Some investment rules are so powerful they can't be rebutted with a straight face. That companies with higher dividends tend to outperform those with lower dividends is one of them. ... Countless academic studies show that dividends are typically the best way companies can reward shareholders.
Oh really, Morgan? I beg to differ. Them are fightin' words. And my face is as straight as the Bonneville Salt Flats.
Correlation is not causation
Morgan claims that dividend payers have historically outperformed nondividend payers. This is absolutely true. I do not dispute this. (Don't worry...the article isn't over.)
What I dispute is that these companies outperformed because they paid dividends, and that therefore managements are putting their shareholders in "peril" by not paying more in dividends. Morgan's academic studies prove that dividend payouts and outperformance are correlated, not that dividend payouts cause outperformance.
And there is a huge difference between correlation and causation in practice.
It's been noted, for instance, that ice cream consumption and crime are strongly correlated. Does this mean eating ice cream causes crime?
No, of course not. It's more likely that there is a third variable at work...like the fact that criminals might prefer to work when windows are open and people are out and about, which happens to be during the summertime when people consume ice cream.
So banning ice cream wouldn't lead to less crime (just more unhappy children).
Similarly, it does not follow that paying more dividends will cause a company to outperform. There could be a third variable at work that explains why dividend payers outperform.
And I think there is.
Companies that pay dividends
Think for a moment about the kind of companies that pay consistent dividends: They're often leaders in their field, possess durable competitive advantages, and have strong and ethical managements.
I'd argue that it's these Buffett-like characteristics -- and not the dividend policy itself -- that causes dividend payers to outperform. Having the free cash flow to be able to pay consistent and increasing dividends is often indicative of a great Buffett-like company. I think this is the real reason why dividend payers outperform.
After all, there is no rational reason -- all else equal -- why changing the dividend policy should lead to outperformance. As I've written about many times, a stock's price falls by the amount of the dividend payment when the stock goes ex-dividend.
To put it another way, giving Research In Motion Cola-Cola's dividend policy won't make it perform like Coca-Cola. Nor will it change the fact that Research In Motion is in trouble (though I think RIM is undervalued and is therefore still a CAPScall).
Similarly, giving Coca-Cola RIM's dividend policy won't change the fact that Coke is an excellent company (and is also a CAPScall). And assuming Coke managed the company in the same way, it won't change the total return given to shareholders. (Er... OK, it would a little bit -- by eliminating double taxation of dividends it would boost after-tax performance.)
Morgan's excellent P/E comparison
My hypothesis -- that dividend payers outperform because they're better companies, not because they pay dividends -- is further validated by Morgan's observation that stocks with higher dividend payout ratios have higher P/E ratios: We know that companies with higher P/Es tend to be better quality names.
Morgan of course has a different conclusion: For the majority of companies, there's a clear trend: Pay higher dividends, and the market will reward you with a higher P/E ratio.
One of the examples Morgan gives is Consolidated Edison (NYSE: ED ) . Morgan points out that ConEd has a higher P/E than Microsoft (Nasdaq: MSFT ) despite obviously having a lower growth rate. He concludes it's because ConEd has a higher payout ratio.
The problem with this is that earnings growth is just one side of the P/E equation. The other is how risky those earnings are. And all else equal, a regulated utility like ConEd is a lot less risky than a software company like Microsoft. Therefore a higher P/E isn't so weird. I also think that Microsoft is a bad example because it's simply undervalued right now (it's been a longtime CAPScall of mine).
The same is even more true of Morgan's comparison of AT&T (NYSE: T ) to Apple (Nasdaq: AAPL ) . Yes, it's true that Apple's growth rate is far greater, but AT&T's business model is also a heck of a lot less risky (which is part of why, you guessed it, it's also a CAPScall). Therefore a higher P/E than Apple is not as crazy as it initially appears, and is consistent with my hypothesis.
Why I fight
Some of you may be wondering why I fight so hard, in so many columns, against dividend evangelism when I clearly like a lot of dividend-paying stocks. Heck, Vanguard Dividend Appreciation is my largest holding and a CAPScall.
There are two reasons why I fight.
The first is that this belief that dividends are manna from heaven is dangerous for investors. I can't count the number of times I've seen investors hold on to bad companies simply because they think their dividends confer magical outperformance powers. Think of the shareholders who held on to old GM or the financials late last decade.
The second is that dividend policy is ultimately a distraction for investors and management. Fools should concentrate on buying the best businesses at the best price, dividend or no dividend. And managements should focus on making their companies worthy of a Fool, dividend or no dividend. Dividend policy does not make a bad business good, nor a good business bad.