The problem of home bias is a hoary old chestnut in the world of investment. Pick a country, and find investors' equity holdings inordinately skewed to domestic securities, often far out of proportion to global benchmarks.

Hong Kong pension funds, for example, allocate on average 28% of their equities to local stocks, when Hong Kong comprises only 1% of global indices. This is hardly unique: American institutions put up to 85% of their equities allocation into domestic stocks, when 50% would be more appropriate. Australia makes up 2% of a global equities index but superannuation funds invest over half their stock allocations domestically. And Japan, 8% of the global equities world, accounts for 59% of Japanese pension funds' equities allocations.

There are some good reasons for this: investing abroad carries currency risk; pension funds and insurance companies' liabilities are denominated in the home currency; local laws limit offshore investment; foreign laws keep their capital out. But there are also questionable reasons for this: a false sense of security and familiarity with local stocks.

At the same time, the traditional idea that global indices are based on - diversification by country - has lost some of its oomph now that global sectors such as technology have become more important. If you're investing in tech companies in your home country, perhaps international diversification loses some of its meaning.

But most markets also have concentrations of particular sectors. Think of Finland, which is dominated by Nokia. Or Hong Kong, where diversified industrials such as Hutchison Whampoa and property companies rule the roost. Pension funds based in such countries, with a home bias, not only fly in the face of global equity indices by country, but also by sector. Either way - from a traditional geography point of view, or a newish sector perspective - pension funds around the world are suffering from home bias to an extent that may not be healthy. And the global index vendors and investment consultants have failed to find a way to persuade pension funds of this.

A modestly sized money manager out of Sydney, however, has launched a concept that addresses home bias in a startling way. Constellation Capital Management's strategy, HomeGlobal, actually argues that pension funds don't need to increase their international exposure to bring reason to their equity allocations: in fact, in most cases they should probably increase their domestic exposures, but in an intelligent way that adjusts for sector imbalances.

"What we're saying is pension funds should invest locally when they can, and go overseas to obtain the required sector exposure," says Douglas Little, managing director at Constellation.

He calls HomeGlobal a country-specific global benchmark, customized for investors in each country to help them invest overseas without home bias.

For example, Hong Kong pension funds investing domestically will have a big exposure to industrials and property companies, but won't have anything in pharmaceuticals, natural resources or IT hardware. If these pension funds simply follow traditional, geographic global indices, they may pick up pharma exposure but they'll also pick up more property companies, which clearly they don't need.

HomeGlobal recognizes that the Hong Kong market is weighted toward property companies by 19%, versus 1.7% globally. So an offshore allocation should avoid property companies, and target oil and gas companies instead.

Constellation calculates that Hong Kong investors can actually get 43% of the global sector universe via domestic investment. That means that the current average of 28% of pension money invested domestically is actually too low. Hong Kong pension funds can invest 43% domestically, if their benchmarks were able to account for global sector correlations. "Don't go overseas if you don't have to," Little says.

It appears to be an elegant solution to home bias - and one that will sorely test the traditional index vending community. Little reports that initial, undisclosed talks with fund managers and investment consultants have sparked interest. The strategy was formally unveiled last week in Tokyo at a conference sponsored by Nomura Securities and the response was warm, he says.

Constellation is initially targeting Japan because it is a large market with a comparatively high level of home bias.

Little realized trying to pitch pension funds directly was likely to be futile, so he has manoeuvred his firm into the existing investment process by teaming up with FTSE and licensing HomeGlobal to it. He approached FTSE in 2001 because he saw it as an innovator, having moved early to free-float adjustments and customized indices.

He is aware that it will take time for HomeGlobal to take off with the global investment world. In the meantime he will begin to market it to retail mutual funds companies and index investment specialists to get them to build products off of HomeGlobal's index method. But this little Aussie outfit has issued a challenge that the world must, at some point, meet.