I think we are all nudged towards DCFs when we start investing BC it seems like a high level of intellectual rigour. I remember first reading Damodaran and he discussed relative valuation (multiples) like a shameful thing lol.
After a while you realise you're just plugging whatever assumptions suit you in there.
Your framework is excellent in that it strips the bullshit out, but narrows in past the "finger in the wind 8x feels about right" stuff.
It looks like a DCF with static growth/ROIC for a "holding period", and then the same concept of exit multiple. Your "minimum expected return over the holding period" sounds a lot like a cost of equity. Am I wrong?
Re the other points:
- Terminal multiple - don't people usually use a sustainable long-term growth rate and ROIC to get to terminal value? You can get this to spit out a multiple if you want, they're the same. But it's not inconsistent
- Assessing risk using a discount rate is exactly what you are doing with your hurdle rate. Either that or you're not considering risk at all?
- I think you're also wrong that a DCF penalises companies reinvesting FCF to grow ("does not really count cash flow"). Isn't the whole point that if ROIC is above cost of capital, reinvesting all the cash flow increases the value?
So yeah, I think the only point I agree with is that DCFs "pay no attention to what happens to the FCF that is generated". I think that's by design
I'm not a DCF purist or anything btw, very far from it
If you stop counting after say 20-30 years and determine terminal value based on discounted net profit, then are you really taking ROIC into account? I guess sort of, partially? So my second point is half wrong I suppose. I have seen people determine terminal value by using EV/EBITDA.
I disagree that my method is more convoluted. I think a DCF is convoluted, especially when potentially more than half your end value is not even determined by actually counting cash flows.
Building one is time consuming and it is easy to make some small mistake/typo and have it spit out the wrong value without noticing. And if I actually keep counting for 70 years, in order to see what goes into my model you have to scroll through rows and rows of kind of laughable assumptions (because the company will likely not even exist 70 years from now).
My method splits up a valuation into high growth and a mature phase and takes explicitly into account:
-ROIC (all the way)
-Payout % when the company goes into its low growth mature phase. This is a pain to do with a DCF (especially when terminal value is a big % of your valuation).
-Period that you feel confident predicting growth. Since most value investors looking to significantly outperform are not interested in holding a low growth company at a fair/over valuation.
-Risk by explicitly calling it a minimum required return (seems clearer than a hurdle rate).
And all of this can be quickly displayed on 1 page, and a calculation can be done quickly in less than a minute to get an overview of what is reasonable to pay for a stock.
And opportunity cost (which is what cost of capital sort of tries to be I guess?) is determined by how far the current valuation differs from the maximum earnings multiple you should pay based on the formula. And is ofcourse heavily determined by growth period and expected holding period when betting on earnings multiple reverting to the mean.
DCF’s are a helpful tool to assess sensitivities to certain factors, help see bear/ bull valuation scenarios.
Just putting a multiple on something w/o understanding what growth, margin, capital investment assumptions are reflected by that multiple makes less sense than using a DCF.
You have no idea what you're talking about, this article is so bad and wrong it actually made me angry
"The fact that a DCF does not really count cash flows if the company is growing" b/c there is no cash-flow... make your forecast period longer until you get to steady state cash flows.
"ROIC is not included in a DCF calculation" why not? Base your CF forecast of ROIC assumptions if you like
"Say you have company A that generates a 10% ROIC and company B that generates 100%. Company B should clearly be valued higher" it will be. "As company B can pay out 90% of its earnings while company A 0%." payout ratio is in no way connected to ROIC, you have no idea what you're talking about
"It pays no attention to what happens to the FCF that is generated." It does, its just you have no idea what are the implicit assumptions behind the cost of equity (hint - it's based off of publicly listed equities, the owners of which all of have the issue of not having direct control over management and payout policy; that's why there is a control PREMIUM assigned to private company valuations ). "I suppose you give the cash hoarder some arbitrary higher discount rate?" it's not arbitrary. You might argue the control premium is arbitrary because it's much more difficult to estimate.
"The concept of “cost of capital” is ridiculous when applied to equity. [...] Most people don’t really understand it" yes, you clearly don't understand it at all. Mostly you don't understand that you do use it yourself, just in the form of a PE multiple.
