Back in November 2013 they reminded investors explicitly. S&P – and other agencies - noted emerging regulation to deal with what are called "domestically systemically important banks" (DSIBS), intended to remove the implicit guarantee. But the agencies concluded new rules hadn’t altered the possibility of taxpayers’ money being used to save banks.
The current debate though is whether several waves of global banking regulation have indeed taken the taxpayer out of the equation – by making banks safer and shifting risk onto bank investors, debt and equity.
S&P's said in 2013 the Australian government's "tendency to support private-sector banks deemed to be systemically important under our bank-rating methodology" was "highly supportive" of its rating, adding that that if it did believe the government was genuinely less likely to support banks, it would cut bank ratings.
The argument is that if banks are implicitly guaranteed, that in turn leads to higher than otherwise ratings and hence, prima facie, lower cost of funding for banks because higher rated debt issuers pay less for their funds.
But the question has always been how much that implicit guarantee is worth. The Reserve Bank of Australia had a stab at working it out and that work, recently made public through Freedom of Information requests, was done in 2015.
The headline numbers, not unexpectedly seized upon by the media, were the subsidy is worth between $A1.9 billion and $A3.7 billion annually for the big four banks and Macquarie.
However, that work, done about two years ago, and as would be expected from central bank economists, was not so categoric.
Critically the researchers were transparent about data gaps, the difficulty in gauging who actually benefits from any subsidy, the lack of consistent correlations between agency ratings and funding costs as well as what happens to it over time.
“These kinds of calculations are not without their problems,” the paper says. “Bank funding costs are affected by a large number of factors, and investors do not focus solely on ratings when pricing bonds. For example, there is a lot of variation in bond spreads across banks with the same credit rating.
“It is also interesting to note that Australian banks recently have paid higher spreads on their bond funding than some lower-rated Australian non-financial corporates.”
This is not news for market participants. For instance, BHP Billiton has a history of relatively cheap bond issues because it doesn’t do them that often and investors will pay for the quality and scarcity. Banks on the other hand are frequent issuers of debt and the price they pay is subject to vagaries such as investors hitting concentration limits.
Large institutional debt investors have their own credit analysts who make their own calls on pricing, using the ratings agency as a benchmark if at all. The attention paid to the agencies’ scores varies from market to market and in different conditions – notoriously the agencies have been criticised for missing both the Asian financial crisis and the global one.
Beyond the role of the rating, the RBA in its paper also notes “to the extent that there is any implicit funding benefit to large banks, then in a competitive banking market which we have, we might expect at least some of this benefit to show up in lower lending rates than would otherwise be the case rather than just showing up in bank profits”.
That is, if banks fund themselves more cheaply due to an implicit guarantee, who is getting the benefit? This is impossible to measure but it will be shared between customers, shareholders and bank staff.
But the most significant concession the RBA paper makes is its analysis is a snapshot. In the RBA’s words, “time invariant”, an average of the subsidy over time. The bank says the subsidy is likely to be larger in times of stress and indeed this looks to be the case in its data – except in the period when it is largest, during the acute phase of the crisis, very little debt subject to an implicit guarantee was issued.
“These estimates show that prior to 2008, there was no significant funding advantage for the major banks. The size of the implicit subsidy appears to peak in 2009, at just over 100 basis points,” the bank says.
“The subsidy has since fallen; estimates indicate that the funding advantage fell to around 10 basis points at the start of 2O14, a level that is not statistically significant.”
Moreover, during the acute phase of the crisis, in response to what foreign governments were doing to protect failing banking systems, the Australian government introduced a debt guarantee. This however was explicit and paid for, not implicit and free.
“During, and shortly after, the financial crisis … most bonds issued were government guaranteed, and hence received explicit support,” the RBA says. “Because of this, estimates from around these times (the periods where the subsidy was found to be largest) are likely inaccurate, relying on few data points.”
Meanwhile, the other issue with the snapshot is global banking regulation since the crisis has been aimed at explicitly reducing the danger of systemically important institutions being bailed out by taxpayers.
Higher levels of capital, new liquidity requirements, explicit deposit guarantees (which level the playing field between smaller and larger institutions) and recovery and resolution plans have all been implemented or are being developed.
Critics argue the only way to truly eliminate the implicit guarantee for D-SIBS is lower leverage and more equity. Presumably they haven’t paid attention to how much new capital the banking system has raised and the current work by global regulators setting limits on leverage – which the Australian banks are well within.
When asking who benefits from implicit government support, one also needs to consider investors. They are the ones who will pay more for bank debt because they think it safer but almost all bank Tier 1 and Tier 2 debt now being issued is what is known as ‘Basel compliant’ – which means it can be converted into equity by regulators in times of stress. Hence bond investors are now explicitly paying for the risk they were previously hand-balling to taxpayers.
So-called 'living wills, recovery and resolution plans for banks, are also designed to eliminate the implicit taxpayer fallback by increasing total loss absorbing capital (TLAC), the details of which are being worked through. The idea is banks can be unwound and their liabilities paid for when they get into trouble without costing taxpayers money or the system to freeze.
Given the trouble gauging the actual value and even existence of an implicit guarantee, post the Basel III reforms of global banking regulations, and considering the cautious work of the RBA, the question is whether the two-notch upgrade S&P and other agencies build in is warranted (and whether it affects the cost of funding).
Clearly the agencies still believe the upgrade is warranted but both empirical analysis and the behaviour of debt investors suggest an implicit guarantee is barely worth the paper it is printed on.
Andrew Cornell is managing editor at BlueNotes