ONE of the winners from an unexpectedly sleepy summer for capital markets has been emerging market (EM) assets. Here, we explore some of the ways we might simplify our attempts to predict the tactical performance of EM assets.
EM asset returns may be closely correlated, but can they be explained by the same variables?
One common hypothesis is that EM assets are a high-beta play on global markets. If this is indeed the case, then EM asset returns may be largely explained by common global factors.
In line with research done by other houses, we formally tested this hypothesis by regressing annual EM asset returns on annual returns of a handful of predictor variables such as the dollar index, developed market equities, oil prices and US 10-year treasury yields.
The results are interesting to say the least. From the results, we can see that all three EM asset class returns are well explained by the global macro factors mentioned above.
Encouragingly, our model also produces intuitive results – US yields and the US dollar (USD) are shown to exhibit a negative relationship with EM asset returns, while oil and developed market equities exhibited a positive relationship.
Admittedly, we are surprised by the unusually high predictive power of our regression model – it seems to leave little explanatory room for fundamental factors such as earnings growth and domestic economic conditions, thus suggesting that EM returns are little else but a function of the global environment.
However, it is important not to muddle correlation and causation when it comes to thinking about EM assets. Just because the factors cited above have high explanatory power, doesn’t automatically imply that they have a causal impact on EM asset returns.
For example, in the period since 2000, we can see that USD and US Treasuries usually fall when global investor sentiment brightens, and this usually coincides with better growth prospects for EM economies.
Our model certainly suffers from some degree of specification bias, similar to the example mentioned above. That being said, this doesn’t mean the model is useless. Its main function is to predict EM asset returns, not to provide causal interpretation of each predictor variable.
Ultimately, the model is just one of several tools to aid us in our investment decision-making process. Fundamental factors such as EM earnings growth, the global macroeconomic backdrop alongside relevant domestic economic conditions will remain important considerations when it comes to thinking about our allocation to EM assets.
So far, the global economy we live in continues to muddle along, with global capital markets remaining predictably if a little disproportionately focused on the precise timing of the next Fed rate hike.
For now, we see some upside to the US dollar and US treasury yields on the back of continuing labour market strength and the return of some inflationary pressure in the US. Tentative evidence of a returning term premium in global sovereign bonds maybe suggests that others are starting to agree.
:: Jonathan Dobbin is head of wealth and investment management NI at Barclays. He can be contacted on 028 9088 2925 or email jonathan.dobbin@ barclays.com
