How to make accurate stock market predictions

As another year draws to a close, it’s natural to start thinking about next year, and for investors, that often means reading and making stock market predictions.

Sadly, stock market predictions are almost always a complete waste of time.  They have virtually no value other than as entertainment or as a way to make us at least feel as if we’re preparing for the future.

However, all is not lost.  In some years, it is possible to make sensible and accurate stock market predictions, but will 2013 be one of those years?

The calculator versus the crystal ball

Many economists, commentators and other market pundits like to predict the stock market in the same way that most people make predictions about the world in general – They just make it up.

That’s entirely reasonable because, in many situations, it’s not obvious that there’s a better alternative than simple rules of thumb like, “There’s a tiger over there, so I’m going to run the other way”.  The person running from the tiger hasn’t used any complex probability theory to work out the best course of action; it just seems reasonable that avoiding a tiger is a good idea.

It’s an approach that has served us quite well for many thousands – if not millions – of years, and so it’s a well-tried approach.

Unfortunately, it isn’t very good at predicting the future in detail.

Saying, “There’s a tiger, so I’m going to run the other way” (and then doing it) is a sensible thing to do, but to start making complex predictions about the tiger’s future position, direction, mental state and who knows what else in order to calculate some probability that you’ll be able to walk past safely in ten minutes time, is both a waste of time and very dangerous.

The additional precision will most likely add nothing other than the illusion of knowledge.

In that respect, the stock market and the tiger are very similar.

We know that at the end of 2013, the FTSE 100 will be at some level, and we know that since it’s currently around 5,500 to 6,000, it won’t be at 30,000 and it won’t be at 1,000.

But the more we try to narrow down the range and become more precise, like trying to predict the exact future position and direction of the tiger, the less likely we are to be right.

But as I said, it’s not all doom and gloom.

There are some tools which we can use to get at least a general sense of where the market might be headed in the next year, and they generally involve a calculator rather than a crystal ball.

PE10, CAPE and Tobin’s Q

If you’ve been reading the right kind of books, then you’ll probably already know these metrics.  In slightly different ways, they all try to make the price-to-earnings or price-to-book ratios more useful and more predictive.

Tobin’s Q is apparently the most accurate, but I think that the differences between them and the impact of those differences in the real world are likely to be minimal, so I’ll stick with CAPE (cyclically adjusted PE), which tends to be the most popular.

In very simple terms (because the idea is very simple), the market is priced at a certain multiple of its earnings which is the price-to-earnings ratio that we’re all familiar with.  The problem is that the earnings part is very volatile, so the ratio is often meaningless and tells us virtually nothing about whether the market is cheap or expensive.

By extending the earnings out to a ten-year average, it becomes a much more stable number.  Comparing price to a more stable earnings number gives us a more meaningful measure of value.

A higher CAPE really is predictive of lower returns, and vice versa.

Over the last century, the S&P 500 CAPE has swung almost randomly between depression valuations all the way up to extreme bubble valuations, and this range is generally between half and double its long-run average of fifteen to eighteen or so.

Extreme events make the world more predictable

Sometimes the weather is easy to predict.

When it’s the middle of summer and the temperature is over thirty degrees, just say, “I think on some random day in the future, it will be cooler than it is today” and you’ll probably be right.

Or, in the middle of winter, when snow is falling and the temperature is minus ten, just say, “On some random day in the future, it will be warmer than it is today” and again, you will probably be right.

But, if it’s May and the temperature is eighteen or nineteen degrees, it becomes much harder to say whether the temperature will be higher or lower on any given day in the future, at least in most of the UK.

I’ve seen eighteen degrees in both February and July.  So for easy predictions, we want extreme situations, while “normal” situations make prediction very hard.

For example, in 1999, the FTSE 100 CAPE reached a high of about 30, which is pretty much double its mid-teen average.  From that point, it was very easy to predict that future returns would be terrible, which is exactly what happened.

In the same way, in 2009, CAPE sank to about 9, which is close to half the long-run average, and again it was easy to predict that returns from that point would be outstanding, as indeed they were.

So if you want to predict the future, make sure you’re at the extreme end of a mean-reverting system (like the weather or the stock market).

So where are we today, and what are the implications for a 2013 prediction?

With the FTSE 100 at 5,800, CAPE is about 13.  This is just slightly below the long-run average, which is both good news and bad news for investors.

It’s good news because it means that the market and shares in general are relatively good value, with future returns likely to be at or above the average rate of 8% or so.

On the other hand, it’s bad news for predicting where the market will be in a year because the market’s valuation is pretty much “normal” and, therefore, not extreme.

So the market today is much like a day when the temperature is eighteen degrees; it’s almost impossible to say whether the temperature (or the market) is likely to be up or down at some unknown point in the future.

But this lack of certainty shouldn’t be a concern for serious investors.  After all, risk, or uncertainty, is something we have to have in order for future returns to be attractive.

If you really need to know whether the market will be up or down next year, you probably shouldn’t be investing in stocks in the first place.

This somewhat unsatisfactory prediction for 2013 only serves to highlight the real point…

The only sensible way to predict the market – and to invest in it – is over the long term, and so my CAPE-based prediction is this:

  • Ten years from now, the FTSE 100 will almost certainly be significantly higher than it is today, and with dividends included, the total return in that time could easily be over 100%.

For some alternative views on market valuation, have a look at this article on  Does the FTSE 100 tumble present a buying opportunity?

Author: John Kingham

I cover both the theory and practice of investing in high-quality UK dividend stocks for long-term income and growth.

One thought on “How to make accurate stock market predictions”

  1. stock market predictions are almost always a complete waste of time. They have virtually no value other than as entertainment, or as a way to make us at least feel as if we’re preparing for the future.

Comments are closed.

%d bloggers like this: