In recent years at 7IM, our ETF exposure had been whittled away to the point that it was close to zero. We found that there were alternatives to ETFs like futures or regular tracking funds. Post-COVID-19 though, the liquidity and diversity of ETFs allowed us to take advantage of a number of opportunities that presented themselves. One of the opportunities that we took was to buy bank-issued contingently convertible bonds or CoCo bonds or more precisely the Invesco AT1 Capital Bonds UCITS ETF, which we bought in July.
Immediately after the COVID-19 crisis in March, we looked for opportunities and we bought into high yield bonds and US investment grade bonds, both using ETFs. We were comfortable with both positions, but they both rallied strongly but by June we worried that the spread over Treasuries on our high yield bond ETFs was not sufficient to compensate us for the defaults we were like to encounter going forward. In particular, we were concerned with the high number of single B names in the index and the probable losses there.
We, therefore, started to look around for something that had a similar profile, but that might be more rewarding. We looked at bank coco bonds or AT1s. These have better credit ratings than high yield bonds, on average they are BBB, and yet they gave a slightly higher yield.
One of the things that concerned us was that we had been down this road before. We had stepped out of our core bond market into a niche bond markets, only to find that the niche bond market suffered greater drawdowns in a crisis. You thought you were doing fine, but then a crisis came along and you had your head handed to you on a platter. Some people quite rightly characterise this as “picking pennies up in front of a steamroller”.
We, therefore, said that despite the superior credit rating of CoCos, we require a higher return of at least 1% a year to compensate us for those sharper drawdowns. We said 1% a year, because in the March drawdown AT1s had sold off more than high yield by about 5% and we believe that the next sell-off was about 5 years away; so 5% divided by 5 years.
Our scenario work showed that we got more than our threshold yield premium with CoCos in all scenarios except a very rapid recovery. In a rapid recovery, the CoCos still beat high yield bonds, but not quite by 1%. With our very pessimistic assumptions, which affect both high yield and bank CoCo issuers, we found that the CoCos still did relatively well even though five banks might have to convert their CoCos to equity and assuming that equity became valueless.
Ironically, our outlook has changed since we bought the Invesco ATI Capital Bond ETF and we do now expect a rapid recovery from the COVID-19 crisis, which does suggest that we will not quite make the large premium over high yield bonds we thought we would. That does not seem to be such a penalty if we do get a return to normality.
Peter Sleep is senior investment manager at 7IM