Investing is all about saving for the future, building a financial cushion to live comfortably on in retirement or building a fund for a rainy day. No matter how you invest, or what you are investing for, you will never be able to get away from forecasting.
Forecasting your needs in retirement, potential investment returns, a company’s projected growth or percentage of savings you can afford every month are all important parts of investing. The problem is we humans are useless at forecasting, which is a big problem.
Bad at forecasting
Humans’ lack of ability to be able to put together an accurate forecast is more troublesome for market timers and equity analysts than anyone else, but it still pays to keep in mind how useless we are as a race at this problem.
The next time you believe or quote some Wall Street analyst or expert's forecasts, remember these figures from James Montier; for analysts, the average forecasting error in the U.S. between 2001 and 2006 was 47% over 12 months and 93% over 24 months. These figures, which were published in 2010, are somewhat out of date but still make an important point.
What’s more, during this period the U.S. economy was experiencing an economic boom, and if analysts cannot predict growth accurately during a bull market, how do they stand any chance of being able to precisely forecast growth or returns during an economic downturn?
Weak stock forecasts
It is not just macro forecasts that are subject to this problem. The same issue exists with bottom-up stock analysis. Specifically, discounted cash flow calculations are especially open to incorrect calculation due to unreliable and unbelievable forecasts.
In 2008, Montier wrote a paper discussing the perils of the DCF calculation and why it should not be used. Instead, he suggested three other valuation methods be used in its place.
The argument as to why investors should not even contemplate using this staple method of valuation comes back to forecasting:
“From the point of view of DCF, the forecasts are central. Most DCFs are based on relevant cash flows years into the future. However, there is no evidence that analysts are capable of forecasting either short-term or long-term growth…In the U.S., the average 24-month forecast error is 93%, and the average 12-month forecast error is 47% over the period 2000-2006. Just in case you think this is merely the result of the recession in the early part of this decade, it isn’t. Excluding those years makes essentially no difference at all. 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%. Frankly, forecasts with this scale of error are totally worthless.”
Further research only supports this conclusion. In the same paper, Montier takes a look at the forecasting ability of analysts with regards earnings growth. Surprisingly, analysts actually provided more accurate predictions of growth for cheaper stocks than they did for high-priced, high-growth equities. Over the period 1985 to 2007, analysts expected a portfolio of the cheapest stocks on a price-book ratio in the market to grow earnings at a steady rate of 10% per annum. Real growth over the period turned out to be 9% per annum -fairly close to forecasts.
On the other hand, the stocks analysts had been projecting the highest growth for (an average of 17% per annum over the period studied) only managed an average growth rate of 7%, less than that of the value stocks.
You can read the full study here.
The bottom line
So what is the takeaway from all of this? For a start, the one main takeaway is forecasting is useless. The second point to make is optimistic Wall Street forecasts for high-growth equities should never be believed.
As well as being useless at forecasting, these analysts are under pressure to produce the highest earnings growth rates or risk falling out with large clients. That said, the data does suggest the most conservative analysts covering low growth, value equities produce the most accurate forecasts.
Start a free 7-day trial of Premium Membership to GuruFocus.
