In the 2012 Berkshire Hathaway (NYSE:BRK-A) (NYSE:BRK-B) annual letter to shareholders, Warren Buffett talked dividends.

He noted that many shareholders didn't understand why the company so highly valued the dividends paid to Berkshire by its subsidiary companies even while Berkshire itself does not pay a dividend. 

Buffett's answer and his application of dividend philosophy paint a picture of a truly masterful dividend investor. Let's examine what Buffett said, and how he puts dividend investing into practice.

What kind of investor is Buffett, really?
Going on conventional media coverage alone, you might believe Warren Buffett is the world's greatest value investor. His $71 billion net worth was built by paying $0.50 for companies worth $1 or more. It's a strong argument.

However, studious investors should also note the nuance to Buffett's value style: he will pay a fair price for an exceptional company. His 2009 megapurchase of the Burlington Northern Sante Fe railroad is one example of this. The acquisition of Heinz brands in 2013 is another.

And then there are dividends. Even a quick glance at Berkshire's top holdings reveal a basket of stable, healthy dividend stocks. Wells Fargo (NYSE:WFC), Berkshire's top holding at 23% of the portfolio, pays a respectable 2.5% dividend. 

Or consider Berkshire's fourth-largest holding, IBM (NYSE:IBM), which constitutes over 11% of its equities portfolio. Big Blue's dividend currently yields 2.7%, and the company has paid a quarterly dividend continuously since March 1916. 

Companies that will still be around in 100 years
The examples continue from the top of Berkshire's portfolio to the bottom, and the reason is simple. 

A fundamental tenant to Buffett-style investing is finding companies that will still be around 100 years in the future. Buffett says these companies are protected by a "wide moat."

They have exceptional brands, products that can't be replicated, and strong market share positions. Companies such as the leading soft drink brand in the world, Coca Cola (NYSE:KO), railroads such as Burlington Northern Santa Fe, or a leading company that makes loans to allow individuals to buy homes for themselves and their families.

These companies are stable. They are big. Oftentimes they are pretty boring. 

Boring companies are predictable; they won't command outrageous price-to-earnings ratios like other high-flying, high-growth companies. Those more reasonable prices make these companies attractive from a value perspective, but also from a dividend yield perspective. 

Cause or effect?
In the 2012 letter to investors, Buffett said paying a dividend makes sense when a company's profits cannot be reinvested into the company at an acceptable rate of return. If that rate of return is adequate, as Buffett believes is the case at Berkshire Hathaway, then it makes more sense to not pay a dividend and reinvest profits fully.

For large, stable, wide-moat company like Coca Cola or Wells Fargo, payment of a dividend can be inevitable. The companies are so successful that they can't produce enough internal returns to justify reinvesting profits fully. They are already highly efficient with large market shares; at a certain point, those internal investments hit a point of diminishing returns for future value creation. The profits they do retain are more about maintaining the machine than reinventing the wheel.

In this way, Buffett's dividend investing can be viewed as a byproduct of investing in these types of wide-moat companies. In the letter, Buffett concluded his discussion of dividends by saying,

Most companies pay consistent dividends, generally trying to increase them annually and cutting them very reluctantly. Our "Big Four" portfolio companies follow this sensible and understandable approach and, in certain cases, also repurchase shares quite aggressively.

We applaud their actions and hope they continue on their present paths. We like increased dividends, and we love repurchases at appropriate prices.

Perhaps the best part of this approach is how easy it is to replicate. You don't need to be a stock market guru or rocket scientist to find winning dividend stocks. Simply look for companies that operate a business or sell a product you think will be around for the next 100 years. When you find the right company with a "wide moat," the dividend will hopefully follow.