"Assessing risk using a discount rate is counter intuitive and opaque." not if you understand the concept. "If the discount rate is increased by a small amount you get a wildly different number without really knowing what that exactly implies." same thing happens to mortgate payments when the Fed raises rates. It's unfortunate but unavoidable. "Using probability trees and making things like dividend/buyback payout, holding period, expected return and ROIC explicit would mean there is more to argue with than some vague hard to interpret discount rate." All the stuff you mention can be incorporated into a DCF, but it all relates to estimating cash flows. It won't spare you the difficult task of coming up with a discount rate.
"Terminal value multiple. I thought we were counting cash flows? And now all of a sudden we are using multiples again? What is this nonsense? " You're so clueless you don't even realize a multiple is the inverse of the discount rate. They're the same thing, just presented differently.
Frankly - this is embarassing. Delete this, start over, read a corporate finance book, come back and try again.
My question is how do you in particular estimate ROIC? There seem to be many ways to do it, so I wonder how you are doing it.
Also, you mentioned ROIC = expected return. Could elaborate on this some more? And you expect that ROIC is equal to EPS growth. Could you also tease this out some more?
"The fact that a DCF does not really count cash flows if the company is growing." This is your opinion, not a fact. A fact is "it's 92 degrees in the shade." An opinion is, "It's hot."
Further, a DCF absolutely allows for growth, reinvestment, and cash flows
Also, I can (and do) include ROIC in my DCF.
Also, a DCF can distinguish between cash from operations and cash from cash, cash equivalents and marketable securities.
Cost of capital - Every stock has a cost of equity, even though the number is not explicit, as is the case with the interest rate on debt. Buffett has said his cost of equity for a company whose outlook is predictable is the yield on the long-term U.S. Treasury plus 200-300 basis points. Easy to measure the impact of a change in interest rates on intrinsic in a DCF.
The P/E is the inverse of investor expectations about the company's risk
Terminal value multiple is the sum of the present values of nominal distributable cash beyond the forecast period.
A DCF is a prediction based on what we know today, and not forever. As we learn new information we need to update our assumptions.
I liken a DCF to the amortization of a schedule of a 25- or 30-year mortgage. If the calculator's formulas are correct, then the monthly payment is correct based on credit score, interest rates, down payment, loan-to-value, etc. The amortization does not predict whether the buyer will be transferred out of state, lose their job, get divorced, annoying neighbors, school quality, crime rates, etc.
Last, your spreadsheet which attempts to predict a stock's natural low price floor is clever and shrewd.
Thanks for sharing.
Hewitt Heiserman
Author, "It's Earnings That Count" (McGraw-Hill, 2004)
The problem is that if you use real FCF and there is significant growth early on you need to keep counting for about 70 years, or use a terminal multiple after say 25 years. You sort of kind of take ROIC into account, but like half your value (if you stop counting after 25-30 years) will be determined by the terminal multiple. So then does it really take ROIC into account? Is it even a DCF model?
It is really a "50% counting cash flows and 50% multiple I just pulled out of my ass on discounted net profit 25 years from now" model.
And it is a pain to introduce payout % into this to see how it affects valuation in a company that is in steady state. Plus I truly start to feel ridiculous making some detailed model on what SG&A will be 20-30 years from now. Might as well just take an average growth % and collapse that into a formula instead.
Using the formulas I provided above you can get a far more intuitive estimate in about 30 seconds with more direct inputs and it forces valuer to make a distinction between growth phase and terminal phase.
If I want to see what is baked into a DCF I would then have to plow through 70 columns or rows of some convoluted DCF. And adjusting it would be a pain, a small mistake somewhere and you might get a very different value.
I recently shared my view on Valuation and DCF methods too, and for me, accurately forecasting returns never guarantees peace of mind, especially in “risk-off” environments.
Personally, studying factors like the business model, competitive landscape, and industry dynamics gives me more confidence in uncertain times. Take $META for example; how many have calculated its intrinsic value, but didn’t buy when it dropped below $100 per share in 2022 🤔?
Love your work, great post.
I think we are all nudged towards DCFs when we start investing BC it seems like a high level of intellectual rigour. I remember first reading Damodaran and he discussed relative valuation (multiples) like a shameful thing lol.
After a while you realise you're just plugging whatever assumptions suit you in there.
Your framework is excellent in that it strips the bullshit out, but narrows in past the "finger in the wind 8x feels about right" stuff.
“ROIC is not included in a DCF calculation” - this is wrong, higher ROIC growth needs less investment so incremental FCF higher
I haven't gone through what you do instead it in too much detail but it seems you are essentially doing a shorthanded/convoluted version of a DCF
It sounds like you stopped reading at the start.
