Warren Buffett is not one to give copious interviews, so when he does, it's a treat worth reviewing.

Although the chairman of Berkshire Hathaway (NYSE:BRK.A) writes extremely lucid and timely thoughts about investing every year in his company's annual report, the few interviews that arrive between those reports always offer new insights into a man who is refreshingly long term, delightfully honest, and successfully "unambitious" in his business dealings. Sometimes, he just sits back in his Nebraska office and waits.

That was his message in today's Barron's online interview. Buffett, his partner Charlie Munger, and the company they run are sitting on $24 billion in shareholder cash with few attractive places to invest it. As Buffett has written in his annual reports over the past several years, stock prices leave much to be desired, and finding good value is like finding a needle in a haystack.

During the last five years, Berkshire has completed relatively few acquisitions, and paid an average of seven times pre-tax earnings for each, about half the market's current 14 times multiple on pre-tax earnings (while the S&P 500 is at 20 times projected 2003 after-tax income). Today, even with the market far below its 2000 peak, Buffett said, "We're not finding anything. We have more cash than ideas. The question is whether that will prevail for an unduly long time."

This said, Berkshire has made two acquisitions in 2003, totalling $3.2 billion, buying a manufactured-home maker and lender, and a grocery wholesaler. Buffett continues to seek businesses that he can project results for well into the future, shunning the unpredictable nature of technology firms. Quoting from the Barron's article:

Buffett has never been comfortable with technology stocks, and he's somewhat puzzled by the current valuation gap between major tech and drug issues. Intel and Cisco command double the valuation of pharmaceutical leaders such as Pfizer and Johnson & Johnson. "Drugs are a better business in the aggregate than technology," Buffett argues, citing higher returns on capital and greater product longevity, owing to patent protection. "If you look at the top 10 drug companies ranked by sales, the No. 2 or No. 3 company from the bottom is still earning a good return. It's hard to find a drug company that has failed," Buffett says. In tech, by contrast, "the outstanding ones are outstanding," but there are fewer of them, he says. Tech companies, he adds, are more sensitive to the economy.

An editorial comment I'll add is that investors are, of course, hoping for large earnings rebounds -- 40% or greater -- at tech giants, while drug companies are maintaining earnings growth rates in the mid-teens. Still, it's interesting that investors are putting a higher premium on hoped-for results than they are on the almost-assured results at drug makers.

Buffett admits in the interview that not selling Coca-Cola (NYSE:KO) and Gillette (NYSE:G) in the late 1990s, when both were much higher than today, was a mistake. However, being on the board of both companies made it nearly impossible to sell and Berkshire has still done well owning both since the 1980s.

In his typical candid and humble fashion, Buffett also admitted that not buying Wal-Mart (NYSE:WMT) years ago because it looked overvalued was a large mistake that cost them $8 billion (so far). He professed to love Wal-Mart's business, arguing -- as the Fool's Lou Ann Lofton has -- that Wal-Mart is killing supermarkets and implying he wouldn't touch those firms despite low multiples. He did not provide thoughts on Wal-Mart's current valuation.

One sector that Buffett did imply looked attractive was banking, where multiples remain below market averages and returns on equity maintain a 20% level.

Despite the company's cash balance, Berkshire will not likely pay a dividend anytime soon. Not unless Buffett & Co. see little hope in reinvesting the cash for strong shareholder returns. A few years ago, Buffett projected that the stock market was likely to return about 6% annualized in the next 17 years, well below its historical average.