By: Mark Hulbert

November 11, 2005, New York Times

The recent decision of Cendant to break itself into four companies reflects an underlying truth about wide-ranging conglomerates: They are sometimes fashionable on Wall Street, but they rarely make much investment sense over the long term.

In most cases, in fact, a widely diversified company would be worth more if its distinct units operated as separate publicly traded companies. Last month, Cendant's directors bet that this would be the case for their conglomerate. They announced that they would break it into four publicly traded companies, each a pure play on just one type of business: real estate brokerage, online travel services, lodging and car rentals.

The board said the break-up was needed to unlock the units' inherent value. This reasoning is in line with numerous studies, which have concluded that conglomerates are worth less than the sum of their parts. Many of the studies are based in some part on the work of the late James Tobin, the 1981 Nobel laureate in economics, and particularly to Tobin's Q ratio, the valuation measure he made famous.

A company's Q ratio is its total market capitalization divided by the replacement cost of its total assets. Companies with relatively high Q ratios are those that investors see as poised for faster growth. While a high ratio is often regarded as an investors' vote of confidence in management, a low ratio is seen as a thumbs-down. Widely diversified companies, on average, have lower Q ratios than those of more narrowly focused businesses - a phenomenon called the diversification discount.

Although the long-term overall odds may be against conglomerates, some can be worth more than the sum of their parts. That is almost certainly the case for Berkshire Hathaway, run by Warren Buffett.

Furthermore, there have been periods in history when investors have reacted more favorably to the combination of vastly different businesses under a single corporate umbrella. During those times, the shares of even run-of the-mill conglomerates have appeared to glitter.

But over the long haul, conglomerates, on average, perform worse in the stock market than the typical focused company. One likely cause is that they tend to do a poor job of allocating capital among their various divisions. Of course, if those units were separate publicly traded companies, the market itself would be making the allocation decisions. And it stands to reason that the overall market is a better administrator in this regard than the average corporate manager.

A study conducted by David Scharfstein, a finance professor at the Harvard Business School, offers evidence of inefficient capital allocation among widely diversified companies. He found that managers of conglomerates generally felt compelled to invest something in all of their divisions, regardless of the divisions' growth potential - a phenomenon that he calls intracompany "socialism." Because of it, conglomerates tend to invest too much in divisions with low growth potential and too little in those with high potential.

Scharfstein's research was conducted for the National Bureau of Economic Research; a copy is available online at http://papers.ssrn.com/sol3 /papers.cfm?abstract-id=226103.

His findings help to explain why conglomerates can unlock value by breaking into their component parts. Once those parts are separate publicly traded companies, the businesses with the highest growth potential should attract capital more easily than the slowest-growth units.

While the decision by Cendant thus makes good investment sense, why did it - or any other conglomerate, for that matter - decide to become so widely diversified in the first place?

A possible answer is suggested in research by the finance professors Owen Lamont of Yale and Christopher Polk of Northwestern University. Writing in the January 2002 issue of the Journal of Financial Economics, they said that companies with the lowest growth prospects tended to be the most likely to diversify. That is probably because such companies see greater growth opportunities in businesses unrelated to their own.

Of course, once a slow-growth company becomes part of a group of unrelated businesses, Scharfstein's socialism kicks in, moving capital out of the fast-growing divisions. If the conglomerate remains intact for a long time, the results may leave investors lamenting its very creation.