Corner of Berkshire & Fairfax Message Board
General Category => General Discussion => Topic started by: nickenumbers on November 25, 2018, 05:22:17 PM

All,
I wanted to get a sense of what you guys think about what value to use as your discount rate in a DCF valuation?
I have been using between 910%.
I was listening to a BRK annual meeting from 2003, and WEB and CM get a similar question and the answer that they gave was that they use the rate of their next best alternative for the money.
So, that got me thinking, and I wanted to get your input.
If we say that a riskfree Treasury is around 2.5 or 2.75%. I understand the concept of the Equity Risk Premium. But, what is my next best alternative of the money, and what rate of return is it earning? Because of PEs, my Rate of Return might be 7.5% over the next 12 months and it might be less.
What do you all think and what do you guys use as your discount rate for a DCF, and why?

I haven't heard
WEB and CM ... answer that they use the rate of their next best alternative for the money
It makes sense though and I think I've used that approach in the past.
And assuming your next best alternative is BRK, then 910% might be a good return expectation and good discount rate.
OTOH, this is a bit circular definition. Assume you say the discount rate is X% because you expect stock XX to return X% long term. But why do you expect XX to return X%? Is that because you did DCF on XX? What discount rate was used on XX then? If it was X%, then where did it come from? If it was not X% then why use X% as discount rate now and why it was not used to evaluate XX?

Aftertax cost of financing that you're going to use for the project.
If you're looking at an investment, look at their bond rating and the market yield on their 5 year bond, & calculate Beta. Thereafter plug the variables into the CAPM model to obtain the discount rate. Just keep in mind that you're calculating off of historic rates, and that this works best with stable companies  not small, or unstable ones. It's also not particularly good, but it will at least give you a ballpark number.
SD

IIRC, Buffett said that he uses the 30year US treasury bond as the discount rate. He said he mainly uses the DCF framework to compare the investments he is contemplating. Thus he would only consider a new investment if its projected return exceeds that of his next best alternative which he usually has already a position in and understands very well.

I use 811 depending on the business and durability of the cash flows, but have learned over time not to focus too much on the discount rate. It's getting the cash flows right that matters. If using 9% instead of 10% makes a difference for me investing, then the value is too close. I'll move on.

I had a conversation with someone about this recently. By using the 30 years, the results can get weird depending on if you go out 5, 10, or 20 years. As value that goes out further gets discounted less in a 3% scenario. So, you can kind of game the DCF by going out 20 years instead of 5 years.
An alternative would be to set the NPV as the current price and look at what the discount rate would be. If you're getting a >15% discount rate by holding the NPV constant, you're probably onto something interesting. If you get a discount rate of >20% if you equate the NPV as the current price, you're likely looking at a very good investment. These things are "Garbage in, Garbage out" So if your assumptions are massively off, then nothing really matters.

As always, the collective input is outstanding. Thanks for your opinions.

I had a conversation with someone about this recently. By using the 30 years, the results can get weird depending on if you go out 5, 10, or 20 years. As value that goes out further gets discounted less in a 3% scenario. So, you can kind of game the DCF by going out 20 years instead of 5 years.
An alternative would be to set the NPV as the current price and look at what the discount rate would be. If you're getting a >15% discount rate by holding the NPV constant, you're probably onto something interesting. If you get a discount rate of >20% if you equate the NPV as the current price, you're likely looking at a very good investment. These things are "Garbage in, Garbage out" So if your assumptions are massively off, then nothing really matters.
The value isn;t important, it's the ability to use a static framework to compare options. You can stick your estimated CFs for BRK and AAPL, use the tbill rate, and see what returns a higher value.
It's like a bathroom scale. We all know the scale isn't accurate, but it can still be used over time to measure relative changes.

IIRC, Buffett said that he uses the 30year US treasury bond as the discount rate. He said he mainly uses the DCF framework to compare the investments he is contemplating. Thus he would only consider a new investment if its projected return exceeds that of his next best alternative which he usually has already a position in and understands very well.
I believe this is how Buffett approaches DCF. Going by the classic Buffett buys, the understanding is that he can reasonably assume that there will be cash flows 30 years hence. The intellectual honesty that is implied in making such an assumption is what separates him from us. The math is the easy part, I’d even wager that Buffett has the DCF tables memorized to nearest round number. At least the range that’s relevant. Munger refered to his dogeared pocket book.
Buffett’s famous (modified) Aesop fable “How sure are you that there are x birds in the bush and if they will emerge” captures the thinking.

IMHO there is a little misunderstanding w/r/t Buffett and Munger's 9% or treasury rate discount rate. They are using it as a way to compare all opportunities to each other.
Most investors use it on a case by case basis to determine if a specific investment is undervalued. Then the discount rate should generally be your desired return (if you would be willing to buy it at the DCF price); you could be a prophet when it comes to forecasting cash flows but you'll still only earn your discount rate if everything goes perfectly.