What's the difference between an economic expansion fueled by top-line revenue growth and an economy that depends on cost-cutting and extraordinary gains in productivity for its growth?

The first results in a strong economy and stock market that lifts the fortunes of well-run companies and inefficient companies alike, much like we witnessed in the 1990s.

The second results in an economy and stock market plagued by slow growth, relentless global competition and rising prices for energy and raw materials. In that environment, inefficient producers are toast. At the same time, disproportionate shares of corporate profits go to the most efficient companies in a given industry. They're able to use the earnings that come from higher margins to reinvest in new capital equipment and faster new product development.

That's the economy I believe we face for the next 10 years. The result will be a winner-take-all stock market where the shares of sector winners will gradually pull further and further away from the pack.

How do you pick winners in a winner-take-all market? First, it's critical to understand the trends that have created this economic situation. Then, you have to be able to identify the companies that are using research and development budgets to create new products to profit from those trends.

Here they are.

Productivity gains fuel corporate profits

Corporate profits are surging, no doubt about it. According to the U.S. Department of Commerce, pretax corporate profits climbed by 29% in the fourth quarter of 2003 over the same quarter in 2002. For all of 2003, pretax corporate profits grew by 18%.

What's called "productivity growth" played an extraordinarily large role in that profit jump. With output per hour worked up 4.4% in 2003, after a 5% jump in 2002, companies are able to produce more goods and services with the same number of workers, the Commerce numbers show. Or the same amount of goods and services with fewer workers. Either way, that's a huge boost to company profits. (Especially since hourly wages aren't going up significantly right now.)

That kind of productivity growth is way, way above historical trends. The long-term annual average for the last 55 years is just 2.3%.

But a surge in productivity isn't all that unusual as an economy comes out of a recession. Companies have spent the last two to three years cutting workers, shutting their oldest, least efficient plants and trimming fat in everything from office support to shipping to sales. As the economy revs up, that's supposed to lead to increased profits that go into capital spending budgets as companies reinvest in their businesses to meet projected demand.

This is the stage that we've been waiting to see kick in for the last year. Capital spending was up a real (that means subtracting for inflation) 14% in the second half of 2003; Citigroup economists project a 10% rise in the first half of 2004. That's a move in the right direction, but it's still anemic compared with what usually happens as an economy bounces back from a recession. Along with the slow jobs recovery, the slow recovery in capital spending is the key reason that this remains a weaker-than-normal recovery.

Some of the weakness in hiring and capital spending comes from corporate skepticism that this recovery will lead to sustained demand growth. Just as important for capital spending, the recovery of the economy as a whole hasn't yet led all sectors and all companies out of the profit desert.

The recovery has done little to solve the long-term profit problems facing the big telecommunications service companies. Prices for long-distance are still falling, and the big wire line companies are still losing customers to wireless providers and increasingly to companies selling phone service over the Internet. Is it any wonder that companies such as SBC ( SBC) and BellSouth ( BLS) have told Wall Street that their capital spending budgets will barely increase for 2004 from 2003?

Capital spending drives growth

But even in sectors that don't face that kind of long-term problem, the distribution of profits has been uneven. For example, Cisco Systems ( CSCO) earned 18 cents a share in its most recent quarter, up 29% from a year earlier. Not a bad recovery.

But many other companies selling communications gear are still wallowing in red ink. Ciena ( CIEN) lost 16 cents in its most recent quarter and hasn't been profitable since its 2000 fiscal year. JDS Uniphase ( JDSU) has been losing money since its 2000 fiscal year. Tellabs ( TLAB) is still losing money, as it has since the first quarter of 2002. Corning ( GLW) has produced (albeit declining) losses in 10 of the last 12 quarters, including the most recent. You get the picture. And companies that are bleeding red don't increase their capital spending.

