SHRINKING HOME BIAS IN INTERNATIONAL

PORTFOLIO INVESTMENT

While it is generally agreed that an internationally diversified portfolio performs better, many countries hold puzzling low levels of foreign assets. The difference between optimal and observed (far too low) holdings of foreign stocks is known as the equity home bias: a costly and worrisome behaviour that has been puzzling economists for quite some time. But a fresh look at the puzzle by Crina Pungulescu and colleagues actually reveals that the home bias is lower for some ‘deserving countries’.

A prime issue when measuring deviations from optimal international investment, is the choice of benchmark – the ‘correct’ mix of domestic and foreign equity. Traditionally, each country is expected not to hold a larger proportion of domestic assets than its own share in the world market wealth.

That means that US investors, for instance, should not hold more than 40% of their portfolios in domestic equity. For Japan, the threshold is 20%, for the UK, even lower, at 10%, and as the size of the market capitalisation within a country decreases, the optimal portfolio mix contains even lower shares of domestic assets.

No country raises to the challenge. Observed domestic holdings are several times higher than these thresholds.

One might question the model that gives such a strict benchmark for optimal portfolio weights. There is plenty of room for reasonable doubt, considering that the model – the International Capital Asset Pricing Model (I-CAPM) – is derived under very unrealistic assumptions.

Recent methodological advances allow us to turn the available data on asset returns into a viable alternative to using the I-CAPM. A new optimal investment benchmark can be obtained from this data.

By involving data on returns in domestic and foreign markets in computing this alternative benchmark, this study lets the performance of home versus host markets play a role in explaining the investors’ international portfolio decisions. An investor that is interested only in the risk and reward of his portfolio, may be more attracted by gains from investing in his (well performing) home market than by the wisdom of the I-CAPM. Using an alternative benchmark, that accounts for country performance, leads therefore to a different measure of home bias than the I-CAPM measure.

The researchers compare the traditional I-CAPM home bias with alternative measures that use data on asset returns in a group of 25 countries over several decades. The results show that in the cases where the I-CAPM was not appropriate, alternative measures reveal substantially lower levels of home bias.

Many countries are in this situation, including Austria, Australia, Belgium, Denmark, France, Iceland, Italy, the Netherlands, New Zealand, Sweden and the UK. In some countries, such as the Netherlands, Denmark and Iceland, the alternative measures ‘correct’ for about a third of the home bias puzzle. The home bias puzzle is substantially smaller for these well-performing, ‘deserving’ countries. In other words, their investors are not as wrong as traditional literature would suggest.

Moreover, the researchers find that the bias decreases sharply in recent years for many countries. They link this welcome development to the larger process of integration, which shapes so drastically the world markets in recent time.

Indeed, home bias appears to decrease faster in countries that are better integrated in the world market. For instance, in the euro area the speed at which the bias is resolved is higher by 7-10% per year compared with the rest of the countries in the sample.

These findings give confidence that time working towards market integration works also towards eroding the home bias puzzle.

ENDS

Notes for editors: ‘Model Uncertainty, Financial Market Integration and the Home Bias Puzzle’was presented at the Royal Economic Society’s 2007 annual conference at the University of Warwick, 11-13 April. The paper is forthcoming in the Journal of International Money and Finance. The authors and their contact details are as follows:

Lieven BaeleTilburg University, Department of Finance, CentER for Economic Research and Netspar, P.O. Box 90153, 5000 LE Tilburg, The Netherlands, email: , phone: +31 13 466 3257.

Crina Pungulescu, Department of Finance and CentER for Economic Research, Tilburg University, P.O. Box 90153, 5000 LE Tilburg, The Netherlands, email: , phone: +31 13 466 2020, mobile: +31 64 283 4648.

Jenke Ter HorstTilburg University, Department of Finance, CentER for Economic Research and Netspar, P.O. Box 90153, 5000 LE Tilburg, The Netherlands, email: , phone: +31 13 466 8211.

For further information: contact the authors or Romesh Vaitilingam on 07768-661095 (email: ).