Monday, May 18, 2015

Don't trust the knockers of Treasury forecasts

People keep asking me whether the budget's forecasts for the economy are "credible". Of course they are. But that's not saying much. And here's a tip: don't believe those saying that Treasury's forecasts are way too optimistic or way too pessimistic. They wouldn't know.

Treasury and the Reserve Bank – whose forecasts are essentially a joint exercise – put an enormous amount of time and expertise into their forecasts, far more than any other outfit you could name.

That doesn't mean their forecasts are likely to be right, of course. Far from it. But it does mean that, on average over time, they're likely to be less wrong than their critics.

On any particular forecast, any individual has a chance of being right while the official forecasters are wrong, just by luck. You can only remove the luck factor by comparing people's forecasting record over time.

And yet at this time every year we have smarties popping up to confidently assert that the budget forecasts are wildly optimistic or "built on artificial assumptions". They do so knowing no one will check how right their prediction proved to be.

They're entitled to their opinion. But you're entitled to say to yourself, what would they know? Some of these people have done no more than run their eye over a row of forecasts and thought, I don't believe it.

Some are sceptical because they don't know as much as they should about the standard dynamics as the economy moves through the ups and downs of the business cycle.

They forget that, when finally the economy picks up after growing below "trend" (the medium-term average rate of growth) for a number of years, it's likely to grow at rates well above trend for a year or two as it takes advantage of all the accumulated idle capacity. The hard part is picking the timing of the turning point.

The I-know-better brigade rarely claim the forecasts are wildly pessimistic – which they sometime prove to be – because they're pessimists and knockers, often with their own axe to grind.

The smarties never preface their pronouncements by admitting that their own forecasting record is just as bad as Treasury's, probably worse. No, they just assert that Treasury's wrong and they're right.

That's why I'll always fly to the defence of the official forecasters. They're the only honest players in this game. They regularly measure the accuracy of their forecasts and publish the results. They regularly remind users that, given their poor record, their forecasts are little more than an educated guess.

To make the point even clearer, in this year's budget papers Treasury has collected its usual warnings, qualifications, disclosures and "sensitivity analysis" into a single new section of the budget papers. It starts by admitting that "the forecasts are subject to considerable uncertainty".

For these purposes, Treasury has combined its forecasts for the financial year just ending and the coming year to give a forecast average annual growth rate of "around 2.5 per cent". On the basis of its record of forecasting errors, it says there's a 70 per cent probability that the actual growth rate will be somewhere between 1.75 per cent and 3.5 per cent. Wow.

The critics tell us that Treasury's forecasts are based on many assumptions. That's true. But it's always true and is just as true of its critics' forecasting models (assuming they bothered to do any modelling before shooting from lip). Assumptions are inescapable.

Some key assumptions are for the exchange rate, interest rates and world oil prices. Treasury takes their average in the period before its forecasts were finalised in April, and assumes this will be their average over the forecast year.

Last week the smarties noted that some of those prices had already moved away from the assumption and concluded that the budget's forecasts had been invalidated already. Nonsense. Who can be sure how those prices will have moved – and thus averaged – by this time next year?

In any case, just because some assumptions prove lower than expected doesn't mean others won't prove higher than expected and thus cancel them out.

Though it's human nature to pretend otherwise, the simple truth is that no one knows what the future holds for the economy: it may be about to take off, about to collapse or about to stay as it has been. We can all have our opinions, but no one can say I'm right and you're wrong.

Point is, we can't be sure Treasury's forecasts will be right, but nor can we be sure they'll be wrong. They may be no more than educated guesses, but they're as plausible as anyone else's.