"The Value of a Company
by: EliasFardo 12/17/97 05:37 pm
Msg: 31 of 166935
I think that Buffett would discount futute cash flows. However, Charlie Munger says that Buffett talks about this but has never seen him do it. I think Buffett has some short-cut method in his head so actually both of these methods may be close. He may make a few marks on the back of a napkin, but don't look for any fancy models. I think that determining what a reasonable PE multiple is one of the really hard parts of investing. What is reasonable is different based on the company and surrounding economic factores. My guess is that he gets a feel for the expected rate of earnings increases, the probability that this rate is sustainable, the cash needs of the company, the extent to which the company has a moat, and comes up with an acceptable earnings multiple. ( Wait a minute, thats what I do. ) I guess I don't have any idea of what he does after all. I have the new book but have not yet read it. The are some other excellent books on Buffett. Buffett, The Making of an American Capitalist by Roger Lowenstein and Of Permanent Value by Andrew Kilpatrick. If you are into Berkshire, these are worth checking out. They have a lot of meat in them. Rober Lowenstein also writes a weekly column in the Wall Street Journal which is excellent. One thing is for sure, I would have a heck of lot more money if the only investment I ever made was stock in Berkshire Hathaway."
"I wanted to find a way to value stocks like bonds."
OK, let's work the following hypothetical bond together.
Assuming you actually wanted to make a "treasury bond like" return of 5% on your investment, what would you pay for an inflation-indexed, investment grade ("AAA"), corporate bond with the following properties?
* It makes a payment of $2,255 one year after your date of purchase.
* It continues to make a payment each and every year (on the anniversary of your purchase) that is an inflationary 2.5% bigger than the previous year's payment.
* After making 300 annual payments the payments cease and the bond expires (no special liquidating payment above and beyond the last normal payment at expiration).
Your schedule of payments and their present value discounted back to the START of year #1 (which is the same as the END of year #0) would look something like this:
Year # ........ payment ... present_value(payment)
....... 1 .......... $2,255 .......... $2,148
....... 2 .......... $2,311 .......... $2,096
....... 3 .......... $2,369 .......... $2,047
... 100 ......... $25,990 ............ $198
... 200 ....... $307,041 .............. $18
... 300 .... $3,627,290 ................ $2
...............................total ... $90,135
* For any year "n", the payment at the end of that year is = $2,200*1.025^n
* For any year "n", the present value of that year's payment is = payment/1.05^n = $2,200*(1.025/1.05)^n
* The present value of the stream of 300 annual payments is simply the total of the present values of the individual payments.
Of course, we could just use the Gordon Stable Growth Discounted Dividend Equation (where "n" = infinity instead of 300):
V = D.o * ( 1 + g ) / ( r - g )
V = $2,200 * ( 1 + 0.025) / ( 0.050 - 0.025 ) = $90,200
Remember: this is the value assuming a 5% return on investment pleases us. If we want to make a higher return then the calculated value is obviously going to be lower. For example, if we want to make a 6% return the present value of that infinite stream of payments is worth $64,429.
DISCLAIMER: Any similarities between this hypothetical bond and any actual security are purely coincidental.
I am sorry I let others side track me. I am really not interested in CAGR it is used to measure short term performance. I wanted to find a way to value stocks like bonds. You can if one normalizes earning and growth. The future amount of a single sum is the orginal sum plus the compound interest earned up to a specific date.
You substitute normalized earnings and growth for interest rate and payment. I never realized a compound interest table is a mental model. I thought the interest tables were quite well respected.
"Actually it takes the share price to 83,268 if you use 15%."
Yeah, I was just doing a quick mental approximation using those last six annual returns.
"I don't feel comfortable using that number since Gen RE composes the major part of float. I prefer to average the low end expectations with the high end expectation."
"CAGR doesn't represent reality."
" It is a pro forma number."
" CAGR is the best formula for evaluating how different investments have performed over time. Investors can compare the CAGR in order to evaluate how well one stock performed against other stocks in a peer group or against a market index. The CAGR can also be used to compare the historical returns of stocks to bonds or a savings account.
When using the CAGR, it is important to remember two things: the CAGR does not reflect investment risk, and you must use the same time periods."
""Normalized earnings are what we think a company would be earning today in a boring year in the middle of a boring economic cycle with nothing extraordinary going on anywhere."
"What's a conservative view of what the company will do in a very boring environment." Maybe it's core earnings, and I know S&P is talking about that. But we are trying to make it even more core by finding the midpoint of an economic cycle, an interest rate cycle, a technology spending cycle and saying what would the company earn then. It's a stable, slow-moving measure. In that way it's more like book value in that it doesn't have anywhere near the volatility of reported earnings. It's a good number to know for a long-term value investor."
Sounds like the 12% number would be correct.
I stil don't know about the 15% rate adding up and dividing. It never sounded right to me I was trying to find a justification for getting to 80,000-85,000. However it doesn't sound like you can only use 5 years. I am still learning about CAGR.
You can make any model you want it is your money.
I was using the rule of 72 for a rough approximation.
I was trying to see how Elias goes about his valuation model. I found out he uses normalized earnings. I went back and read the annual reports
and felt WEB's implied growth rate was 9.5%-15%. The 15% is the very giddy range. So a approximation would be 12% for steady growth would seem to apply. That makes sense because the averages will grow around 6.5% and WEB usually doubles his returns compared to the overall averages.
So I favor Elias's model for valuation purposes and feel anything higher implies that nothing will go wrong and everything will go perfect forever. I am sure Elias double checks his numbers. This is just my rough anlysis of seeing whose work I like better.
Normalized earnings implies to me that you take an average of the years and that becomes your norm to use for compound interest tables.
Actually it takes the share price to 83,268 if you use 15%.
I don't feel comfortable using that number since Gen RE composes the major part of float. I prefer to average the low end expectations with the high end expectation.
I see how Elias formed his number and prefer to use his model for valuation porposes. It is a personal decision. If I was a seller I would prefer to use the high end model ;-)
Elias is an accountatnt and knows the company very well. His estimate is close to 70,000. Elias normalizes his earnings. I trust him. He has no other agenda. IV is not an exact science. If you can convince a buyer it's worth 80,000 good for you. I just feel it is closer to 70,000 than 80,000.
BRK is a very hard company to value. Everyone needs to make their own valuation decision for themselves. Gen Re accounts for 2/3rds of float. I just feel more comfortable with conservative estimates as a buyer. Now if you are a seller 80,000 sounds much nicer :-). Everytime it comes close to 80,000 WB issues some sign and makes the stock go down. Isn't that a clue?