To see this, consider all the countries in MSCI’s database for which we have equity returns in the past 10 years - 56 in all. Across these 56 countries, there is a good correlation (0.53) between equity returns (in US dollars) and real GDP growth since 2003. This might suggest there is indeed a case for investing in fast-growing countries.
It doesn’t. For one thing, the sensitivity of returns to GDP growth is small. A percentage point higher GDP growth is associated with only 0.12 percentage points higher annual equity returns. This is consistent with earlier research, which has found that most of the benefits of GDP growth flow not to shareholders but to workers, governments or unquoted companies.
One reason for this small effect is simply that there’s not much variation across countries in equity returns. For example, 33 of our 56 nations have seen their stock markets rise by between 2 and 8 per cent (unannualised) in the last 10 years. Many markets deliver similar returns, simply because they are closely correlated with global equity returns.
What’s more, there’s a big difference between actual GDP growth and expected growth. One reason for our apparently strong correlation is that unexpectedly good growth drives share prices up and unexpectedly bad growth drives them down. The most obvious of these shocks in our sample - though not the only one - has been the euro crisis. If we exclude the PIIGS (Portugal, Italy, Ireland, Greece and Spain), the correlation between GDP growth and returns in the last 10 years falls from 0.53 to 0.37.
But of course, surprises are - by definition! - unpredictable. Knowing that a country will generate good equity returns if GDP growth exceeds expectations is useless as a basis for investment unless we can predict growth better than the average investor can. And we can’t do this.
By all means invest in countries you expect to deliver fast growth if you have superior foresight than the market. But if you are human rather than a god, you might prefer another strategy.
And there is one. Instead of looking for future growth, which cannot be predicted, simply look for poor countries. There’s a strong negative correlation (minus 0.39) between a country’s GDP per head in 2003 and equity returns since then. A lower GDP per head of $10,000 per year (roughly equivalent to the gap between South Korea and the UK) is associated with 0.18 percentage points higher annualised equity returns.
Economic theory tells us there’s a strong reason why this should be. Other things equal, the marginal utility of consumption is higher in poor nations than in rich ones; when you’re poor, a given percentage fall in spending hurts you more than when you are rich.
This should mean higher equity returns for two reasons. One is that it means you have to make a bigger sacrifice if you are to save and buy equities at all. When you’re poor, buying equities means not being able to buy other nice things, but foregoing your third Aston Martin is less of a hardship. This means that, in poor countries, more people are deterred from buying shares, which means prices will be low and returns high for those willing to make the sacrifices necessary to buy shares.
Secondly, it means shares are riskier because a given fall in prices does more damage. If a bear market means having to take your children out of school it hurts you more than if it merely means having to postpone your world cruise. And if equities are riskier, they should carry a higher-risk premium.
There is, however, a big caveat here. This thinking only applies if stocks are bought by local investors. If instead markets are driven by international capital flows, share prices will depend upon the global marginal utility of consumption, and so returns will be similar across countries.
And this is true to some extent - returns are indeed similar across many countries, as January’s concerted drop in emerging and developed markets reminded us. But this is not an argument against investing in poorer countries. Instead, it’s a case for preferring so-called frontier markets to more established emerging ones. Established emerging markets are more integrated into the global economy and so sensitive to capital flows whilst frontier markets are less well integrated and more likely to be driven by local investors.
You might object here that it’s risky to invest in markets you know little about. That’s true, but local risks - such as political instability - are by definition idiosyncratic and so you can diversify against them by holding other markets. By contrast, more familiar markets are likely to fall at the same time as other stock markets fall - which is something that’s harder to diversify against.
There’s something else. In the last 10 years there has been a strong negative correlation (minus 0.67) between GDP per head in 2003 and subsequent real growth; poorer countries have grown faster than rich ones. It could therefore be that the correlation between growth and equity returns exists only because both are caused by the past level of GDP, so that if we control for the latter there is no correlation at all between growth and equity returns. (Whether this is actually the case depends upon how we cut the data: it is true if we exclude the PIIGs, but not if we don’t.)
The message of all this is rather encouraging for investors. It suggests that if we want to invest in overseas markets we don’t need futurological guff about spotting the next fast-growing countries - be they BRIC, Mint or any other acronym. We need only to look at more easily-available data such as GDP per head. Sometimes, investing is easier than the men in suits pretend.
