Bond ETFs are unlike equity ETFs. Equity ETFs like to be broad market and passive. The broader the market and more passive they are, the more assets they tend to gather. But bond ETFs tend to go the other way. That is, those that target stratified niches and avoid pure passivity tend to gather the money. There is a good reason for that, argues Gareth Parker, chief index officer of Moorgate Benchmarks: passive bond indexes can be off-putting.

ETF Stream: Market cap weighted indices can be pretty simple to build. I’ve seen 22-year-olds with two months’ experience make them in spreadsheets. But building fixed income indices are a bit more challenging. What are the differences in index build between market weighted equity and bond indices?

Gareth Parker: The main challenge over the past few years is how you define for market cap in bonds. The closest you can get to a direct equivalent is issuance, i.e. the total dollar value issued. The problem there is that it often leads to most heavily weighting the most undesirable corporates and sovereigns. (Editors note: this is often called the “bums problem”). So if you take a European market weighted bond index, you get more Greece than Germany and more Italy than Scandinavia. Who wants that?

But it also makes it easier for active managers to outperform purely passive bond indices. If you’re an active manager, and you want to outperform that index, then obviously just avoid Greek bonds. In equity things are different. It’s really hard to beat market weighted indices, especially the broad ones. Years of evidence shows that. In bonds, lots of people can outperform because most of them are market cap weighted.

This has led index developers to head off in other directions and figure out how to weight them. I get the impression there isn’t a clearly agreed replacement for doing it by issuance at this stage.

There are alternative suggestions: issuer caps, equal weight, or GDP weight. But until there is agreement on how to do it or what the standard way to index it there’s always going to be this problem.

So there are strong arguments for smart beta approaches when building bond indices?

Yes. There is a good reason to think smart beta is a good way to go. But very few people have made that jump to passive in fixed income. So while potentially promising, smart beta in fixed income could be a while away.

Critics argue that the bums problem is fake. They say market weighting works the same for debt and equity. That is, the market determines how debt and equity a company can issue. The market then sets the cost of capital for both. There’s no distinction.

Maybe. But I don’t think fund buyers have the same understanding of what this involves in debt as in equities. When people buy market weighted equity indices they understand that they’re buying lots of large companies. They also understand that those companies are large for good reasons: your Microsofts of the world make a lot of money.

But do people understand that if they buy a bond index, your most heavily weighted issuers aren’t Microsoft and Google – which are big for good reasons. Rather they’re Greece, Portugal, that have a big footprint in that market for bad reasons? I’m not sure.

How easy is it to build a smart beta bond index? I know they can be hard for equity: when MSCI first launched a value index they had to pull it because it kept picking all the wrong stocks. Meanwhile, Vanguard has opted for actively managed ETFs for its quant products, thinking indices aren’t up to the job.

They’re easy enough to build. The difficulty to me is a lack of understanding between the index provider and ETF issuer as to what the index is doing and whether it’s doing what it’s meant to.

When things have failed in the past its often because index buyers haven’t done enough due diligence on the risks involved in smart beta strategies. These risks include a tendency towards significant underperformance at certain stages of the market cycle although generally outperforming over the full cycle.

Where do you think the potential for innovation in smart beta bond indices is? For me, it’s most obviously on the corporate debt side.

I agree. You’ve got a bigger universe of entities and much more data. If I had a lot of data, I would be thinking of smart beta for corporate debt where I can use the same sorts of strategies and use the same data we have for equity in the fixed income side. So, I can put an ESG screen and use company fundamentals for the weighting methodology. It doesn’t solve the problem at the sovereign debt side.

Am I correct in thinking we’ve seen less insourcing and self-indexing for bond index funds?

We’ve absolutely seen less of that. There is less indexing full stop as more people think active works. And to the extent that active outperforms the standard indices they’re exactly right as the standard indices are dominated by poor performing countries. Most people start in fixed income with the assumption they’ll have to build it themselves.

Gareth Parker is chief index officer of Moorgate Benchmarks

This article first appeared in the Q3 2019 edition of our new publication, Beyond Beta. To receive a full copy, click here.