Focusing on expected returns instead of intrinsic value…

But they're [Nestle] trading at a 3.3-3.4% dividend. Their growth, because they are in high-growth areas -- if you read their financials, it looks like the growth is slower because historically it's been in Swiss francs and the Swiss franc has been appreciated -- strongly relative to the currencies in which they sell, but their growth is probably, especially in earnings because they have operating leverage, maybe 6%, which is a little above the 4.5% nominal world GDP growth, or even 4% nominal world GDP growth. 

You had a 6% earnings growth, a 3.3% dividend, and probably some buybacks on top of it because they really don't need all that much capital, so you're looking at a 10% return on a stock. That is extraordinarily safe. 

You look at what happened to them after 2003, where their input prices when through the roof, and there was no general inflation, and their margins go up steadily. You look at them in the crisis and they're barely affected by it in '08-'09. 

It's sort of recession-proof and inflation-proof, and it's a 10% return. Maybe as low as 8, but in a world -- even where 30-year Treasuries, which have all the inflation risk, are 3% -- that is not an expensive stock, and there a fair number of opportunities.

…

When growth has value, what you're trying to do is take a cash flow number and multiply it by the appropriate multiple for a growing stream. Costs of capital these days are probably 8%. Growth rates for anybody with a franchise... and growth has no value if you don't have barriers to entry, because your return on capital is driven by the cost of capital, so you invest $100 million in growth, you earn 8%, you have to pay 8% to the people who provided it. 

In the stocks where growth matters, you've got, say, an 8% cost of capital and, because you've got operating leverage, probably a 5% growth in earnings. That's 1/3%, which is 33 times, which is a crazy multiple. 

If you're off by just 1%, you could have 100 times. If, for example, the cost of capital was 7 and the growth was not 5, it was 6, which is 1/1%, or if the cost of capital was 9 and the growth rate is 5, you're 25 times, and even lower. 

You make small mistakes in those numbers, you get massive mistakes in valuation, so when you've got a growing stream, it's almost impossible to put a value on it. If the growth doesn't matter, you can get an earnings-powered value. It ought to be supported by assets because there are no barriers to entry, and you'll get a very good valuation. 

You can get a valuation metric in terms of price/earnings ratios, price to tangible book, ultimately price to replacement cost, and price to sustainable earnings. 

In the non-growth stocks, which are the non-franchised stocks, it's always been the case that traditional valuation metrics apply, and they continue to apply. In the growth stocks, on the other hand, I think we now have learned, to our cost -- and by the way, not just in this crisis but in the sell-off for the tech bust in '99-2000 -- that you cannot put a sensible value on things. 

If you can't put a sensible value on things, the question is how do you decide if they're worth buying? The answer is, you can put a sensible return on things. 

In those cases, notice what I did. I said, "OK, here's the cash return." It's the dividends plus the buybacks, probably -- expected level of buybacks, which is like 4.4% -- the growth is what it is, and the growth in earnings may be slightly higher than the growth in sales, but the growth in sales is forecastable. 

It's going to be either slightly above or slightly below the level of GDP growth, depending on how successfully they allocate capital to grab market share or whether they're in a category that's growing faster, like pet food, than GDP or much slower, like newspapers or automobiles, than GDP. 

You can assess those numbers and notice that now it's not 1/R-G. It's just linearly related to the return. What I did for you is I basically calculated a total return for Nestle, and that has always been the sensible way to look at growth stocks. 

You notice you can compare that directly to returns on Treasuries, to returns on junk debt, to returns on regular debt, because they're all in returns space so you've got a much better idea of what you're buying. 

That's the good news about what we've learned. I think it's not just been from the crisis. I think, as I say, it goes back to the tech bust where people put unrealistic and crazy valuations on things. 

… 

It's what, really, people like Warren Buffett, who have pioneered the process of investing in these franchise businesses -- and it's only the franchise businesses where growth has value, because that's where you earn above the cost of capital. 

I mean, if a market grows with no barriers to entry, you just get a lot of entry. It drives prices down and it eliminates the profitability, so you have to have the barriers to entry. 

What they always said is, you can make a sensible buy decision. It is incredibly difficult to make a sensible sell decision. That's because you have to pick a price at which to sell. I get a price from the market today and I can calculate, just as I did for Nestle and other things, a return at that price. 

But when I decide to sell Nestle, I've got to pick a price to sell it at. That, of course, gets into all the problems of at what point it's overvalued or not. There, I think what you have to do is not hold on too long. That's the temptation. 

That's what kills Buffett when he won't sell Coca-Cola at $80 a share. It's what kills him when he holds onto the Washington Post for too long. 

It's a very difficult decision to make, but I think the evidence is that you've got to be pretty conservative about it so that if the dividend return for Nestle, for example, plus the buyback return, goes down to 3.5, even though the overall return may be reasonable relative to returns on fixed income, you're still going to have to sell at that point. 

But I think what we've learned in, as I say, as much the tech bust as the current crisis, is you cannot effectively put numerical values on stocks, or have price-related multiples on stocks, that are growth stocks. 

If you'll let me, I'll do just one more example that shows you. 

… 

People got really excited about newspapers when they got down to like 8 times earnings. Eight times earnings is an earnings return of about 12.5%. They were distributing 80% of that, so it was a 10% cash return, either buybacks or dividends. 

In spite of what investments they were doing, those businesses were shrinking at at least 5% a year in terms of their earnings power, and they had no good alternative investments. They tried and failed, and the basic businesses were shrinking. You had all the loss in operating leverage, because you had this big fixed cost infrastructure to fill and sell newspapers and a shrinking revenue base against it. 

… But let's say it was 7% a year, which means it takes 14 years for it to fall in half. You start with 10 and subtract 7, that's a 3% return, which is a substandard return in a very high-risk investment. 

When people are excited by the multiple, which has never been as low as 8, if they do the return calculation, they'll understand that the multiple is almost completely uninformative, that you have to look at the implied returns. That's what I think people have learned to in these franchise stocks.

His idea of a fair price, by the way, is a price that gets him a return going forward on that investment without any improvement in the multiple of somewhere between 13 and 15%. By normal standards, when the average market return is 7-8%, that’s a really good price, he’s looking for a very good price. They call it a fair price because the multiple may be fairly rich, it may be 13, 14 times earnings. But the value of the growth may bring his return up to 13-15%. So when he says a fair price he’s talking in terms of normal value metrics, and there the reason that you prefer that to a poor company at a really good price is that because the good company can grow through reinvestment and things like same-store sales growth, your return will come in the form of capital [gain], not distributive income, and that, after tax, is much more valuable.

At Giverny Capital, we do not evaluate the quality of an investment by the short-term fluctuations in its stock price. Our wiring is such that we consider ourselves owners of the companies in which we invest. Consequently, we study the growth in earnings of our companies and their long-term outlook. 

Since 1996, we have presented a chart depicting the growth in the intrinsic value of our companies using a measurement developed by Warren Buffett: “owner’s earnings”. We arrive at our estimate of the increase in intrinsic value of our companies by adding the growth in earnings per share (EPS) and the average dividend yield of the portfolio. This analysis is not exactly precise but, we believe, approximately correct. In the non-scientific world of the stock market, and as Keynes would have said: “It is better to be roughly right than precisely wrong.”