Looking at a deal in wholesale distribution? Given the sector's size and today's low trading multiples, chances are that someone in your firm is exploring the fundamentals of a wholesaler who delivers complex valves to oil refineries or a middleman who ships hamburger buns to your local family eatery.

Even in a flat year like this, $7 billion worth of wholesale distribution deals have closed, from acquisitions to minority stakes in sectors as diverse as restaurant supplies and specialty steel.

It's too bad that most will generate poor returns. Over the past five years, only 20% of public wholesale distributors surveyed by our firm managed to beat the S&P 500. In fact, most actually destroyed shareholder value.

Time to blacklist wholesalers? Not so fast. The Internet was supposed to be the death of the middleman. But there are still about 900 companies in the steel-distribution sector alone, and estimates put the universe of all U.S. distributors at more than a quarter of a million.

Hidden in there are great performers. Our research shows some companies' shareholder returns are 1,000% better than the worst of the bunch. So the question becomes: How do you identify those stars?

Of course strategy differences drive varying returns for competitors within a segment, be it steel, pharmaceuticals or some other. But we suspected the structure of a segment could itself be a factor, so we launched a research project, tracking a cross-section of segments including drug, steel, chemicals, food service, office supply and grocery.

The findings are intriguing. They show you can explain almost all of the variances in returns between segments by using two rules of thumb: the shape of a segment's supply chain and, to a lesser extent, the percentage of products bypassing wholesalers entirely.

Note what is missing: Contrary to popular belief, traditional financial measures such as margins and cash flow are not key drivers of differences in returns between segments.

But consider the shape. If the supply chain looks like an hourglass ù if there are more suppliers at the top and retail customers at the bottom than wholesale distributors in the middle ù that's good news for investors.

It means that the value of the role played by the middleman is high. Customers can buy from only a few distributors, and suppliers have limited channel choices.

A for-instance: Companies in the drug wholesaling segment, which has a ratio of suppliers to distributors ù a leverage ratio ù of almost 24 times, have yielded five-year shareholder returns that average almost 350%.

By contrast, in a low-ratio segment such as steel or chemicals, competition between distributors is brutal, and often they must take on added capital-intensive investments to convince customers not to order directly from suppliers.

Direct ordering is the other part of the story. We have found that the percentage of shipments going directly from supplier to end customer is an excellent proxy for the degree of a distributor's value-added from the customer's perspective. Our estimate is only 43% for drug wholesaling, but 80% in office supplies.

Investors should also track how those metrics change over time. Has the waist of a sector's hourglass shape been getting thinner or thicker? And what's happening to the distribution-direct rate?

The pharmaceutical sector has gotten skinnier and busier. A decade ago, it had a 15-times leverage ratio and a 50% distribution-direct rate, for a 150% shareholder return.

Perhaps you'll want to think twice about entering a sector with decreasing ratios.

These days, the grocery segment is one to watch. Consolidation keeps increasing the leverage ratio, although direct shipments are up slightly as the large retailers have moved distribution in-house over the past decade.

The reverse of that trend—retailers shedding their distribution assets—would be an important signal that returns of companies in the segment may increase.

The tools work well in signaling the success of roll-ups. Most roll-ups to date have been too small to improve a segment's fundamentals.

Take Metals USA Inc., which went public immediately after its creation from the 1997 merger of eight steel distributors. The steel distribution segment's leverage ratio of 0.1 should have been the telltale sign; today, Metals USA has a negative 80% five-year shareholder return.

One message is clear: Do not attempt to roll up a segment that does not have an hourglass shape. It is simply too hard to consolidate enough companies to bring the leverage ratio above 1, particularly if suppliers are consolidating as well.

In fairness, we're not saying that investors should avoid all segments with poor leverage ratios and distributor-direct metrics. But we are saying that successful investments in these segments are the exception and not the rule.

The segment-shape tools are one critical way of sorting one kind from the other.