Even those companies in the sector that are making money aren't anywhere near as profitable as Cisco. For the trailing 12 months, Cisco's gross margin has been 76% and its net profit margin 22%. That's above the company's five-year averages of 67% and 10%, a tribute to the effectiveness of Cisco's cost cutting. Nortel Networks ( NT), which has been in the black for all four quarters of 2003, shows trailing 12-month gross margins of 51% and net profit margins of 8%. (Assuming expected restatements of its 2003 results don't seriously change the numbers.) Net margin to net margin, Cisco is almost three times as profitable. When it comes to return on invested capital, Cisco has a similar huge edge with a return on invested capital of 16% for the last 12 months to 10% at Nortel Networks. Guess which company finds it easier to finance capital spending out of earnings and finds it easier to justify that investment to Wall Street and shareholders?

And the more that a Cisco, an Intel ( INTC), a Schlumberger ( SLB), a Taiwan Semiconductor Manufacturing ( TSM) outspends competitors on more efficient new machinery, the more that profitability gap will increase.

Open that gap wide enough, and it's almost impossible for a competitor to challenge a more profitable top dog. That's the problem that Advanced Micro Devices ( AMD) faces as it continues to chase Intel. Intel's trailing 12-month gross margin is 73% to AMD's 62%. Intel's net profit margin is 19%, but AMD hasn't shown a full-year profit since 2000.

In 2003, Intel spent $3.7 billion on a new plant and equipment to make the world's most efficient and profitable chip maker more efficient and profitable. AMD spent just $577 million.

This isn't a game that AMD can win. Over the long term, this kind of imbalance results in a winner-take-all industry.

Research and development is the way to win

Companies chasing bigger and more profitable competitors do have a way out of this no-win trap. It's called research and development. Come up with new and better products, as AMD has been doing recently with its Athlon and Opteron processors, and you've got a chance to fight and win against a more efficient, more profitable competitor. Despite Intel's edge in manufacturing and capital spending, AMD has taken market share from its archrival.

Which is why the big battle right now is over research and development spending. And why the best way to separate the winners from the losers in a winner-take-all stock market is to look at this key financial number.

The ferocity of the battle between Intel and Advanced Micro Devices right now shows up in the trend of research and development spending at each company. Intel, through the technology crash and into the recovery, has kept its R&D budget almost constant: $4 billion in 2000, $4 billion in 2001, $4.1 billion in 2002 and then $4.2 billion in 2003.

But as Advanced Micro Devices has seen a competitive advantage created by its newest products, the company has sought to push that advantage to the limit by massively upping its R&D spending. It jumped from $650 million in 2001 to $816 million in 2002 to $870 million in 2003.

That's a very different situation from the telecommunications equipment sector. There, Cisco has been adding to its profitability and capital spending edge over a competitor such as Lucent Technologies ( LU) by building up an equally huge edge in R&D. Lucent, the home to Bell Labs, once one of the top 10 research organizations in all of technology, spent just $1.5 billion on R&D in 2003. I say "just" because the company spent $5 billion in 2000 and $3.5 billion in 2001. Cisco, however, spent $3.1 billion in 2003 on R&D.

The National Science Foundation reported in February that national spending on research and development from all sources rose just 1% in 2003 (after inflation).

That 1% gain includes declines in R&D spending at Ciena (down 14%) and Lucent (down 35%) as well as increases at Microsoft ( MSFT) and Donaldson ( DCI) ( both up 8%) and Johnson & Johnson ( JNJ) (up 12%).

If the research and development gap is big enough and lasts long enough, it will give sector leaders a chance to make already huge market shares even larger, and to lock in those shares for years to come. These companies, in this economy, could indeed take it all.

If you liked this article you might like

The 50 Best Stocks in the World

The 50 Best Stocks in the World

OPEC Demand Drives Oil-Price Surge

OPEC Demand Drives Oil-Price Surge

Five Stocks for the Rocky Times Ahead

Five Stocks for the Rocky Times Ahead

10 Small-Cap Stocks With Big Prospects

10 Small-Cap Stocks With Big Prospects

An Oil Stock for a Grim New Reality

An Oil Stock for a Grim New Reality