Coincidentally while I was noodling about bond indexes on Friday so was Goldman credit strategy research. Here I will show you a chart they made:
So what this is is performance of ETF, index-y bonds versus performance of otherwise similar but non-ETF, non-indexy bonds. Goldman took the bonds that are in the iBoxx US liquid investment grade index (IBOXIG), which underlies the $21 billion LQD ETF, and compared them to other bonds that are - take their word for it - similar, but not in the liquid index (but in the broader ICPRDOV index), and saw that the ETF-y ones outperformed the non-ETF-y ones - by about ~4% of price / 60bps of spread over the last three years. Here are some more words on the word you have to take for it:
[W]e use a factor approach and construct hypothetical portfolios, one with only ETF bonds and the other with non-ETF bonds, that have otherwise identical compositions. These portfolios are constructed using the constituents of the iBoxx USD domestic index of which the LQD ETF is a sub-index.
One immediate challenge that such an exercise raises is that our 'ETF' factor might be hard to disentangle from a 'liquidity' factor. In the extreme case where the ETF only include onthe-run bonds, the ETF factor might just be an artefact for the on-the-run/off-the-run premium. Partly to address this issue, we include an “on-the-run” dummy variable in our regressions.
They regress this spread versus ETF funds flows and find about a 24% R2, meaning that the outperformance is partly explained by just the amount of money flowing into bond ETFs helping the bonds in those ETFs. Conversely, or conversely-ish, though, they note the ETF bonds underperformed during the second half of 2011 "as the European crisis intensified only to resume their outperformance during the LTRO rally and then more recently following Draghi’s speech in July."
Note that this is about relative performance: it's not just that bonds weakened when Europe did, it's that popular/ETF/indexy bonds weakened more than unpopular/non-ETF/quieter bonds. Bonds that are in the big ETF aren't simply better; they are more intense: they react more dramatically to economic news than the bonds that aren't. Some of this is perhaps liquidity, but it may also be dissociation between some bonds that have been drafted into service as proxies for macroeconomic theses, and other bonds that are just bonds.1
One other, somewhat related thing that we talked about on Friday was that the new CDX high-yield credit derivative index will include some names on which there is currently no single-name CDS. In a sense that makes CDS in those names sort-of pure proxies for macroeconomic theses: nobody wants to buy CDS on them as companies, but maybe somebody will want to buy CDS on them as index components. The FT today gets antsy about this:
Wall Street financial engineers have devised a new way to combat declining trading in the credit derivatives market – they are revamping an index to add financial instruments that do not exist. ... This week, the index provider, Markit, will cross a Rubicon and begin to include three companies in its North American high-yield CDX index for which no bank is offering a CDS.
Markit and derivatives traders hope the addition of CIT Group, Charter Communications and Calpine Corp will force banks to launch CDS on the three companies. ...
The prices of Markit’s CDX indices are set at a daily fixing using quotes provided by dealers, excluding outlier quotes, a process the company says mirrors the one used to calculate Libor – the interbank lending rate that has become the subject of a slew of regulatory inquiries into possible manipulation.
Deepak Agnani, head of US credit indices at Markit, said the process for choosing additional names was transparent and conducted in consultation with dealers. “We have picked the three companies that have the highest amount of debt outstanding,” he said. “They are names that should be put in focus by the CDS community.”
So, one, Markit, did you really say "oh our process is just like Libor, no problem"? Could maybe do with some tightening of the message there.
But two, that last quote is right, right? CIT and Charter and Calpine are weird companies with big debt complexes that don't have traded CDS for as far as I can tell mostly technical recent-bankruptcy reasons. They do seem to have been included because it's a good guess that one day they'll grow up to have real single-name CDS trading - and not because they're good pure macro-thesis proxies.
If I were running the index stupidly I'd have picked, like, Groupon or something. Sure they're unrated and have no public debt to default on so no actual human would ever buy CDS on them for Groupon-specific purposes. But if you were just using high-yield CDX to express a general view about whether, like, lots of moderately crappy U.S. companies are going to go bankrupt, wouldn't they be a neat addition to the index?
1. This isn't due entirely to the LQD ETF, or all ETFs: funds flows only have a 24% r-squared to the price differences, for one thing; for another bond ETFs still aren't that big. From Goldman:
According to data from Lipper, IG and HY mutual funds manage roughly $1.3tr and $279bn, respectively, up from $525bn and $126bn at the end of 2007. Even when scaled by the overall size of the US corporate bond market, these figures suggest a bigger ownership share of credit mutual funds. Perhaps more impressive is the growth of corporate bond ETFs, whose size for IG and HY has increased to $105bn and $31bn, respectively, from just $12bn and $288mn at the end 2007.
So investment grade ETFs are still under 10% of the total IG mutual fund market. You could imagine that part of the issue is that, as we discussed Friday, some non-ETF mutual funds are committed to tracking the index in more or less explicit ways. Can't have tracking error, etc.