide an optimization specification, covariance estimates, bond-level constraints, or evidence that the proposed yield target would translate into benchmark outperformance. Yield and expected return are distinct, and implementation would also depend on credit, duration, liquidity, concentration, and transaction-cost constraints, none of which the brief response analyzes.

Key ideas

• Historical bond returns can inform portfolio weights and benchmark-relative beta estimates.
• Mean-variance optimization is proposed as a way to choose diversified corporate bond weights.
• A yield or return premium may reflect alpha, additional benchmark beta, or both.
• The response expects beta risk to explain much of the return of a diversified portfolio.
• The proposal gives no empirical validation or detailed treatment of credit and implementation constraints.

Tags

# Constructing a Corporate Bond portfolio?


# Constructing a Corporate Bond portfolio?












Is it possible to create a corporate bond portfolio such that its yield is 100bps higher that its benchmark, while still outperforming the benchmark (BBG Corporate bond Index)? I guess my question is if I am looking at constructing a higher yielding portfolio and still trying to beat my benchmark, what are some factors, constraints etc I should consider and how would I adjust them. Assuming I am starting with a 100M AUM and implementing from scratch without the use of ETFs.

## Answer by phdstudent (score 1)

https://quant.stackexchange.com/a/78508

The best thing you should do is to download returns of a corporate bond portfolio, and then create a diversified bond portfolio whose beta against the benchmark is such that it yields an expected return that is 100bps higher.

One way of doing this is just to run a Mean-Variance problem, to choose weights of corporate bonds such that the average return you achieve is 100bps higher than the average return of the benchmark. This return could come from either alpha or beta against the benchmark, but if it is truly a well diversified bond portfolio, most of it should be beta risk.

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.