PsyFi Search


Sunday, 23 October 2011

HONTI #5: Home Is Where The Risk Is

Rule #5: Reduce your risk by diversifying your investments internationally.

Familiarity Is A Poor Basis For Investment

Home bias1 afflicts almost all investors, of all shapes and sizes. It’s the tendency to invest in firms that are close to home in preference to firms that are far away. This is true in both a geographical and a mental sense, and the net effect is that many portfolios are unsafe, under-diversified and unbalanced. A bit like most of us investors, really.

It isn’t any great surprise that people prefer to invest in stuff that they’re familiar with2. Indeed there’s a strong argument that we should only invest in things we understand – the firms within our personal circles of competence. Unfortunately merely visiting your local supermarket on a regular basis doesn’t mean that you understand the global logistics and brand marketing that lies behind shelves bulging with produce from all over the world. Competence, in investment terms, means understanding how a business makes its money, not being an eager consumer of their products or services.

The tendency to extrapolate from personal experience to broad generalisations is the stock in trade of many media commentators, but it’s a capital mistake for investors. Personal experience is not a good basis upon which to make sweeping judgements but it’s something we’re evolutionarily built to do for good survival reasons3.

For similar reasons familiarity is a powerful determinant of personal judgement: we tend to cleave to that with which we’re most familiar. In investment terms, however, this tends to have a nasty side-effect, which is that if we only invest in companies we have some knowledge of then we’re going to create portfolios with oddly skewed risk profiles. As of 2007, US investors still held over 80% of their portfolios in US based stocks, despite clear evidence that an international spread provides significant risk diversification4

The worst case, by far, of a home-bias affliction is the case of people who are heavily invested in their own employers. Often this is because of stock option or preferential stock purchases schemes, which sometimes have significant tax advantages. Clearly people would be foolish to pass up a free lunch, but simply working for an organisation doesn’t mean you understand the business. Ending up in a situation where your salary, your pension and your investment portfolio are all dependent on one single company isn’t a wise position to get yourself into. Sadly it’s exactly what happened to a lot of innocent employees when Enron and Worldcom went bust: working in one part of a large organisation doesn’t tell you much about its inner stability5.

More generally home bias has been found in all sorts of situations. After AT&T was forced to spin off its regional telco’s, the so-called Baby Bells, research uncovered that the geographical distribution of shareholders was correlated with the regional coverage of the companies6.  Regional investment preferences have also been demonstrated in other ways. So, for instance, there’s a demonstrable bias that local newspapers preferentially report stories about local businesses and that this reporting, rather than any fundamental changes in the company, drives trading in the stock of the company7. This would only happen if local investors are heavily invested – implying, once again, quite strongly that they’re investing at home on familiarity grounds.

This bias, though, is more than a tendency to invest in what we know. We also expect that firms we know about will yield higher returns than firms we don’t know about: which is ridiculous, if you think about it. The side-effect of this is that companies which are more successful will tend to be more familiar: but investing in companies which are already successful will tend to be putting yourself on the wrong side of mean reversion – after all, if you only invest in the next Big Thing after it’s become the next Big Thing then you’ve already missed a large part of the growth8.

You might think that this is purely a problem for amateur private investors, but it’s not. It appears that the same biases that drive private investors occur in the professional sector: they feel that they have more knowledge of their own market and are over-confident about their ability to generate higher returns9.  Unfortunately it also turns out that they’re no better than anyone else at forecasting their home markets. In addition, their overconfidence means that they do less fundamental analysis of the companies that they’re familiar with, relying on other sources of information – with predictable consequences10.

Despite these issues anyone investing internationally in 2008 wouldn’t have seen a lot of difference to their returns regardless where they were invested, as pretty much everything, everywhere, fell together. However, a recent study has shown that if you take the long view then avoiding home bias saves you from the worst possible mistakes:11
“International diversiļ¬cation might not protect you from terrible days, months, or even years, but over longer horizons (which should be more important to investors) where underlying economic growth matters more to returns than short-lived panics or global coordinated events, it protects you quite well.”
The point made is that although there are occasional global crises there are far more localised country or regional specific ones. International diversification is, in the long run, both prudent and necessary – and in an age where acquiring such a spread can be achieved by buying a few ETFs there’s simply no excuse for ignoring it.

