I worked a bit more on deposit guarantees and the resulting note is over the fold.
Until the recently announced proposal to guarantee bank accounts up to $20 000, the security of Australian bank deposits rested on two main presumptions
(i) the Reserve Bank would act to maintain stability. While this has always excluded an explicit guarantee of deposits, it was generally understood to mean that anyone who had a standard bank account (not precisely defined) with a bank complying with APRA regulations would be fully protected against default either through the lender of last resort facility, an RBA-facilitated rescue by another institution or in the worst case by the RBA taking over the bank and assuming its liabilities to depositors
(ii) the fact that depositors rank ahead of all other creditors, and account for only about 50 per cent of liabilities meant that, even if the Reserve Bank decided to let an institution fail, depositors would almost certainly recover their money in a liquidation
The second of these defences no longer seems adequate in the light of overseas bank failures in which losses far exceeded 50 per cent of the capital value of the bank concerned.
The general acceptance of the Reserve Bank as implicit guarantor has remained strong (if shaken by recent events). However, this can scarcely be combined with an explicit and limited guarantee of deposits up to some fixed amount.
In this context, the proposal to guarantee bank accounts up to $20 000 is worse than useless. According to statements made in support of the proposal, the result would be to protect 85 per cent of depositors, or equivalently, to withdraw protection from 15 per cent of depositors. The proportion of deposits left unprotected is presumably larger.
Under a scheme in which individual accounts are protected to a fixed amount, depositors have an obvious incentive to diversify, holding a number of accounts in separate institutions, each of which falls below the statutory limit.
The process of diversification would require large-scale withdrawals. While the money withdrawn would be rapidly deposited in new accounts, such a churn process might not be symmetrical, and would in any case be likely to generate considerable alarm. The possibility of this turning into a run is far from remote.
The Opposition response, suggesting a guarantee of $100 000 is an improvement. But this is still below the level guaranteed in most other deposit-insurance jurisdictions (the US level was recently increased to $US 250 000). More importantly, most jurisdictions are abandoning limited insurance in favour of unlimited guarantees. In the current environment, only an unambiguous guarantee appears sufficient to avoid the risk of panic.
The financing of an insurance or guarantee scheme is straightforward in principle, though the calculation of premiums will be a difficult task in the absence of any historical failures, but in the presence of greatly increased risks in the immediate future. However, at this point it is more important to get an adequate guarantee in place than to get the details of financing correct.