Wednesday, May 15, 2013

What's wrong with DCF?

DCF or Discounted Cash Flow if you do your research, has been said to be one of the most preferred valuation model. It is preferred due to its assumed accuracy in terms of projecting what the valuation of a business based on what is projected down the road.

I like cashflow, it is one of the most important measurement on how well a company or business is doing. But DCF is all bullshit. Why?

I have done tonnes of them (DCF) and everytime I feel that I am bullshitting. Most of the time, the DCF that I do is for the company I work for or to make someone happy that the company is worth so much. From a DCF, one can just make a major change to a valuation based on a slight change in the assumption.

Before we go further, what is a DCF? DCF is a method of projecting a company's worth using mainly the future cashflow of a business and a discount value. Both are equally important. But both are equally difficult to project.

A future cashflow - imagine for a business, can some company project its cashflow say 5 years down the road? Or perhaps 10 years down the road? If that person being the CFO, CEO can project that close to accuracy, then he / she is almost superhuman. Future numbers are amazingly hard to predict. It is like someone (a super analyst) projecting what is the Bursa KL Composite say in 31 December 2018. People change, situation changes, management move around. And are you telling me management is not important in a business? Well DCF is saying that, isn't it. DCF can be assuming a scenario 20 years down the road and we are supposed to be taking the number say 2034 as correct for the valuation to be accurate.

In a DCF, the projection is not only for one year, but for a period. If it is for 10 years, then it is for 2014 to 2023 or even in perpetuity. And it has to be almost accurately provided. Can a business be consistent? Some yes, mostly no. In fact, very few yes. Then how are we to provide a DCF accurately?

Then comes the discount value! A 1% difference, in a lot of times cause significant changes to the valuation. So which number to use 10%? 11%? 12%?

Why the hell then someone still wants to use DCF. Most probably, the current valuation is not good enough. And most of the time that current value is not good enough, some guy has the indigenious thought of why don't they come out with a projection over a long stretch of years and project the value backward, considering time value of money.

Well, if someone wants to bullshit another, you can contact me to do the future cashflow for you and discount them to today's value, cause I am pretty good at that!

For investment though, go for the tried and tested, based on historical track record and what you feel the business is heading in the future.

There are only some businesses which a DCF can probably be used - mostly concessions. Even then other factors are so important.


K C said...

DCF analysis is no doubt one of the most controversial valuation methods in investing. However, it is the "accurate" method of valuing a business, at least theoretically, albeit with some "wild" guesses especially in term of the future cash flows. However, isn't that useful to have a feel of the value of a business before one pays a price for it?

Out of the two important assumptions used in DCFA, the discount rate, though can affect the outcome greatly, may not be that controversial as each investor could have his own rate which he is willing to use. It is the future cash flow projection which is highly unpredictable. Hence the usefulness of the concept of margin of safety.

May be it is better to use the reversed DCFA to find out what growth expectation the market is paying for the company if the growth expectation is realistic or not. Often one can see market paying exorbitant prices as if growth is 50% or 100% for the next 10 years for some companies, and also sometimes the market price has incorporated some huge negative growth expectation when the business does not appear to be so.

felicity said...

You are right but usually the prospects would be more coming out with aggressive projection more often than not.


panaceaasia said...

Thank you for your intelligent views on equities and investing.

Multi said...

I agree with Felicity that more often than not, DCF was used to justify the valuation by playing around with those variables, because Professionals and authorities need the quantitative or scientific method to support their decision...we as investors have to exercise our own judgment.

Betronist said...

I respectfully disagree with the writer, though I was in the same shoe before.

I think DCF is still by far the most useful valuation method. Just that it is much more subjective than the others. Well, even Buffett uses it.

DCF is the ONLY valuation method that value companies in an absolute way, not a relative way like P/E, P/BV, EV/EBITDA, etc. Investors should value stocks based on what they want, not comparing to what others have gotten.

However, there are many shortfalls of using DCF:
1) Your holding period has to be aligned with your DCF period. What is the purpose of discounting 10 years cash flows when you are just holding the company for 2 years?

2) DCF valuation is meant for the controlling shareholders. The generated free cash flows always stay with the company and the minorities like us have no says on it. And that's why, it is more appropriate to use Dividend Discounting Model for minorities (which itself has many shortfalls too). Anyway, one should factor in a minority discount of 10% to 20% after deriving your DCF valuation.

3) Like what the writer said, any slight tweak on the assumptions will result a huge change in valuation, which to me actually is non-issue. I normally apply conservative assumptions, and anyhow I don't forecast up to 10 years, 5 years is my maximum, I used 3 years most of the time. As for discount rate, I use my own universal 10% required return rate, which is slightly above the historical long term stocks market return of 8% (if my memory serves me correct).

Anyway, I think the best way to value using DCF is to create a range based on 2 scenarios i.e. optimistic growth (say 10% growth) and pessimistic growth (say 5% growth), but remember to use same discount rate on both scenarios. Ultimately, take the average valuation of both scenarios, that should be your calculated value.

I can see why people hate it (I did too) because many times the generated value is far below the market price. Well, that's how Buffett does it. If the price is not right, just skip it and move on.

Hope that I'm not too cheong hei. Hehehe

investbullbear said...

DCF is still a useful method.

1. The company should be generating cash flows that are predictably consistent.

2. Then use very conservative growth rate in your projections.

3. Having done so, apply a discount rate that is appropriate in your opinion.

4. You would have then arrived at a valuation that is conservative enough to guide your investing.

5. This value is not exact but a good guide.

Just share with you all here. Your final decision to buy or to hold or to sell a stock, though guided by your valuation method(s), is still a SUBJECTIVE decision.

Faced with 2 companies you like, your decision favouring one to the other is yet a SUBJECTIVE ONE. :-)

REMM2012 said...

I do find some value in using DCF though not looking at the absolute number but as a ball park guide.

I am using DCF in 2 ways (using the latest FCF and project it at 5% pa, and line of the best fit for the past FCF).

I am also using discounted earnings (similar to DCF but using earnings instead of FCF). Also, 2 methods - latest earnings and line of best fit for the past earnings.

Results using the 4 calculations vary from one end with all 4 methods giving non intrinsic valuation to the other end with all 4 methods giving intrinsic valuation (leaving out safety margin for the time being).

E.g. analysis on Petronas Gas gives me non intrinsic valuation for all 4 methods.

I would be more willing to invest for the ones which have all 4 or 3 intrinsic valuations (after considering other key parameters).

My 2 cents.

felicity said...

Thanks, knew that my post is going to be controversial. I guess for a lot of companies with insider information, DCF is the most preferred valuation method.