The number that reveals when Asian markets are cheap

In the middle of the eurozone panic, it's easy to lose sight of what matters most in investment.

Yes, the flailing and failing of European governments could rattle markets for months more, if not years. If the crisis gets worse – which is quite likely – markets could even go lower.

But for long-term investors, this steady stream of bad news is much less important than a very simple question: How cheaply are you buying?

Over the long run, what drives investment returns is valuation. It's certainly been true in Asia in the past. So let's see what history says about when you want to buy, when to sell and where we are today.

The best way to value Asia

Regular readers will know that one way I like to look at market valuations for Asia overall is the price/book ratio (also known as the price to net assets). Price to book has the advantage that the value of assets used in this ratio doesn't vary as much as earnings over the cycle.

As a result, we don't need to worry so much whether we are at cycle-peak earnings or in the trough of the recession. So price/book looks through the cycle much more effectively than price/earnings.

It helps that Asia is very cyclical when it comes to valuations. The chart below shows the price/book ratio since 1975 (based on a combination of modern data from MSCI and Bloomberg and a reconstruction of historic data from Citigroup). As you can see, it frequently soars into a bubble or plunges into a bust.

This is very useful, because given such pronounced cycles, there's a good chance that comparing valuations and subsequent returns will give us some idea of what to expect. A market that always trades on a price/book ratio of around two is likely to have much less information in the valuation than one that frequently over- and undershoots.

What history tells us about returns

To see what we might expect, I looked at a series of one-year, three-year and five-year returns for the MSCI Asia ex-Japan and compared it to each end-month price/book ratio since 1995. (I used 1995 rather than earlier years to keep all data strictly comparable, because this is when the modern data series from Bloomberg begins).

Initial numbers on this don't look very impressive. The r-squared value – which measures the proportion of one variable that can be accounted for by the other – was 0.29 for one-year returns, 0.33 for three-year returns and 0.43 for five-year returns. While investment models usually have a significant level of error, you might hope to get a better fit than that.

But if you look at data as a scatter plot – price/book on one axis, subsequent return on the other – it looks more interesting. Price/book clearly is not going to predict your exact returns – but it looks like it may give you some idea of whether you're likely to get a decent gain or a severe loss. 

Over one year, buying at very low values has almost invariably been associated with a good performance over the next year. Buying at the bubbly levels above 2.2 or so has usually led to a significant loss.

It's a similar pattern on three-year and five-year returns - although the longer your time period is, the more likely you are to get a decent return from buying at higher valuations. Note that this doesn't include dividends, which will start to be a significant part of returns around five years.

When should you buy and sell?

There's a limit to how confident we can be about any conclusions. We only have 15 years of data – although helpfully this covers some significant ups and downs, including the Asian crisis, the dotcom bubble and the global financial crisis.

Obviously conditions could change. So could markets: the MSCI Asia ex-Japan of 1995 had some significant differences to today. The further we go back, the more notable this becomes: in 1973, Asia ex-Japan consisted of Hong Kong and Malaysia-Singapore, which was still a linked market – a very different world.

Nonetheless, history – together with a best guess at the future – is an important guide to what we can expect.

Buying much above two looks decidedly risky. In the past, I've always reckoned you're safe until about 2.2-2.4 – but looking at these charts, I'd probably start questioning the market a bit earlier in future

But now, the MSCI Asia ex-Japan is on a price/book ratio of 1.6, which compares to a median of 1.85 since 1995. A look at those scatterplots suggests that at this level, we have a better-than-even chance of good returns over the next year. Over longer periods, this should improve further.

Clearly, the worse the situation gets in Europe and elsewhere, the more likely the Asian markets are to go lower. And it's at this point that most people pull out of the market.

Yet it's exactly this time that you want to be buying. Panics are helpful for the long-term investor, because they let you buy more cheaply. History suggests that if the price/book ratio drops further to, say, around 1.3, the argument in favour of buying would be extremely strong.

And at very depressed levels – eg below one – are short-lived, panic-driven opportunities that have usually delivered very strong returns in a matter of months. In the unlikely event we see that this time, investors should be willing to go against the crowd and buy enthusiastically when everyone else is selling.