Bank stress testing is an interesting thing to think about. The goal is to model what would happen to bank income statements and balance sheets in the wake of “shocks” like house prices returning to realistic levels or people realising that paying debt back just gives executives bonuses and less disposable income for cool stuff.
The Reserve Bank’s latest Financial Stability Report includes some discussion of recent bank stress tests done in collaboration with the Reserve Bank of Australia.
The impaired asset expense estimations even when the “shocks” happen are extremely low. An impaired asset expense is recognising on an income statement that you’re unlikely to have a loan repaid.
This flows through to whether the bank makes a profit or loss. If they make a loss, they have to raise additional capital in order to retain high Tier 1 capital ratios under Basel III requirements.
There is $35 billion in wholesale funding that banks have to rollover to replace over the next 3 years. With the euro-zone crisis still nowhere near being resolved, and the LIBOR scandal, will there even be a functioning wholesale money market in 3 years?
To say that there will be is arrogant in the extreme. Just look at how quickly the overnight lending markets evaporated after Lehman Brothers. We have no idea what could trigger another freeze. What if Greece exits suddenly and the charade that the euro-zone still has some semblance of solvency evaporates?
The RBNZ and RBA worked with the banks’ own models in order to conduct stress tests. They do not say whether they adjusted the banks’ figures to reflect the incentive they have to under-report substantial losses they would incur if certain shocks happened.
There is also not much discussion on whether impaired asset expenses peaking at 2% is realistic at all. This is odd. Why? Because if a major shock happens and New Zealand goes into a deep recession, we have no way of telling how the dominos will fall.
What I mean by this is that we can’t say what will cause different businesses to make shutdown decisions in the wake of a deep recession. Some business owners might fear the worst, layoff their employees promptly and suddenly an extra 30-40 mortgages start falling behind.
What happens if impaired asset expense ratios go above 2%? And if banks aren’t able to raise capital either domestically or on the wholesale money markets?
They become insolvent. Their capital ratios fall below their Basel III requirements. There is a risk that no capital injection can be arranged and so we turn to the proposed Open Bank Resolution strategy.
While I won’t go into detail about how the Open Bank Resolution process will work, I will make one comment : what if the initial losses are far greater than what the Reserve Bank has considered in building their model?
Banks actually have a major incentive, if the government is going to guarantee solvency support of the unfrozen portion of their assets, to make hay while the sun shines.
Paying bank executives 7-figures salaries in gratitude for shifting potential taxpayer liabilities into the tens of billions of dollars, when there is no recourse for clawing back excessive compensation and excess returns that have been clawed off the table, is folly in the extreme.
We cannot tell what the actual level of losses will be. We cannot tell what the actual impaired asset expense will be if a major bank collapses a year or so after a deep recession.
This means that the most honest value of government solvency support to a bank admitted to the Open Bank Resolution scheme is $X billion to $XX billion.
Furthermore, if the RBNZ takes over an insolvent bank, are there measures in place to insure futures contracts are rolled over to avoid cash settlement? What if a $10 billion notional US Treasury position isn’t rolled over? Do term depositors or taxpayers stump up for that?
If I was employed by the RBNZ I would be actively “war gaming” things outside the model. What if the Australian government, for political reasons, doesn’t want its super funds to eat the losses of an insolvent big bank? What if the Australian High Commissioner takes a trip to the Beehive and tells Bill English that the New Zealand government will eat the losses in partnership with the Australian government?
Bank stress testing based on the numbers is better than no bank stress testing at all. But there needs to be some clear thinking about what “outside the model” potential hurdles might need to be overcome.