from streetdogs.co.za articles archive home
Advice on forecasts: Don’t! Michel Pireu     | Business Day Wednesday, January 26, 2011
“The folly of forecasting is one of my pet hobby-horses,” says James Montier in The Tao of Investing, which he wrote while still with Société General. “I simply can’t understand why so many investors spend so much time engaged in an activity which has so little value, and so little chance of success.”

“For instance, let’s say you invest according to the following process: forecast the economy, forecast the path of interest rates, forecast the sectors which will do well within that environment, and finally forecast which stocks will do well within that sector. Now let’s assume you are pretty good at this and you are right on each forecast 70% of the time (massively above the rates actually seen). However, if you require all four forecasts to be correct, then you have just a 24% chance of actually getting it right! (This assumes that each of the forecasts is an independent event). Now think about the number of forecasts an average analyst model contains. Sales, costs, taxes, etc. no wonder these guys are never right. In addition, even if by some miracle of divine intervention your forecast turns out to be correct, you can only make money from it, if (and only if) it is different from the consensus. This adds a whole new dimension of complexity to the problem.”

“The evidence on the folly of forecasting is overwhelming,” says Montier. “Economists haven’t got a clue, frankly the three blind mice have more credibility than any macro-forecaster at seeing what is coming. The analysts are no better. Their forecasting record is simply dreadful on both short- and long- term issues. My colleague Rui Antunes has examined the accuracy of analysts. In the US, the average 24-month forecast error is 93%, and the average 12-month forecast error is 47% over the period 2001-2006. The data for Europe are no less disconcerting. The average 24-month forecast error is 95%, and the average 12-month forecast error is 43%. To put it mildly, analysts don’t have a clue about future earnings. And their performance in divining the longer-term future is sadly no better than their performance in the short term. Even a cursory glance at the [research data] reveals … analysts have absolutely no idea about forecasting long-term growth.”

Montier’s final rant on the folly of forecasting concerns target prices. “Why do analysts persist in trying to forecast prices?” he asks. ”As Ben Graham said forecasting security prices is not properly a part of security analysis. Analysts have an embarrassing track record with respect to target prices.”

In a nine year study of analysts’ prediction of prices 12 months ahead he found that on average they predicted stocks to be 25% higher each year. “”I then contrasted this implied analyst view with the actual returns achieved across the same universe,” he says, ‘and in four out of the nine years, analysts have not even managed to get the direction of change in prices correct!”

“The bottom line from this whistle stop tour of the failure of forecasting,” says Montier, “is that it would be sheer madness to base an investment process around our seriously flawed ability to divine the future. We would all be better off if we took Keynes suggested response when asked about the future, which is: We simply do not know.”
 
contact us |  add to favourites
disclaimer
Michel Pireu & Associates cc, the owners of this website, are active stockmarket traders. The company disclaims all liability for any loss, damage, injury or expense however caused, arising from the use of, or reliance upon, in any manner, the information provided through this service and does not warrant the truth, accuracy or completeness of the information provided. The company's permission is required to reproduce the contents in any form including, capture into a database, website, intranet or extranet.