The Value of Value

Ask an investment person their view on markets and the main thing they will probably want to talk about is valuation: favour Asset A because it is cheap and avoid Asset B because it is expensive. Why this fascination with value? Sure, everyone loves a bargain, but there is more to it than that. In a world of uncertainty, valuation analysis can give a surprisingly clear window into the future and for many, measuring value is deemed easier than and preferable to reading the tea leaves of the global economy. This is unfortunate, because the one thing that is overvalued is the role that valuation can play in tactical strategies.

There are two main problems here. The first is that valuation is hard to define and even harder to measure. When we say that an asset is cheap, we are saying that we can buy it below some measure of fair value and that we will be able to generate decent returns when the valuation reverts to the mean. But how do we determine fair value?

Any measure of fair value must make sense and needs to be relatively stable. This later point is important. There is no point have a measure of fair value that is inherently volatile, because otherwise, assets that appear cheap one minute never get the chance to ‘mean revert’ because the yardstick closes the gap first. The price-earnings multiple (PE) for equities is a clear case in point. Using this metric, equities are deemed to be cheap when share prices are low relative to company profits. The problem is that profits are more volatile than share prices and hence PE’s have very limited forecasting power. How often have we seen stock markets peak when PEs are low and bottom when they off the scale ‘expensive’ as per 2009?

The second problem relates to the fact that even the better valuation yardsticks only work over long periods of time. Take for instance the Shiller PE – which compares share prices to a 10-year average of real historic earnings – which correlates relatively well with future returns. But its closest fit with future returns is over a 10-20 years’ time horizon. To say the least, that is beyond the patience of most investors.

The one time that valuation can be more powerful in terms of forecasting returns over shorter periods are when it is close to historic extremes. But even here, we need to tread with care; records are there to be broken and expensive markets can become more expensive. Just ask all those analysts that predicted disaster for US equities in 1995-6 when valuation yardsticks reached what was then the top end of their historic range. Many of them were no longer employed when the market eventually peaked at much higher levels in March 2000. It is also important to remember that even when markets peak in overvalued territory, they rarely – if ever – peak because they are expensive, rather it is because of fundamental developments, e.g. rising interest rates and/or bond yields.

Bonds are more mean reverting than their equity cousins, with the result that valuation has more predictive power over shorter holding periods, but even here we are talking years rather than months or even quarters. Plus, even bonds are prone to periods of considerable overshoot – just look at what is happening right now or what happened in the other direction in the early 80’s.

Valuation deserves its place in the analyst’s toolbox, but needs to be approached with caution when it comes to tactical asset allocation. Where valuation can be very helpful is in framing investors’ expectations about what they can realistically expect from their portfolios in the years ahead. And here the news for traditional portfolios is not good, with valuation experts forecasting negative real returns from developed world government bonds and equities in the next 7-10 years.

With this in mind, what is the investor to do? The first thing to say, is don’t panic. The whole point of this article is that valuation is a very poor market timing tool. Markets may well be disappointing in the long-run, but that does not mean that they cannot do great things in the short to medium term. Indeed, that is essentially what we expect. The second point to make is that even if the main indices perform poorly, that does not mean that everything will be similarly bad. The S&P500 was similarly expensive in March 2000 and went on to produce negative returns in the next 10 years. However, investors were still able to generate positive returns from value and (most notably) small caps.