Fixed income has been the fastest-growing corner of the ETF market this year, pulling in $31.5 billion in new capital through September. Those inflows worry me, for a number of reasons.
First, as I’ve written about previously, those inflows have forced some corporate bond ETFs to trade at large premiums to their net asset values. Those premiums are sustainable so long as investors continue to buy. Unfortunately, like all ETF premiums, if fund flows reverse, those premiums can collapse and turn into discounts, and investors will be left holding the bag. (See related story here.)
But there’s an even more fundamental problem with corporate bond ETFs, which stems from the way their indexes are constructed. This is the elephant in the room in corporate bond indexing, and it amazes me that it is rarely discussed.
Ready? Most corporate bond indexes weight holdings based on their total debt outstanding. To put it more bluntly, the more debt a company has, the higher its weight in the index.
On the face of it, this is crazy. The idea that investors want to pile more money into the most indebted companies flies in the face of common sense. Where are the fundamental indexers when you need them?
This isn’t a new insight. Laurence Siegel of the Ford Foundation has been writing about it for years. It’s called the “bums problem”: Bond indexes tend to overweight corporate bums that run up huge levels of debt.
A look at the holdings of corporate bond ETFs like the SPDRs Barclays Capital High Yield ETF (NYSEArca: JNK) confirms these fears. Among the top 10 holdings by weight are gems like AIG, GMAC, Harrah’s and Intelsat.
I realize that the idea of junk bond indexing is to buy into notes that pay high yields, but there’s risk and then there’s RISK. Overweighting companies that issue more and more debt just seems absurd.
I’m not sure how you get around the bums problem. Some have suggested equal-weighting as a preferred methodology, but that “hands-off” approach has never appealed to me. There must be a way that combines liquidity, ability to repay and true credit quality.
As a way of measuring the market, weighting an index by debt outstanding makes sense. But as a way of investing in that market, it makes me nervous.
Index-weighting methodology isn’t the only obvious problem with bond indexes. The other big fat elephant we dance around is credit ratings.
As far as I know, all the major bond indexes use official bond ratings from S&P, Moody’s and Fitch Ratings to determine which bonds are “investment grade” and which are “high yield.”
Didn’t we learn anything from the financial crisis? Can we really trust the ratings agencies to do their job?
Here I think there is a relatively easy solution. Why not use credit default swap rates to determine credit quality rather than the “official” bond ratings? We could define the split between “investment-grade” and “high-yield” debt based on the cost of insuring against default for the next five years. Let the market make the determination, in real time, rather than the troubled credit ratings agencies operating with a lag.
If CalPERS is taking hedgies out, ETFs may be coming back in.
As valuations grow uncomfortably high, ‘quality’ ETFs makes more sense—if you can figure out just what quality means.
‘Smart beta’ almost surely means loss of more market share for active managers.
Be careful of your assumptions (and headlines!) about volatility ETFs.