« Why I'll not vote | Main | Bold, true or trivial? »

May 02, 2005



Is your complaint that exchange rates are unpredictable in the short term, or over any term? The pound declined vs the dollar for decades. Is this about to turn round, over decades?


After working as a forcaster for decades I developed a degree of difficulty ranking.

Easiest -- the stock market -- I use leading indicators of the market pe.
next -- macro indicators like real GDP, inflation, etc.
middle -- micro indicators like copper prices, GM earning, etc.,
harder --interest rates
hardest -- exchange rates , you have too forecast at least two interest rates in addition to other variables.


lol - very nice

I recall a very brief discussion I has with a university prof (Martin Weale as it happens) concerning a Cambridge finals question. The question was

economic forecasting is a mug's game. discuss

Clearly students were expected to explain why the economy is difficult to predict, talking about animal spirits, endogenous cycles, sunspots and the like.

Weale made the comment that he personally would have given first class marks to students pointing out that forecasters are generally disproportionately well paid

you're in good company :)

Rob Hayward

Isn't a lot of 'forecasting' really just talking about something. I always look at it as being like forecasting a football match. Liverpool play Chelsea tonight and the chances of me picking the winner are probably 50:50 - the same as my wife's chances. However, I could talk for hours about why I think one team or another will win, while she could only guess.


Even though Warren Buffet has "lost a packet", his views on an overvalued dollar could still be the correct one. it's just he has timed out on the bet!!. It seems very similar to taking a view on the UK housing market. Whilst some commentators believe the market is overvalued and ready for a correction, no one can predict exactly when the market correction will happen.

Psychology rather than economics may be what's required.


Yes but if it only grew at 1pc a year for 10 years then that is a correction as inflation is 2pc, so ignoring compound interests and stuff, its a real terms correction of 10pc off the value of each house.
And this is what I think would be desirable for the Gvt, the Bank of E, banks and homebuyers. Plus because prices rise its an invisible loss to homeowners.



that point is what keynes was on about in his stuff on probability

your football example - you put the odds at 50:50 but have little confidence in the estimate basically because you have litte information on which to judge the reliability of the estimate - the weight of argument in favuour of this probability estimate is weak

contrast - you estimate probability of heads and tails for a coin flip - also 50:50, but your confidence level is likely to be much higher becuase the estimate is underpinned with arguments that have weight (balanced coins land equally probably heads or tails)

so probabilities alone are not the whole story if you want to understand how uncertainty and risk affect decision making and expectations - for example, expectations based on arguments with low weight are likely to be highly elastic and prone to volatility.

The comments to this entry are closed.

blogs I like

Blog powered by Typepad