Forecasting is difficult, especially about the future.

I once went to the finance director and asked what next year’s profit would be. “What number do you want?” he said.

I have to suspect that something like this happened in preparation of the figures just released by Swan.

Anyone who has worked in business knows that forecasts more than a few months out are a bit, but not much, better than guesswork. In most businesses I was involved in forecasts were not treated as fact but as something to trigger actions if reality was turning out too far from the forecast.

We also learned that forecasts of sales and costs should not change much in the short term unless there was a big event – earthquake, factory burning down or somesuch. You don’t get significant new information over a period of a few months. What you do get is swings between optimism and pessimism which is a dangerous way of running a business.

The change in the latest figures from the budget nine weeks ago seems to be the result of changes in forecasts of commodity prices over the next few years.

As my finance director would say “If the Treasurer can forecast commodities that accurately he wouldn’t be here – he would be in the South of France with his feet in a bucket of champagne.”

This entry was posted in Uncategorized and tagged , , . Bookmark the permalink.

8 Responses to Forecasting is difficult, especially about the future.

  1. Peter Brady says:

    We have had all the banks plus a few others in and they say privately the deficit will be lower than forecast this financial year and the next budget will be in surplus if commodity prices hold up ( which they think will).

    I have yet to meet any economist who examines the budget who believes the budget figures will worsen for what it is worth.

  2. In software engineering we have to perform estimation tasks knowing, for certain, that we will be wrong. Often wildly wrong.

    One tool used to convey this is the “cone of uncertainty”. At all points estimates ought to be conveyed as a range. As a project progresses, this range shrinks. This forms the ‘cone’.

    I haven’t seen the technique used elsewhere but I’m sure it must be in use under different names.

  3. JC says:

    That’s all true, Ken.

    However my concern is that something nefarious is going on and these trogs are just plucking numbers out of the air to provide a foundation for their policies, which is to spend more money.

    Even though forecasts are not worth much there still is a process that gets you there. I think the process has been politically subverted by the Treasury to help the ALP and the numbers come from thin air.

  4. Pingback: The $6 billion assumption at Catallaxy Files

  5. Greg Lindsay says:

    Ken, your finance director is just a little bit of a plagiarist. He’s right of course, but the champagne quote belongs to the late and great Modest Member, Bert Kelly:

    I don’t think Bert would mind though!

  6. Labor Outsider says:

    I will repeat what I said on the other related post. Unless you want to propose that Treasury stops producing budget projections it is simply essential that it also forms a view about a range of variables whose future path is uncertain. Even without MRRT, commodity price movements have an enormous impact on the budget. Look at what happened under Howard. Using the same methodology as Treasury currently uses, commodity prices were systematically underpredicted and hence the size of budget surpluses was also underpredicted.

    Unless you can produce evidence that the change in commodity price parameters is not simply a function of new information since the budget projections, but a change in methodology that is biased toward yielding greater revenue forecasts, then this line of attack makes little sense.

    Of course, given that future commodity price movements are hard to forecast, the government should be careful about committing the higher revenue to higher spending or lowering other taxes.

  7. JC says:


    Do me a favor, get off your soap box.

    I don’t have to produce any fucking evidence. It’s the other way around. We should demand Bernie Wombat-Henry produce the methodology he’s used to keep changing the numbers in the budget.

    Brain-dead SwanDive initially reported that the recent changes to the my fair share tax would produce 1.5 billion loss in receipts. That’s now gone to 7.5 billion.

    From the surrounding information it’s been reported that the government was expecting $24 billion as a result of the my fair share tax and not 13 billion.

    All this is now being attributed to mining tax receipts. Yea right.

    As I said, the ALP Treasury should be renamed Madoff & Co as they have numbers coming out of their arse.

    Do you like the name, LO?

Comments are closed.