22-07-15 9:15 | Markets and investments | Fredrik Wilander

In my previous blog post , I wrote about the investment behavior of private individuals and noted that it is not uncommon that it is characterized by under-diversified portfolios, excessive trading and a preference for owning domestic equity (the home bias).

A home bias may not only be the domestic equity market, but even closer to home. It could for example be investing in companies headquartered in your town, or even in the company where you work. The latter may for example be through an employee stock ownership program.

As far as I know, there is not much research documenting the performance of private investors when they invest very close to home. One exception is a 2005 study by Ivkovic and Weisbrenner in Journal of Finance, who utilize the same data set as Barber and Odean used in their 2001 "Boys Will Be Boys" study; individual accounts at a large US retail broker. They find that private investors earn higher returns on local small-cap investments than non-local small caps, indicating some sort of information advantage with investing in small companies close to home.

At the same time, we often anecdotally hear "horror stories" of when things do not work out well, as in the recent Swedish case of Northland Resources, a mining company located in northern Sweden. Its bankruptcy supposedly left many employees both unemployed and deprived of their financial savings, which had been invested in the company stock. The very local version of the home-bias is a case in point when it comes to portfolio diversification if one includes the value of assets such as real estate and human capital in the "household portfolio".

Recent research (The In-State Equity Bias of State Pension Plans ) points to that also institutional investors fall victim to the desire to invest close to home. By examining 27 U.S. state pension plans who manage their own equity portfolio, the authors document that state pension plans substantially over-weight stocks that are headquartered in the same state, despite overall holding well-diversified portfolios. The average allocation to in-state company stock is about 10%, versus the 5% that would have been the allocation had the plans held a portfolio mimicking the value-weighted market portfolio.

This in-state bias is substantially larger than that of regular institutional money managers, which was documented in a2010 study . The in-state bias is especially interesting for state-pension plans. As noted by the authors, state pension plans should probably under-weight in-state companies " because each state's economic activity, tax revenue, and the income of state residents is more positively correlated with the performance of in-state stocks compared to other stocks".

As such, for the state of California to over-weight Silicon Valley tech companies in their state pension plan is a bit like the Swedish Northland case. Luckily for the pension plans, there seem not to have been to many Northlands, at least not so far; also in this study the investment performance is better for local investments than out of state investments. Interestingly, by about the same magnitude as in the 2005 study on private investors.

Substantial academic evidence points to that holding the value-weighted market portfolio is not efficient; there are numerous ways to "tilt" an equity portfolio to improve returns. Using information advantages to invest in local small cap stocks may be one of those. But I would underline the world tilt, meaning a slight over-weight. Not placing one's entire life-time savings in the town's corporate pride. For that, the risk for the "household portfolio" is simply just too great.

U.S. state pension plans favor local companies ( Source: Brown et al. 2015)


distributed by