It looks like a DCF with static growth/ROIC for a "holding period", and then the same concept of exit multiple. Your "minimum expected return over the holding period" sounds a lot like a cost of equity. Am I wrong?
Re the other points:
- Terminal multiple - don't people usually use a sustainable long-term growth rate and ROIC to get to terminal value? You can get this to spit out a multiple if you want, they're the same. But it's not inconsistent
- Assessing risk using a discount rate is exactly what you are doing with your hurdle rate. Either that or you're not considering risk at all?
- I think you're also wrong that a DCF penalises companies reinvesting FCF to grow ("does not really count cash flow"). Isn't the whole point that if ROIC is above cost of capital, reinvesting all the cash flow increases the value?
So yeah, I think the only point I agree with is that DCFs "pay no attention to what happens to the FCF that is generated". I think that's by design
I'm not a DCF purist or anything btw, very far from it
If you stop counting after say 20-30 years and determine terminal value based on discounted net profit, then are you really taking ROIC into account? I guess sort of, partially? So my second point is half wrong I suppose. I have seen people determine terminal value by using EV/EBITDA.
I disagree that my method is more convoluted. I think a DCF is convoluted, especially when potentially more than half your end value is not even determined by actually counting cash flows.
Building one is time consuming and it is easy to make some small mistake/typo and have it spit out the wrong value without noticing. And if I actually keep counting for 70 years, in order to see what goes into my model you have to scroll through rows and rows of kind of laughable assumptions (because the company will likely not even exist 70 years from now).
My method splits up a valuation into high growth and a mature phase and takes explicitly into account:
-ROIC (all the way)
-Payout % when the company goes into its low growth mature phase. This is a pain to do with a DCF (especially when terminal value is a big % of your valuation).
-Period that you feel confident predicting growth. Since most value investors looking to significantly outperform are not interested in holding a low growth company at a fair/over valuation.
-Risk by explicitly calling it a minimum required return (seems clearer than a hurdle rate).
And all of this can be quickly displayed on 1 page, and a calculation can be done quickly in less than a minute to get an overview of what is reasonable to pay for a stock.
And opportunity cost (which is what cost of capital sort of tries to be I guess?) is determined by how far the current valuation differs from the maximum earnings multiple you should pay based on the formula. And is ofcourse heavily determined by growth period and expected holding period when betting on earnings multiple reverting to the mean.
DCF’s are a helpful tool to assess sensitivities to certain factors, help see bear/ bull valuation scenarios.
Just putting a multiple on something w/o understanding what growth, margin, capital investment assumptions are reflected by that multiple makes less sense than using a DCF.
You have no idea what you're talking about, this article is so bad and wrong it actually made me angry
"The fact that a DCF does not really count cash flows if the company is growing" b/c there is no cash-flow... make your forecast period longer until you get to steady state cash flows.
"ROIC is not included in a DCF calculation" why not? Base your CF forecast of ROIC assumptions if you like
"Say you have company A that generates a 10% ROIC and company B that generates 100%. Company B should clearly be valued higher" it will be. "As company B can pay out 90% of its earnings while company A 0%." payout ratio is in no way connected to ROIC, you have no idea what you're talking about
"It pays no attention to what happens to the FCF that is generated." It does, its just you have no idea what are the implicit assumptions behind the cost of equity (hint - it's based off of publicly listed equities, the owners of which all of have the issue of not having direct control over management and payout policy; that's why there is a control PREMIUM assigned to private company valuations ). "I suppose you give the cash hoarder some arbitrary higher discount rate?" it's not arbitrary. You might argue the control premium is arbitrary because it's much more difficult to estimate.
"The concept of “cost of capital” is ridiculous when applied to equity. [...] Most people don’t really understand it" yes, you clearly don't understand it at all. Mostly you don't understand that you do use it yourself, just in the form of a PE multiple.
"Assessing risk using a discount rate is counter intuitive and opaque." not if you understand the concept. "If the discount rate is increased by a small amount you get a wildly different number without really knowing what that exactly implies." same thing happens to mortgate payments when the Fed raises rates. It's unfortunate but unavoidable. "Using probability trees and making things like dividend/buyback payout, holding period, expected return and ROIC explicit would mean there is more to argue with than some vague hard to interpret discount rate." All the stuff you mention can be incorporated into a DCF, but it all relates to estimating cash flows. It won't spare you the difficult task of coming up with a discount rate.
