Call on me... emerging market bank capital and calling CoCos

The structure of banks’ subordinated debt (for example, contingent convertible bonds, or CoCos for short) developed quickly after the Global Financial Crisis.

CoCo structures generally offer a higher yield than traditional senior unsecured bonds issued by banks and corporates, but are more nuanced and may entail additional risks.

An Additional Tier 1 (AT1) CoCo typically has a perpetual non-call structure (for example, Perpetual Non-Call 5 Year, abbreviated to PerpNC5), while a Tier 2 (T2) will usually have a dated non-call structure (for example, a 10NC5), meaning it can be called at par after five years.

If the bank fails to call its AT1 or T2 CoCos at their first call dates, their coupons will reset to an underlying rate (for example, the prevailing five-year treasury) plus the initial credit spread. The initial credit spread is determined at the outset of the issuance.

In the first instance, market convention has been to price these instruments off the five-year point of the bank’s senior yield curve.

For example, let’s say a bank has a five-year senior USD bond outstanding that yields 5.85%.

This equates to a spread over the US 5-Year Treasury of c.180 basis points (bps)

Market convention

When this bank issues a 10NC5 T2 bond the market convention has generally been to price this instrument at a certain spread or multiple over the five-year senior bond spread to reflect the bond’s subordination. As such, the market would expect the bond to be called on the first call date. If the assumption was that the bond was not to be called after five years, the market would price the bond off the ten-year point of the curve.

When you explain this concept, a common response is that the curve between the five- and ten-year point of the US Treasury curve is flat (or inverted), so there would be little difference in the overall yield of the bond at the time of issuance.

However, the credit spread between a five-year senior and a ten-year senior bond is not flat. In some emerging markets, the bank credit curve is worth c.50bps for a one-year extension. A 10NC5 T2 that is priced off the five-year point and not called should have arguably built in a further 250bps to the spread, plus the associated multiple at the time of issuance.

To date, the vast majority of T2s (and AT1s) have been called at the first call date, so the market convention has given bank issuers the benefit of the doubt.

Furthermore, a Tier 2 CoCo typically loses capital treatment in the last five years of its life, so it usually makes sense for a bank to call and replace this bond as it essentially becomes expensive senior debt for a bank.

Reputation versus economics

However, the discussion has shifted as market volatility has increased, and in some cases it has become uneconomic for issuers to call their outstanding CoCos. This is because the cost of issuing a replacement instrument would be greater than leaving their current issue outstanding at the new reset rate.

In other words, the market convention has allowed non-callers to price their instruments off the five-year point of the curve, when they should have been priced off the ten-year point for a 10NC5 and perpetuity for an AT1.

 Once bitten, twice shy

Issuers generally understand the market dynamics here; this is where reputation comes into play. If they do not call their instruments, when they come back to the market after a non-call event investors should ask for the new issue to be priced off the appropriate point of the curve (i.e. the ten-year point for a 10NC5 or perpetuity for an AT1).

Over time, we believe the market will likely differentiate between callers and non-callers, with the non-callers having to pay more to access the market. Arguably, non-callers may not be able to issue with the same call structure again, instead opting for a Tier 2 bullet, for example. Furthermore, one bank’s decision may affect investor confidence about future issuance from the wider banking sector.

In our view, subordinated debt can provide an additional source of alpha for our portfolios, but we also understand it is a specialist topic. To add further complexity to the subject, AT1 and T2 structures are not uniform across the globe, hence we rely on our experienced bank analysts when assessing an investment in the asset class.