Arnott: Buy Treasuries Today, Short Them Tomorrow

February 18, 2010

Rob Arnott discusses the flaws of traditional bond indexes, and forecasts a 'once in a generation' trading opportunity in the Treasury markets.

Rob Arnott, chairman of Research Affiliates, thinks most bond indexes are dumb. According to Arnott, the flaw in bond index construction—whereby they assign higher weights to companies or countries with higher debt—is self-evident.

In an exclusive interview with, Arnott examines the flaws in traditional bond indexes and explains how his newly launched benchmarks promise to fix those flaws. He also looks at the current state of the bond markets, and forecasts a “once in a generation” trading opportunity in the Treasuries market. ( What’s wrong with existing bond indexes?

Rob Arnott, chairman, Research Affiliates (Arnott): There are two problems with existing bond indexes. First, they include a lot of bonds that don’t trade, so you simply cannot replicate most bond indexes. The original bond indexes were created to serve as a pricing mechanism for thinly trade bonds in the broker dealer community. They were not really intended as investment vehicles.

The second thing that’s wrong with them is, why on earth would you want to own more bonds from a company just because it has more debt? The fallacy of that is self-evident. If they’re so fundamentally flawed, how did they become so popular?

Arnott: Because, as with cap-weighted stock indexes, they do in fact represent the broad market. If you want to own the broad market and make no active bets, you have to use cap-weighted indexes. That’s as true for bonds as it is for stocks.

Also, value-added is a zero-sum game. If I pick the better bonds, someone else has to own the worse bonds. These are the same arguments used for cap-weighted stock indexes, and they are just as true in bonds as they are in stocks.

That said, if the market gets prices wrong—if the market thinks a company is more or less creditworthy than future events reveal them be—investors are going to own too much of the bonds that are not creditworthy and too little of the bonds that are creditworthy. Just as with cap weighting, if the prices are wrong, you wind up emphasizing the wrong assets. The argument, then, is almost identical to the argument for Fundamentally Weighted equity indexes?

Arnott: Yes, except in the bond world, it’s much clearer and more self-evident. In stocks, you could argue that cap-weighting is a form of popularity weighting. I would say that the flaw in owning more of a company because it’s more popular is self-evident. But in bonds, the argument is more striking: The more a company owes, the more you own of its debt in a cap-weighted index.

That’s the reason bond indexing has never caught on to the extent that stock indexing has: The structural flaw in cap weighting in bonds is even more obvious than in stocks. On the surface, it would seem like your Fundamentally Weighted bond indexes would outperform when investors are punishing credit risk, and under-perform when risky credits are rallying. Is that a fair assessment?

Arnott: Not necessarily. Bigger companies are sometimes out of favor. Ford and GM are big companies, so their fundamental weight may not be small. It just wouldn’t be as big as it is in a cap-weighted index right before they run into trouble.

We find that fundamental bond indexes outperform in bond bull markets and outperform even more in bear markets.



Find your next ETF

Reset All