How to invest like Warren Buffett's

Increasingly worried that the Dow Jones Industrial Average ($INDU) is too expensive after reaching a record closing high of 12,961 on April 20? Sitting on cash waiting for the market to correct and yet fearful that the rally will leave you behind?

Well, then, let legendary value investor Warren Buffett's recent purchase of 11% of railroad Burlington Northern Santa Fe (BNI, news, msgs) -- worth about $3 billion -- remind you there is another way to think about the stock market that doesn't involve attempts to predict market tops or bottoms.

The key to Buffett's buy is a measure called return on invested capital (ROIC), which indicates a company's power to grow its business profitably for investors over time. If a company's ROIC is high enough, it puts the massive power of compounding on the long-term investor's side. If return on invested capital isn't high enough, no share price is low enough to justify purchase by a long-term buy-and-hold investor.

The utility of Buffett's logic and the ROIC measure isn't limited to railroad stocks. It's especially well-suited to today's equity markets and increasingly global economy, and it's a great guide for picking stocks for your portfolio when the stock market indexes are at historical highs.

Of course, to apply Buffett's logic you first have to understand exactly why Buffett bought shares of Burlington Northern and two other railroads now.

The fundamental reasons an investor wants to own railroad shares in his April 18 column, "Buffett on the right track: Buying railroad stocks." But although these fundamentals explain why Buffett would want to own railroad shares, they don't explain why he'd buy them at current prices.

At first -- and even second -- glance, the timing seems odd even for Warren Buffett, who says he never tries to time the stock market. Even though Buffett bought his stake in Burlington Northern a few months ago, he was still loading up near an all-time high for the railroad's shares and near an all-time high for the Dow. To make things even more puzzling, he was buying as the U.S. economy seemed to be slowing. Railroads are notoriously cyclical stocks that do well when economic growth is speeding up and head rapidly downward when growth slows.

So, for example, Burlington Northern shares hit a high in April 1999 and bottomed in February 2000 as investors began to anticipate an economic slowdown. Real growth in the U.S. economy was a very robust 3.7% in 2000 and sank to a barely visible 0.1% in 2001.

It is not possible for me to believe that an experienced investor like Buffett has forgotten the cyclical record of a railroad stock such as Burlington Northern. So what was he thinking as he pulled the trading trigger and put $3 billion into this cyclical stock at what looks like a market and economic peak?

In my opinion, he was thinking that, as he has so often reminded investors, he doesn't buy stocks; he buys companies. And if you're buying a railroad company, this is exactly the right time.

Why? Because thanks to all those fundamental factors mentioned in Brush's column, the company's return on invested capital has finally inched above its cost of capital.

For years, railroads, even such recently well-run railroads as Burlington Northern, have earned a lower return on the capital they invest in their business than that capital cost them in the marketplace. That's not a recipe for bankruptcy -- a company can chug along in this position for pretty much forever because, thanks to operating leverage, it doesn't have to invest a dollar of capital to produce a dollar in earnings.

But earning less than the market cost of your capital sure doesn't encourage reinvesting profits in your business: The return is so low that it's more attractive to pay dividends or buy back shares. And that eliminates the power of compounding from the company's growth. Sure, the company can throw off lots of cash, as railroads do, but since they aren't reinvesting that cash, the business and the cash flow aren't growing at a compounded rate every year.

Standard & Poor's puts the weighted average cost of capital for Burlington Northern at 8.5%. The company didn't make that on average over the past five years. Look at the company's investment returns on our site. The five-year average is just 4.5% for the company and the railroad industry.

What our data don't show is that Burlington Northern and railroads in general haven't earned a return on invested capital above their cost of capital for considerably longer than five years. The last time these businesses were consistently that profitable was in the early years of the 20th century, roughly 90 years ago.

But that's been changing recently. Return on invested capital has climbed to 6.8% over the past 12 months. And in the past few months, return on invested capital has grown even more, so that it is either close to or actually exceeds the company's cost of capital. Railroads as a whole are riding this trend. In the past few months, Canadian National Railway (CNI, news, msgs) and Norfolk Southern (NSC, news, msgs) have also recorded returns on invested capital near their costs of capital. CSX Corp. (CSX, news, msgs) and Union Pacific (UNP, news, msgs) seem likely to pass this mark in the next few quarters.

So Buffett bought into the railroad sector just at the point where railroads could put the power of compounding to work: reinvesting this year's earnings and earning a high enough return to justify investing the earnings on those reinvested earnings next year.

For a long-term investor like Buffett, that turning point in a company's (and sector's) fortunes is more important than any temporary spike in the stock price -- because reinvesting and compounding each year's earnings grows the business at a rate that, in the long term, overwhelms any effect of buying at a temporary top.

This analysis -- and I readily concede it's only my reconstruction of what might be Buffett's thought processes -- would lead me to prefer Union Pacific to Burlington Northern. The latter railroad has raced ahead at the front of the ROIC pack because the company invested during the years when it was tough to justify investing in double-tracking, which allows trains going in opposite directions to pass each other more easily. The company now has far more of its routes double-tracked than a competitor like Union Pacific. That makes Burlington Northern more efficient now, but it also means that Union Pacific has more future opportunity to reinvest capital playing catch-up.