Currency traders were right to trim the value of the Australian dollar on comments by Reserve Bank governor Glenn Stevens that the central bank still thinks the $A is too high and remains "open minded" about intervening in the markets to try to push it down.
On Friday the Aussie was trading around US92.4¢. That was above a low of just over US89¢ at the end of August, but a half a cent down after Stevens' speech on Thursday night, US4.5¢ below a mini-peak of just over US97¢ in late October, and US13¢ below April this year, when it finally began adjusting to weaker commodity prices.
That's about right weight. Stevens didn't flag an immediate attempt to wade into the markets and swap Australian dollars for foreign currencies, creating sell-side pressure on the $A in the process.
What he did make clear, however, is that there is a risk-reward trade-off for doing so - and it has moved in the central bank's favour.
Stepping in to sell the $A not so long ago would have been the equivalent of jumping in front of a runaway train, and that is no longer the case.
Stevens said buying or selling the currency when it moved on either side of its "equilibrium" value became more risky as the resources boom expanded and the global crisis resulted in interest rates in northern-hemisphere countries falling to zero, opening up a profitable currency "carry trade" into Australia, where rates are more attractive.
A question arose about whether a new equilibrium value for the $A had been created. The currency was probably still above its longer-term equilibrium value, he said, but it was not entirely clear what the new equilibrium was. The forces that had pushed the $A up in the last half-decade were unlike those at work in the first 25 years of its life as a floating currency, and were very powerful.
Stevens was first talking about balance sheet valuation risk with those comments - the risk of asset value write-downs if it sells the $A and buys foreign exchange only to see the $A rise further. Ahead of any new buying, it owns about $36.6 billion of foreign currency. A 10 per cent rise in the $A's value would create a paper loss of about $3.6 billion on that base alone.
He also noted, however, that there is also a direct revenue and profit hit. If the Reserve sells Australian dollars and buys foreign currencies, it is swapping money that is earning about 3 per cent and buying money that is earning zero per cent, or very close to it. That is a loss-making "carry trade", the reverse of the profit-making carry trade that has been sucking international capital into the $A.
Stevens went on to discuss cost-benefit equations on intervention. It might be argued, he said, that the negative carry created by selling the $A and buying foreign currencies and the accompanying "very large valuation risk" were "a price worth paying" if intervention corrected "a seriously misaligned [that is, overvalued] exchange rate".
Intervention might turn out to be profitable in the end, he said. That would be the case if it assisted in a sustained decline in the value of the currency. The larger point, however, was cost-benefit sums had to be considered along with the likely effectiveness of intervention. So far, the Reserve had not been "convinced that large-scale intervention clearly passed the test of effectiveness versus cost", he said.
He then said what the currency traders seized on. "That doesn't mean we will always eschew intervention. In fact we remain open-minded on the issue. Our position has long been, and remains, that foreign exchange intervention can, judiciously used in the right circumstances, be effective and useful. It can't make up for weaknesses in other policy areas and to be effective it has to reinforce fundamentals, not work against them. Subject to those conditions, it remains part of the tool kit."
If the Reserve wants to try to push the $A down by exchanging it for foreign currency, it can. It owns the $A printing presses and so has an unlimited supply. Australian dollar buying operations on the other hand are limited by the amount of foreign currency the Reserve has to sell - $36.6 billion at the end of October, the same amount it owned at June 30.
The $A is, however, a very deep currency market, the fifth most heavily traded in the world. It could take a lot of money to move it. The Swiss National Bank sold about 175 billion Swiss francs and bought foreign currencies worth about $US191 billion in 2011 and 2012 to halt a rise in its currency, for example. That was about a third of Switzerland's annual GDP: a similar $A selling operation would be worth $500 billion.
The Swiss operation was, however, mounted at a time when there was a massive, one-way river of money pouring into the Swiss franc as investors fled from Europe's sovereign debt crisis. It was not easily sand-bagged and when the Australian dollar was soaring the task here would have been just as daunting.
Now, however, as the decline in the value of the $A from its highs shows, the $A trade is more two-way. It is also smaller than it was when the dollar peaked.
There is not enough selling yet to get the currency down as far as the Reserve, the government and currency-exposed industries want, but there's enough to give a $A selling program more leverage.
Stevens is signalling that if it wants to intervene, the Reserve should now get more bang for its buck. That changes the odds on intervention and the markets are right to take note.