U.S. retirement savers are increasingly diversifying into international equities, but are still leaving much of what has to be considered a free lunch on the table.

A new study of 3.8 million U.S. savers in 401(K) retirement accounts over the 2006-2011 period shows a general trend towards better diversification internationally, but with most accounts still significantly under-diversified.

And yet, given the diminishing share of U.S. companies in global equity capitalization, they are, as a group, far less diversified than they ought to be.

"One key fact emerging from the data is that there is an upward trend in the extent of international diversification. We show that part of this, but only a small part, is he potentially rational response to the slowly decreasing importance of the US market in the world equity markets," according to the study's authors, Geert Bekaert and Enrichetta Ravina of Columbia Business School and Kenton Hoyem and Wei-Yin Hu of Financial Engines, a retirement advice company which supplied access to the data. (http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2417976)

The average international equity allocation is now 17.8 percent, with 17 percent holding no exposure at all to the asset class. Given that 64 percent of all global equity capitalization is now outside the U.S., that's a huge underweight.

Diversification in general is critical to investment performance, reducing volatility while increasing returns.

Diversification internationally of equities is particularly important because the U.S. over time, while it may perform very well, will inevitably see its share of global output and equity value decrease. This is not simply about expectations of higher growth rates elsewhere, but also about deepening capital markets in emerging countries and falling costs to access them.

A variety of reasons have been proposed for why savers tend to stay close to home - familiarity, cost, consumption and currency risk, and the feeling that the investor will be at a disadvantage to investors closer to the companies and markets on which they trade.

Some measure of international under-diversification is to be expected, and may indeed be wise. To simply hold shares relative to their global weight exposes the investor to real risks which the benefits of diversification may not fully cover, particularly currency risk.

U.S. investors are saving to meet dollar-based liabilities, and so theoretically dollar assets have an extra value not fully expressed in investment returns.

Still the data on the overall benefits of diversification is very strong. In one study by investment managers Gerstein Fisher covering 1999-2011, a globally diversified portfolio outperformed a U.S.-only one in 96 percent of rolling three-year periods, with a total outperformance of 35 percentage points over 11 years.

Familiarity Bias

Much of the Columbia and Financial Engines study backs up the familiarity hypothesis for why people under-diversify.

For example, having an export oriented factory open up near you seems to be a very good thing for your investment sophistication.

The study found that states with higher relative export components in their economies tended to also see higher international exposure in 401(K) accounts. At the zip code level, living in a place with more foreign-born residents also correlates meaningfully with being more willing to hold foreign shares.