The collapse of Credit Suisse and Silicon Valley Bank in 2023 has spurred a rethink on how banks should manage their capitalisation.
The Swiss regulator's forced writedown to zero of Credit Suisse’s additional tier one capital bonds in March 2023 was quickly followed by statements of support for the asset class from European and UK regulators. But some 21 months later, the Australian Prudential Regulation Authority has chosen a different path.
In stark contrast to almost every other major regulatory regime, Apra has decided the best thing to do is completely phase out the AT1 asset class — which at June 2023 had some A$42.5bn ($27bn) outstanding.
Instead, Australian banks must hold 25bp more common equity tier one (CET1) capital — real equity — as well as 125bp more tier two capital.
Apra’s big decision is a bad one for tier two bonds, which already do a lot of heavy lifting in the Aussie bank capital stack.
Of course, Australian regulators have a history of going their own way.
In 2019, for instance, Apra mandated that Australia’s domestic systemically important banks (D-Sibs) — ANZ, Commonwealth Bank of Australia, National Australia Bank and Westpac — would have to meet their total loss-absorbing capital (TLAC) requirements with tier two bonds.
Other major markets met the TLAC need by creating a new bail-in-able senior debt instrument. Instead, Australia piled more responsibility on the tier two layer.
At the time, a further A$50bn ($32bn) of tier two was expected to be needed. This build-out passed without a hitch — Australia’s four major banks now meet this target.
However, Apra’s latest policy swerve will push tier two — now a quasi-senior product — back towards its roots as a capital product.
That is likely to put upward pressure on Aussie banks’ funding costs.
Right now, Australian tier two spreads are tight. Many of the notes in euros close to senior unsecured bonds from similarly rated countries.
The most recent Australian tier two in euros was Commonwealth Bank of Australia’s €1bn 10 year non-call five green note, issued in late May. The 4.266% June 2034 non-call June 2029 bond was priced at 135bp over mid-swaps — on Tuesday it was bid at an I-spread of 120bp, according to Tradeweb.
This level is not far off where European banks’ higher rated senior unsecured notes are trading.
Crédit Mutuel Arkéa’s €500m October 2034 green senior preferred note — the most recent senior euro bond at a similar tenor to CBA's, according to GlobalCapital’s Primary Market Monitor — is bid at 106bp, Tradeweb shows.
There is only 15bp between them, but Arkéa’s note is rated two notches higher than CBA's by Moody’s and three notches higher by Fitch, at Aa3/AA-.
Similarly, ANZ priced a €1bn 5.101% 10 year non-call five tier two in January 2023 at 215bp. On Tuesday, it was bid at an I-spread of 114bp.
These levels are far inside where European tier two bonds trade. The most recent issue, De Volksbank’s Baa2/BBB rated €500m November 2035 non-call November 2030, trades at 180bp, for instance.
Even better rated, longer dated notes like Danske’s €500m November 2036 non-call November 2031 or Handelsbanken’s €500m deal with the same tenor and call date trade wider — 147bp and 128bp, respectively.
These tight tier two spreads are a boon for Australian banks that rely on the product to build up their TLAC buffers.
However, if Aussie tier two must now take on part of the riskier AT1 load, it is only natural to assume that these spreads will widen, affecting a wider band of the capital structure than AT1 previously occupied.
Aussie banks will start replacing AT1 with tier two from 2027, while at the same time refinancing their maturing and called TLAC paper.
Between them, the four Aussie majors have just under A$109bn of tier two debt outstanding, of which A$21.1bn needs refinancing before 2027.
For instance, Westpac has A$2.9bn of tier two paper maturing or up for call for the first time in 2027. Now, in the same year, it will also have to refinance a further A$3.5bn of obsolescent AT1.
Of course, Apra is right to wean Australia's retail investor base off risky AT1s.
But perhaps there is a better way to reform the most deeply subordinated part of the capital stack than piling extra strain on already stressed tier two.