Notes to the article:
  1. The classic paper on this is Investor Diversification and International Equity Markets by Kenneth French and James Poterba. The idea is that there are "information asymmetries" - that's lack of knowledge to you and me - about foreign markets, and that there are underlying behavioral biases linked to familiarity which both drive the effects, although how these interact still isn't really understood.
  2. See, for example, Fooled by Fluency, which looks at the mere exposure effect: merely exposing people to a new idea makes them more likely to prefer it.  It's why advertising works and very annoying it is too.
  3. In particular we tend to overweight the probability of rare events, as in Investor Decisions - Experience Is Not Enough and to generate illusory patterns out of limited data and random events, as in Deep Time and the Fallacy of Frequency.  
  4. Statistics from Home Bias in Open Economy Financial Macroeconomics which also gives an up-to-date, although extremely complicated, explanation of factors implicated in home bias.  Probably not one to read unless you're feeling particularly brave.
  5. The sad case of innocent Enron employees was covered in Puke: Don't Invest In The Familiar. Which also helpfully explains why you shouldn't eat Indian food when feeling a bit queasy.
  6. Gur Huberman's paper on Familiarity Breeds Investment: "This paper offers a novel explanation for the home country bias: people simply prefer to invest in the familiar".  Sometimes the best explanations are the obvious ones ...
  7. This is a taken from an elegant piece of research from Joseph Engleberg and Christopher Parsons, discussed in The Language of Lucre.
  8. The idea that home bias is linked to investor overconfidence about their ability to assess home based stocks is developed by Norman Strong and Xinzhong Xu, although they do point out that it's not possible to figure out whether people are overconfident because of home bias or have home bias because of overconfidence.
  9. Lots of research on this, starting with Home Bias at Home by Joshua Coval and Tobias Moskowitz although I especially like this piece by Jungwon Suh.
  10. Quite extraordinary stuff, this, in What Drives Home Bias?  The researchers show that home based and home biased institutional investors base their expectations about getting above average returns on chart analysis, technical indicators and the views of economic opinion leaders rather than actually crunching numbers or asking hard questions of managements.  It's only mildly surprising that they don't use chicken entrails and astrology.
  11. International Diversification Works (in the Long Run) by Clifford Asness, Roni Israelov and John Liew.   Well worth reading.


  1. Thanks for the article
    There are also good reasons for investing at home.

    1. Taxes and fees
    As as French investor I can invest in all European listed companies in a tax-deferred account. But in in practice broker fees are much lower for Euronext Paris (for my broker ar least). For US-listed stocks, it's ~minus 33% in taxes on each $ I make.

    2. Correlation
    I've tried investing in ETFs (CAC40, DAX, MSCI Emerging Markets) for a while. They're highly correlated especially during downturns.

    3. Competition
    I may have an edge for French small caps (for a start, their accounts are published only in French, which rules 99% of world population) (ok maybe a little less). What's my edge on the LSE or on the US markets ? The competition from English-reading investors from all over the world is much more intense.

    4. You can get a large international exposure from domestic stocks. LVMH sells handbags in China and emerging countries. 75 % of Air liquide sales are made outside France. The same thing applies of course for export-oriented German or English or US large caps.

    Anyway, again thanks for your blog, I'm an avid reader.

  2. Good article. Rightly pointed out that we do not look outside. Primary rule for great investors like Warren Buffett: If you don’t understand it, don’t do it. –

    Investment itself is complicated as we do allow people to drive a vehicle without a driving test but allow them to enter complex financial world without any formal education. I think that is what keeps people from investing home

  3. Hi Caque

    In respect of ETFs, which are the only sensible way for most investors to look at foreign investments (mainly due to the points you raise about costs), the main point is that the last few years aren't typical. Historically country specific investments have been non-correlated, leaving aside rare events like the Wall Street Crash and the Sub-Prime Subsidence. I think we're in danger of thinking the last few years are typical of economic downturns: of course, globalisation means that they might be, but history suggests that they're not.

    In terms of small caps I'd absolutely agree: if you can find an edge in analysis then that gives you a head start: it's just that when everything French goes down then so will the small caps. I know my edge overseas is minimal but I still feel that some overseas exposure decreases my overall risk.

    As for multi-national stocks: the case isn't proven. There's some evidence that company performance is heavily correlated with its home country even if it's internationally exposed: this seems to be related to capital requirements and legal systems. However, this research is nearly a decade old, and globalisation has moved on so although I thought about including it in the end I chose not to. I also think it would be very difficult to assess how much diversification you got from LVMH or Sodexo or whatever compared to buying an country specific ETF.

    So I don't disagree with anything you've said, but there are arguments for and against ... as always :)