About 10 years ago, before he passed away, my grandfather showed me something that forever changed my approach to investing.

One evening from his office drawer he pulled out a dividend check bearing the name of a prominent Canadian company. As he explained, the annual dividend was worth half of what he paid for the stock. 

Of course, that didn't happen overnight. I think my grandfather had been holding onto those shares since the late 1960's. But the point is that even a modest yield can become a cash-flow machine if given enough time. 

Unfortunately, many investors reach for the highest-yielding stocks they can find. But as my grandfather taught me, when building a stream of retirement income, it's not yield but dividend growth that really matters. It's about finding a cash flow trickle that could one day become a mighty river. 

Chevron (CVX 0.37%) is a great example of what the power of compound growth can do for a stock's yield. Over the past decade, the company has increased its dividend at a 10.6% compounded annual rate. If you had bought and held the stock over that time, the yield on your original investment would be 11.5% today.

Take a look at the chart below to see what I mean. This example assumes you purchased 100 Chevron shares at around $34 near the beginning of 2003.

The magic of compounding

YearDividend Per ShareTotal DividendsYield On Cost
2013 $3.90 $390 11.47%
2012 $3.51 $351 10.32%
2011 $3.09 $309 9.09%
2010 $2.84 $284 8.35%
2009 $2.66 $266 7.82%
2008 $2.53 $253 7.44%
2007 $2.26 $226 6.65%
2006 $2.01 $201 5.91%
2005 $1.75 $175 5.15%
2004 $1.53 $153 4.50%
2003 $1.43 $143 4.21%

Source: Yahoo! Finance

Let's continue this thought experiment for another decade. Assuming that Chevron can continue to grow its dividend at a 10% compounded annual rate, by 2023 our yield on cost would be 30%!

But it gets even better. Over the previous 10 years, Chevron also repurchased 10% of its outstanding shares. This increases our stake in a wonderful business and is something we can expect to continue in the upcoming decade. 

But can Chevron realistically maintain this growth rate? Actually, it's not really a crazy assumption. Many of the company's big investments, such as the Alberta oil sands, are just beginning to bear fruit. Chevron estimates that total production will grow 27% by 2017.This should generate plenty of cash flow for additional dividend hikes and share buybacks. 

Will most investors do this? No, probably not for a couple of reasons reasons. 

Sucker punched
First, investors are often suckered in by high payouts. Take Encana (OVV 0.23%), for example. Earlier this year, the natural gas producer sported a yield north of 4.5% financed by debt. But today, low commodity prices and the need to conserve cash for its reorganization have forced the company to cut its distributions. Shareholders have learned that it's far better to have a sustainable yield that can grow over time. 

Second, it's boring. Compounding growth over decades is about as fun as folding laundry. It's far more exciting to follow today's sexy names like Facebook, Twitter, or BlackBerry. But I think we all understand intuitively that excitement and profits usually don't go together. 

Chevron is evidence that when it comes to income investing, dividend growth is far more important than a big yield. As my grandfather taught me, even a humble yield can become an income torrent if given enough time.