Think local, act global: a plea for global equities

Post with image
Wouter Weijand wrote an interesting post on residential property entitled “Home Bias”. However, it was not originally for the real estate asset class that the term “Home Bias” was coined. It was first discovered and used with regard to equities by Kenneth French and James Poterba in 1991. The general idea is that investors tend to, for no rational reason, overweight stocks from their own region, although it’s detrimental to diversification. Almost a quarter of a century later, even institutional investors are still usually strategically overweight their own domestic stock market. The same could be said for retail investors, who often consider international equity as an “alternative investment”, although they will systematically subscribe to the latest local IPO.

Home bias is therefore a strong and resilient effect. It works even within one country, and is not a residue from inherited investments. For instance, when AT&T had to spin off its regional telcos, research showed that the geographical distribution of the new shareholders was correlated with the regional coverage of the companies.

Behavioural finance explains this bias as a form of familiarity bias: people like what they know, even when what they know is bad. As politicians and communicators often say a bad reputation is still better than no reputation at all… Overconfidence, which is the tendency to think that we can make better use of information than we actually can or that others can, also plays a role. To put it in a nutshell, our arrogance tricks us into thinking that we can outsmart others on things that we know, although we have no specific information advantage. To some extent, there is a similarity with the low-volatility anomaly: local stocks, like volatile stocks, attract more attention and more investment inflows than they should.

The consequences of the home bias are manifold, apart from the obvious local overweight. Here are some:

  • As explained in this post, emerging markets tend to get more than their fair share of speculative and indexed flows, since most investors from developed countries only consider these markets as an “alternative” solution.
  • Investors tend to structure their equity portfolio by regional allocation. Although sectorial or factorial breakdowns would be less correlated, and more coherent with the asset class level breakdown, you need a regional allocation to overweight your neighbourhood.
  • Consequently, investors keep on asking for regional versions of quantitative stock-picking strategies.

This last element is important to us, as it multiplies the number of local versions we are required to run for quantitative processes that are actually global by concept. It is not only costly to run many portfolios for limited client bases, but it also destroys value for investors: When you have found a stock-picking process that has some interesting features (whether it’s a good return, a good sharpe ratio, a good information ratio or anything else), it seems obvious that reducing the opportunity set, by fracturing the investment universe, will reduce these features and increase the cost of risk constraints.

So, continue to think local, but act global!

Etienne Vincent

Global Head of Quantitative Management

Leave a reply

Your email adress will not be published. Required fields are marked*