"Terminal value multiple. I thought we were counting cash flows? And now all of a sudden we are using multiples again? What is this nonsense? " You're so clueless you don't even realize a multiple is the inverse of the discount rate. They're the same thing, just presented differently.
Frankly - this is embarassing. Delete this, start over, read a corporate finance book, come back and try again.
Yeah the article contains a few misleading arguments. And is far from my best. But it seems you got a bit carried away yourself here.
Nice article.
My question is how do you in particular estimate ROIC? There seem to be many ways to do it, so I wonder how you are doing it.
Also, you mentioned ROIC = expected return. Could elaborate on this some more? And you expect that ROIC is equal to EPS growth. Could you also tease this out some more?
I saw in another comment where you describe how you calculate ROIC. For PPE, do you take net PPE (depreciated)?
I like this article because it made me think.
However,
"The fact that a DCF does not really count cash flows if the company is growing." This is your opinion, not a fact. A fact is "it's 92 degrees in the shade." An opinion is, "It's hot."
Further, a DCF absolutely allows for growth, reinvestment, and cash flows
Also, I can (and do) include ROIC in my DCF.
Also, a DCF can distinguish between cash from operations and cash from cash, cash equivalents and marketable securities.
Cost of capital - Every stock has a cost of equity, even though the number is not explicit, as is the case with the interest rate on debt. Buffett has said his cost of equity for a company whose outlook is predictable is the yield on the long-term U.S. Treasury plus 200-300 basis points. Easy to measure the impact of a change in interest rates on intrinsic in a DCF.
The P/E is the inverse of investor expectations about the company's risk
Terminal value multiple is the sum of the present values of nominal distributable cash beyond the forecast period.
A DCF is a prediction based on what we know today, and not forever. As we learn new information we need to update our assumptions.
I liken a DCF to the amortization of a schedule of a 25- or 30-year mortgage. If the calculator's formulas are correct, then the monthly payment is correct based on credit score, interest rates, down payment, loan-to-value, etc. The amortization does not predict whether the buyer will be transferred out of state, lose their job, get divorced, annoying neighbors, school quality, crime rates, etc.
Last, your spreadsheet which attempts to predict a stock's natural low price floor is clever and shrewd.
Thanks for sharing.
Hewitt Heiserman
Author, "It's Earnings That Count" (McGraw-Hill, 2004)
@HewittHeiserman
The problem is that if you use real FCF and there is significant growth early on you need to keep counting for about 70 years, or use a terminal multiple after say 25 years. You sort of kind of take ROIC into account, but like half your value (if you stop counting after 25-30 years) will be determined by the terminal multiple. So then does it really take ROIC into account? Is it even a DCF model?
It is really a "50% counting cash flows and 50% multiple I just pulled out of my ass on discounted net profit 25 years from now" model.
And it is a pain to introduce payout % into this to see how it affects valuation in a company that is in steady state. Plus I truly start to feel ridiculous making some detailed model on what SG&A will be 20-30 years from now. Might as well just take an average growth % and collapse that into a formula instead.
Using the formulas I provided above you can get a far more intuitive estimate in about 30 seconds with more direct inputs and it forces valuer to make a distinction between growth phase and terminal phase.
If I want to see what is baked into a DCF I would then have to plow through 70 columns or rows of some convoluted DCF. And adjusting it would be a pain, a small mistake somewhere and you might get a very different value.
Great post on such an important topic 🙏👍
I recently shared my view on Valuation and DCF methods too, and for me, accurately forecasting returns never guarantees peace of mind, especially in “risk-off” environments.
Personally, studying factors like the business model, competitive landscape, and industry dynamics gives me more confidence in uncertain times. Take $META for example; how many have calculated its intrinsic value, but didn’t buy when it dropped below $100 per share in 2022 🤔?
Very interesting article
How do you calculate ROIC?
There are two types of ROIC, the type that includes acquisitions and excludes it.
I take Property plant and equipment + net working capital (so receivables + inventory - account payable - accrued expenses) - deferred revenue.
And then divide unlevered adjusted (in case of amortization of intangibles) net profit by this amount.
If a large portion of growth depends on acquisitions then include Goodwill as well.
Thanks!
What do you exactly mean by unlevered adjusted net profit?
I assume the company doesn't have debt, even if it does. And add back amortisation from acquisition related intangible assets.