Collateral questions

Look what is (not) in bond ETFs

I’ve been looking at the fixed income corner of the world for a bit recently and it reminded of something I was told a few years ago about stock lending in bond funds. Because demand for gilts and other government debt as a liability match is so high some funds lend out a large proportion of their holdings and for a prolonged period. And, what’s more, the collateral lenders hold in return can be very different in nature to the headline exposure of the fund.

Trawling around for information on a fixed income fund, I stumbled on a stock-lending report. This showed that for a particular UK gilt fund 20% of the holdings by value were on loan at the time of the report, and the 12-month average on loan was over 22%.

A bit more trawling turned up a much bigger stock-lending report across a range of ETFs. This includes maximum on loan and average on loan figures so here are the top ranked by average on loan. A UK gilt fund tops this list - it has gone over 50% out on loan and averages 42% loaned out.

Across those ETFs, what is likely to held as collateral? The top 10 holdings across the funds are shown below and are a mixture of bonds and equities. US tech stocks are a major component.

Finally, here’s another gilt fund I found stock-lending information for. As you can see at the time of the report about a third of the fund’s holding by AUM were on loan and the 12-month average on loan was over a quarter of the total.

For this fund I was also able to find the collateral held, and this I found really interesting. The top holding, valued at 8.5% of the AUM of the fund, was a German equity rather than a UK bond.

All of this bears out the point that at any time the bond ETF you are invested in might hold a load of securities that aren’t either bonds, or issued in the country you sought exposure to.

Given that the function of many such funds is to provide investors with a return that matches the market return of the asset class, I guess many investors won’t care as long as it does that. But the fact that for much of the year the funds hold things that are completely different to the what it says on the tin is a bit odd.

Presumably it does matter somewhat in terms of stewardship. Those equities held as collateral might cause some funny things to happen. What if it pushes the total holdings of the manager holding the collateral above regulatory reporting requirements? What happens to the voting rights? If engagement with the issuer is taking place who should the issuer be focused on - the investor who holds the collateral or the one who provided it?

I think of many investors’ approach to stock-lending as somewhat analogous to Airbnb. You’re allowing someone to use your asset for a price during a period where you don’t need it. Which is fine if we think bonds, equities and other securities are just lines on a spreadsheet. But now that a superstructure of ‘stewardship’ has been built over equity holdings in particular this doesn’t feel quite right to me.