Investment legend Bill Miller recently banged on about how large-cap stocks are the investment opportunity of a lifetime. He offered some more bullish musings on CNBC yesterday:

If the [dividend] payout ratios were dramatically higher than they are today, then people would have a lot more money in the stock market. If you look at IBM (NYSE: IBM), IBM is the only one of the megacap stocks that were popular in the late 1990s to go to an all-time high. The reason for that is, IBM has an absolutely unequal record in capital allocation. They basically return cash to you in the form of dividends and share buybacks; they do it consistently over time.

Hewlett-Packard (NYSE: HPQ) hasn't raised their dividend in five or six years. If Hewlett-Packard had a 70% payout ratio, which they should have -- or Microsoft (Nasdaq: MSFT), which they should have, or Intel (Nasdaq: INTC), which they should have -- they would all have over 7% dividend yields today ...

Obviously, they wouldn't trade at 7% dividend yields. The telecom sector is the highest-valued sector in the overall market, and it grows at 3%-4%. But it has a high dividend payout ratio, which is why it trades where it does. If those [tech] companies actually did what they ought to do, which is pay out 70%, the stocks would be up 50%. Easily.

Heck, why stop at those few tech companies? Over the past 12 months, the S&P 500 average paid out 33% of earnings as dividends, vs. a long-term average 51%. The SPDR S&P 500 ETF (NYSE: SPY) currently yields 2%. If the index simply reverted to its long-term payout average, the ETF's yield would go north of 3%. If the S&P earns $83 in 2010, as it's on track to, and paid out 72% of earnings as dividends, as was the average from 1930-1950, the index would yield more than 5%. And as Miller points out with telecoms, the market is more than willing to bid up valuations once yield is brought into the equation.

Which is, I think, mostly silly. Investors should be willing to pay a similar multiple for cash flow as they are dividends, provided management doesn't have a fondness for squandering cash (and some do, to be sure). Apple (Nasdaq: AAPL), for example, pays no dividends, yet the immense cash it generates every quarter increases the company's market cap because of its larger cash balances. It's hard to see how higher dividends would materially benefit Apple shareholders. If Apple paid a $10 per-share dividend, mark my words: Its stock would fall $10 on the ex-dividend date. This is precisely what happened to Microsoft when it paid a large one-time special dividend in 2004. The company paid a $3 dividend, and its shares fell $3 on the ex-dividend date to reflect its lower cash balances. Shareholders actually faced a net-negative return once taxes were factored in. The return characteristics of dividends aren't as straightforward as they seem.

But dividends are what investors crave these days. Bonds yield almost nothing. Cash earns less. Yield is in great demand wherever it can be found. It's hard to imagine the S&P dividend yield surging without stock prices surging in response. And the yield could surge without the economy changing a lick if companies just brought payout ratios back to historic averages.

Some will say higher payout ratios makes stocks less attractive because it leaves companies with less cash to invest and grow. This would be a fair argument in normal times. Today it's mostly moot. In aggregate, companies are holding record levels of idle cash and continue to add to those stockpiles. As Miller rightly went on to note, "they don't need the cash; they're hoarding cash." There's plenty of fuel to invest with without tapping into cash flow.

One more comment, although not quite as specific, on current valuations leading to outsized returns. I had a chance to speak with renowned Wharton professor Jeremy Siegel last week while researching another article. Here are his thoughts on the odds we'll undergo another decade of abysmal stock returns:

Extremely small. The super-bull market of the 1980s and 1990s ended in the technology/Internet bubble that pushed stocks too high. But the bear market of 2008-09 sent them below their long-term values. I now estimate that stocks are about 30% below their long term trends and fair market value.

That was the end of our discussion on the topic. No specifics, although the 30%-off level doesn't seem dramatically different to how Bill Miller's argument would apply to the overall market. A 30% increase in the S&P would put the index at around 1,530, which is about 16 times estimated 2011 earnings. Not cheap, but hardly unreasonable. This chart I put together last week that gives some historical context to the issue:


I'll give Bill Miller the last word:

The last ten years have conditioned people to think short term and tactically as opposed to long term. Everybody wanted to think long term in 1999 after 17 years of a great market and ten years of returns that were 20% per year. And that was the wrong time to think long term; things were really expensive. Now things are really cheap, and it's time to go back to thinking long term.

Check back every Tuesday and Friday for Morgan Housel's columns on finance and economics.