Despite constant attempts by analysts and the media to complicate the basics of investing, there are only three ways a stock can create value for shareholders:

  1. Dividends.
  2. Earnings growth.
  3. Changes in valuation multiples.

In this series, we drill down on one company's returns to see how each of those three has played a role over the past decade. Step on up, Union Pacific (NYSE: UNP).

Union Pacific shares returned 329% over the last decade. How'd they get there?

Dividends were a big boost. Without dividends, shares returned 261% over the last 10 years.

Earnings growth was incredibly strong. Union Pacific's normalized earnings per share grew 12.7% per year from 2001 until today. That's about double the market average -- an earnings boon shared by other railroads like CSX (NYSE: CSX) and Burlington Northern before it was purchased by Berkshire Hathaway (NYSE: BRK-B) in 2009. Here's how Berkshire vice chairman Charlie Munger described the rail boom a few years ago:

Railroads -- now that's an example of changing our minds. Warren [Buffett] and I have hated railroads our entire life. They're capital-intensive, heavily unionized, with some make-work rules, heavily regulated, and long competed with a comparative disadvantage vs. the trucking industry, which has a very efficient method of propulsion (diesel engines) and uses free public roads. Railroads have long been a terrible business and have been lousy for investors.

We did finally change our minds and invested. We threw out our paradigms, but did it too late. We should have done it two years ago, but we were too stupid to do it at the most ideal time.

There's a German saying: Man is too soon old and too late smart. We were too late smart. We finally realized that railroads now have a huge competitive advantage, with double stacked railcars, guided by computers, moving more and more production from China, etc. They have a big advantage over truckers in huge classes of business.

Back to business. Have a look at Union Pacific's valuation multiple:


Source: S&P Capital IQ.

This is interesting. Union Pacific's P/E ratio has stayed in a fairly tight range over the last 10 years. That's rare. Most large-cap stocks were overvalued 10 years ago, and have seen their P/E ratios drop considerably ever since. Indeed, a fairly large group of stocks achieved earnings growth similar to Union Pacific's over the last decade yet saw shareholder returns go nowhere. The reason Union Pacific shareholders didn't suffer a similar fate was because shares were reasonably valued 10 years ago. It can't be repeated enough: Starting valuations determine future returns.

Why is this stuff worth paying attention to? It's important to know not only how much a stock has returned, but where those returns came from. Sometimes earnings grow, but the market isn't willing to pay as much for those earnings. Sometimes earnings fall, but the market bids shares higher anyway. Sometimes both earnings and earnings multiples stay flat, but a company generates returns through dividends. Sometimes everything works together, and returns surge. Sometimes nothing works and they crash. All tell a different story about the state of a company. Not knowing why something happened can be just as dangerous as not knowing that something happened at all.