There is a lot of soft headed piffle that’s published on the subject on the Australian Dollar. This week’s article by the Australian’s David Uren is a rare exception.
In particular, Uren is the only popular pundit I have seen who points out that RBA sales of the AUD are not cost less. This is an important point.
IF the RBA were to sell the AUD, they would exchange high-powered money (reserves) for foreign currency. In the trade, the RBA’s counter-party would obtain an AUD deposit at some clearing institution, and the RBA would obtain a foreign currency deposit at some clearing institution.
While the RBA has unlimited ability to supply Australian Dollars, the transactions create a ‘cost’ for the Bank – in the form on interest bearing deposits at the RBA.
Like all central banks, the RBA’s balance sheet must balance (see here so the AUD reserves it creates to pay for the foreign exchange it purchases on the open market must re-appear as deposits at the central bank.
These reserves cannot leave the system. The central bank’s balance sheet must balance, so they must re-appear as a deposit at the central bank, in the form of exchange settlement balances (by the way, this is why it is total nonsense to look at Fed and ECB deposit balances and declare that the balance sheet expansion isn’t being lent). Thus, they ought to be thought of as a loan from the banking system.
The RBA has the privilege of choosing the rate it pays on those liabilities (that’s what’s so special about central banks), so it has some control over this cost – however it can only move this rate in a way that’s consistent with its inflation target. Appreciating this is what distinguishes the Swiss and Australian cases.
Before the SNB started selling CHF, it first lowered its cost of funds to zero to fight deflation. Once that step had been made, it had access to zero nominal interest rate loans – and it used this cash to buy foreign exchange (and ultimately, foreign bonds).
In executing this policy, the SNB purchases EUR at the 1.20 ‘cap-rate’, which is 0bps loan from the CHF banking system, and invests in foreign assets. It is hoping to make a profit both from the positive carry (the difference between their cost of funds, and the yield on the asset they purchase) and FX gains when the foreign exchange value of the CHF eventually declines.
What’s its risk? That the EUR collapses, and the bonds it is purchasing are not repaid in full; and that it becomes appropriate to allow the foreign exchange value of the CHF to rise.
In Australia’s case, the RBA would be selling AUD with a cost of funds of 25bps below the target overnight cash rate. Assume for the moment that the RBA’s policy rate is steady at 3.5% (a weaker AUD would mean that there is little prospect of further rate cuts), and that the RBA needs to sell about 500bn of AUD to make a meaningful impact.
The RBA is likely to purchase a mix of highly rated foreign Bonds with about a 5yrs maturity — on that portfolio it would be lucky to earn an annual yield of 1% per annum.
If this were the case, the RBA would be paying 3.25% on 500bn (~16.25bn per annum), and earning 1% on the amount (~5bn per annum) – for a total cost of 11.25bn per year.
Who pays this? The Australian Taxpayer — for the RBA’s balance sheet forms a part of the Australian Government’s consolidated accounts.
And that’s not the final cost to the tax-payer — most taxpayers are net importers, and the weaker value of the AUD would lower their real consumption wages.
In reality, this policy would be a huge tax on Australian households – for the benefit of a narrow range of exporters. And once we re-open FX intervention as a policy option, I guarantee that we are going to have to re-open tariffs, quotas, and other trade subsidies – for the losers from interventions to move the foreign exchange value of the AUD will certainly be looking for ‘compensation’ from the Treasury.
A free float and free(er) trade have served us well. It would be a grave and costly mistake to re-open this Pandora’s box.