My last few entries highlighted some recent AQR papers with hopefully some interesting commentary. I know it’s a little weak to do that again so soon, but some of the papers are just too cool not to opine on. I have a particular weakness for pricking the bubble of conventional wisdom, and, much geekier, for arguments that hinge on the subtleties and great value of true diversification. This one has got it all, and I couldn’t resist discussing it.
There is a common belief that goes something like this: “Sure active stock picking doesn’t work, but in fixed income, active management really shines.” Various reasons are given for this including the notion that fixed income is, for some reason, a less efficient market or that the benchmarks are worse investments. In other words, fixed income is ripe for the value-add that comes from traditional active management.
The problem is that any perceived alpha has been, at least to a great degree, due to a passive long-term overweight of credit.1 If we found equity managers out-performed largely because they were strategically, not tactically, higher beta than their benchmarks, would you get excited and pay a lot for that? I hope not. For fixed income managers, the credit exposure doesn’t necessarily eliminate all of their alpha, but once you account for the credit exposure, fees and other simple factor exposures (for example, short volatility exposure) there may not be much left.
It gets worse. The reason most invest in fixed income is because it’s a positive expected return asset that offers diversification from equity risk, which dominates most investors’ portfolios. Well, credit is the part of fixed income that is highly correlated with equities. As a result, not only is there less alpha than people think, the whole premise for investing in fixed income is compromised on account of the average2 active manager’s passive overweight to credit as it non-trivially raises fixed income’s correlation to equities.
So, once again fans of active management are just wrong, and Jack Bogle is just right.
Except… Yeah, except for the stuff AQR and other quants believe in (I know, there he goes again). We do find, again much like in equities, expensive traditional active fixed income management is a bust; however, the same factors we find effective elsewhere (value, momentum, carry, defensive/quality)3 show up in bonds. So, not surprisingly, we think there is hope, but it’s not because “active management works in bonds,” but because “the same basic things that work everywhere also work in bonds.” Basically, don’t go traditional “active” in fixed income, go factor (and make sure the fees fit that!). That’s your self-serving message from me for the day.
[1] “Passive” is important here. We don’t mean they’ve added alpha from going long credit (or other exposures highly correlated with credit) when it was particularly smart to do so. We mean they have a persistent long exposure to credit. You don’t need to pay active fees for one-time “take more risk” decisions. [2] I mention the “average” here to give an overall sense of what managers are doing. Obviously, that doesn’t mean that every active portfolio shares this characteristic, and we're talking averages, so, of course, there may be many good ones, but the paper goes further to show that the persistent long credit exposure is pretty pervasive across the universe. Please see the paper for further description of average active manager. [3] See, for example, “Investing with Style,” Journal of Investment Management (2015) by Asness, Ilmanen, Israel, Moskowitz.
Article by AQR
