Buffett is noted for saying,
"You can't compensate for risk by using a high discount rate."
Does that mean we shouldn't use a high rate? Does it mean that using a high rate is risky?
How the discount rate affects calculations
The higher the discount rate, the lower the valuation and vice versa. But valuing businesses is not just about the discount rate—you need a margin of safety. The two go hand in hand.
Discount rate and margin of safety
If you use a higher discount rate, you'll end up with a lower valuation. When that happens, you can buy great companies with a lower margin of safety than if you use a low discount rate.
Let's look at the valuations. For this example, we'll use the Coca-Cola valuation in 1996:
At 15%: We estimated Coca-Cola's value at $30.45 per share. Because it is an industry leader, I said we could comfortably buy it at a 25% discount, or $22.84 a share. The total valuation for the company came in at $75.8 billion. At 9%: If we were to discount the future cash at 9%, Coca-Cola's value would have come in at $125 billion, or $50.25 per share. But if we're using 9%, we should have a 50% margin of safety (I'll explain in a second) giving us a target buy price of $25.13 per share.
Why change the margin of safety?
Hey, I'm not making this stuff up—I analyzed Buffett's past purchases and found that:
At a 15% discount rate, he was buying at a 25% margin of safety for big companies. Using a 9% discount rate (8.85% for Coca-Cola in 1988), he had a 50% margin of safety.
Why did I use 8.85%? That was the rate that the ten-year, risk-free US bond was offering in 1988. Why did I use 15%? That is the minimum return I expect when I buy a stock.
The results are close enough in this inexact science
Either way, the purchase prices, with their respective margins of safety, are so close together that the result ends up roughly the same.
So, what was Buffett saying?
I believe that Buffett was saying that a risky company is a risky investment, no matter how you justify the purchase. Using a 15% (or higher) discount rate means you end up with a lower overall valuation and purchase price. But in the end, a bad business is a bad investment no matter how cheaply you bought it.
To demystify and paraphrase a bit, Joe Ponzio says (if you really care, that is):
"Bad companies are bad investments, no matter what price you pay for them. The more time you spend rationalizing your purchase, the less money you should put into it."
So, what rate should I use?
Whatever rate you want, so long as you factor in an appropriate margin of safety to compensate for your (our) inability to predict the future. The lower the discount rate, the wider the margin of safety. Discount rates aren't about risk...they're about valuing businesses.
Personally, I use 15% and require a 25% margin of safety on large, stable companies and a 50% margin of safety on less-than-sure companies. For additional details and justification, take a look at Michael's question and my response on the Johnson & Johnson valuation.