- Due to the mechanics of bond indexes, taking passive fixed income exposure can be problematic
- Additionally, there are good arguments for why active managers may be able to add value in debt markets
- However, care must be taken when evaluating active strategies; not all outperformance is “alpha”
- Separating skill vs. risk-taking is increasingly important in the current yield environment
The Problem with Passive Fixed Income
Selecting an appropriate passive benchmark can be challenging no matter the asset class, but fixed income indexes can pose some special challenges.
Most benchmarks weight securities in proportion to their market value. This may be okay in equities, but in fixed income it means that the more debt an issuer has outstanding, the larger weight it will tend to hold in the benchmark. This is a peculiar feature, since it results in aggressive debt issuers being rewarded with larger weights relative to more conservative issuers. There’s no reason to believe that the issuers with the most debt outstanding offer the best risk-adjusted returns.
An issuer’s idiosyncratic yield compression may likewise cause it to receive an increased index weight. In other words, the more expensive a company’s bonds get, the more weight they could potentially receive. This further magnifies the counter-intuitive structure of passive indexes, inducing investors to buy more of bonds that might potentially merit sale due to their appreciation.
These two effects have likely contributed to the remarkable “tale of two cities” witnessed in sovereign debt markets over the past few years, wherein near-mechanical buying of major issuers (such the US, Japan and EU countries) has led to continued spread compression and negative yields, while at the same time emerging markets have faced increased real borrowing costs.
Beating the Benchmark
Given the inefficiencies just discussed, it may be possible to generate systematic outperformance simply by reweighting the benchmark. But the case for active management does not end there. Two notable additional justifications include:
- The debt market is much larger than the equity market, and there are many more individual fixed income securities outstanding. There may thus be a greater potential to identify pricing inefficiencies, especially given the increasing concentration of major indices toward a relatively small number of large issuers.
- The presence of large forced buyers in certain areas of the market – such as risk capital requirements which compel regulated financial institutions to hold large quantities of AAA-rated sovereign bonds – may create distortions which can be capitalized upon by active investors.
Benchmark Choice and Excess Return
While a strong case can be made for utilizing active management in fixed income, it does not follow that excess returns can be automatically attributed to manager skill.
Because performance statistics such as alpha and excess return are relative measures, they’re highly dependent upon the choice of benchmark index. For example, consider the annual total return of a hypothetical US “core” bond fund, whose manager made some discretionary allocations to issuers with higher credit risk:
|Active Core Bond Fund||8.55%|
|Core Bond Index||7.70%|
|High Yield Index||9.22%|
This strategy outperformed the core bond index by 0.85%, but underperformed the high yield index by 0.67%. As a result, this manager would rank well relative to the core category, but poorly relative to high yield managers. The question of which of these comparisons is appropriate – or whether some third benchmark would be more accurate – depends on many factors, such as the manager’s stated investment strategy and universe, and whether the lower-quality bias is stable or fluctuates over time.
Fixed Income Performance Attribution
Prospective investors must measure to what extent excess returns are attributable to structural biases in the manager’s portfolio – not only in terms of credit quality, but also other factors such as duration and curve positioning. Adding these factors as well as sector and/or country tilts to an attribution model usually results in most if not all alpha disappearing as the variance is explained away, leaving little residual variation.
The key judgment that must then be made in the due diligence process is to what extent a manager should receive credit for these factors: for example, is the manager skilled at taking credit risk, or has he simply been lucky over the strategy’s history with his riskier bets? Often, a strategy’s track record will not be sufficiently long to definitively conclude one way or the other, given that default events are relatively rare in normal times even among junk bonds.
The general view taken by Empirically is that managers should receive credit for successful factor timing, but not for one-way factor bets. The reason is that constant factor bets come at the cost of bearing certain risks, and can usually be replicated at a lower cost. By contrast, factor timing skill is valuable because it cannot usually be reliably replicated at lower cost.
To provide a concrete simplified example, a manager of a core bond fund who juices returns by allocating an average of 10% of his portfolio to high yield bonds will generally not receive credit for the excess returns attributable to the high yield component. However, if the manager still beats a custom benchmark of 90% core bond index and 10% high yield index – either by presciently increasing and decreasing the high yield weight or by selecting better-than-average high yield securities – will receive credit for that difference as “alpha”.
Separating true skill from factor tilts is of heightened importance in the current environment. Yields are near all-time lows, capping further price appreciation potential and reducing the compensation provided for potential defaults. Likewise, volatility has also reduced dramatically, reducing trading opportunities. Therefore, asset owners should measure their active fixed income managers not only against flawed benchmark indexes, but also against investable smart beta strategies and replicating portfolios, which may be able to provide a similar return profile with greater transparency and lower cost.
Author Information: Jordan Boslego is a Partner at Empirically.