Investing is an art. But it’s also a craft and a science at the same time. In this article, I’ll teach you everything you need to know about an Earnings Growth Model. This models tells you which return to expect from your investments.
thanks a lot for this - really interesting to explore and Im trying to get as much as possible into a google colab notebook. It will be interesting to see how nicely it matches up with the buy-hold-sell sheet
Out of interest, you always use 30% margin of safety? I've just been rounding down the lowest number out of the 10 year average and the LT growth estimate
Ive been looking at a few entries in the buy-hold-sell sheet. Sorry if its been answered elsewhere but I'm a bit confused on how to get to the fair value price based on the calculations here.
10% = 12.0% (EPS Growth) + 0.7% (Dividend Yield) - 0.1* (26.3x - Multiple you could pay)/ Multiple you could pay
"
but I did not understand how to calculate the 'Multiple you could pay'. I tried to reverse engineer it using the Adobe entry in the sheet but it gave me a completely different answer.
It's a good question and there is no golden truth for this.
The multiple you could pay is the PE which I consider fair in the long term. In the buy-hold-sell list, we usually use the 10-year average forward PE for the company in question. Does this make sense?
@Pieter, do you always use the 0.1 in the EEG model for all the stocks in the Watchlist? I am trying to work backwards on all the data to understand and sanity check the different expected return numbers :D
Follow up question. In the Watchlist sheet. Sometimes you use the 10 Year Average Forward PE as the Fair Exit PE. Other times you adjust. For example Adobe:
Fair Exit Multiple 25
10 Year Average Forward PE 32.
A -21% downward adjustment.
For some you just round it it seems, and others you use the exact 10 year average.
It's a nice article, no doubt. But as far as I can find, it does not explain *why*.
Translating that into a simple question: *why* does it make sense to *sum* the dividend per stock price with the change rate (!) of the earnings per share?
In this model, I assume the Forward PE goes back to ‘normal’ (reversion to the mean) within 10 years. 1/10 = 0.1 —> That’s why a factor of 0.1x is used.
It’s a smart idea to also do the calculations under the assumption that the valuation goes back to normal in 5 years. In that case, you need to use 0.2x as a multiplier (1/5). You’ll see that some high-growth companies will be less attractively valued when you do this.
thanks a lot for this - really interesting to explore and Im trying to get as much as possible into a google colab notebook. It will be interesting to see how nicely it matches up with the buy-hold-sell sheet
Out of interest, you always use 30% margin of safety? I've just been rounding down the lowest number out of the 10 year average and the LT growth estimate
It's definitely not exact math. Sometimes I use a 30% margin of safety but I don't always do it that way.
In general, I think it's just important to be conservative with your estimates. It's better to surprise on the upside. :)
thanks
I thought that was probably the case :)
Ive been looking at a few entries in the buy-hold-sell sheet. Sorry if its been answered elsewhere but I'm a bit confused on how to get to the fair value price based on the calculations here.
I saw referenced in another article:
"
Expected return = EPS Growth + Dividend Yield +/- Multiple Expansion (Contraction)
10% = 12.0% (EPS Growth) + 0.7% (Dividend Yield) - 0.1* (26.3x - Multiple you could pay)/ Multiple you could pay
"
but I did not understand how to calculate the 'Multiple you could pay'. I tried to reverse engineer it using the Adobe entry in the sheet but it gave me a completely different answer.
It's a good question and there is no golden truth for this.
The multiple you could pay is the PE which I consider fair in the long term. In the buy-hold-sell list, we usually use the 10-year average forward PE for the company in question. Does this make sense?
I'm sorry, just getting a bit confused and my maths is not all the best. In the other article it uses Visa as an example:
"
Here are the assumptions used for Visa:
EPS Growth = 12.0% per year over the next 10 years
Dividend Yield = 0.7%
I consider an exit PE of 26.3x (= historical average Forward PE) as fair in the long term
Now you need to calculate the multiple you could pay for Visa to receive a return of 10%.
Expected return = EPS Growth + Dividend Yield +/- Multiple Expansion (Contraction)
10% = 12.0% (EPS Growth) + 0.7% (Dividend Yield) - 0.1* (26.3x - Multiple you could pay)/ Multiple you could pay
"
So in the this case 26.3 would be the 10-year average forward PE in that formula. What would be the 'Multiple you could pay' figure in this case?
Sorry Im probably being daft and missing something really obvious
It's just a mathetematical question:
10% = 12% + 0.7% + 0.1*((26.3-X)/X))
If you take out X here, you get at a PE of around 24x!
Thanks Pieter - got it now.
A mathematical question: how do you get to the -1.7% contraction? I know it’s (new-old/old), but i don’t understand the 0.1.
Thanks you!
It's because we estimate that the valuation will normalize in 10 years. That's where the 0.1 comes from :)
0.1 is over 1 year, over 10 years this is 1
With the Earnings Growth model, are you usually conservative with the multiples?
Absolutely.
Better to surprise on the upside :)
That was very easy to read and follow, Pieter! Well done! 🙂
I need to re-read this and run through the exercise in my journal to have the lesson glue itself to my brain. 🖋️
Always feeling blessed to see you here, Boris!
thanks for sharing this simple and valuable model
It's a true honor!
Thank you for continuing to write these in a simple, straightforward way.
It's a true honor, Gary!
Thanks for detailing the model. Always helpful to see how others value companies.
I appreciate you, Joel!
I would replace Dividend Yield with Total Shareholder Yield, which includes share buybacks and debt payments.
Buyback is included in the EPS Growth. That's why we just use the dividend yield.
Otherwise you would include the buyback yield twice. :)
How did you get that the fair P/E was 16x ?
Yeah I couldn’t figure how the P/E ratio in example was 16x either.
Think it’s just an assumption he was making as an illustration
What The Curious LP mentions is correct.
To determine a fair Exit PE, the Microsoft example provides more guidance
@Pieter, do you always use the 0.1 in the EEG model for all the stocks in the Watchlist? I am trying to work backwards on all the data to understand and sanity check the different expected return numbers :D
@Peter Nielsen when you assume that the forward PE will go back to normal after 10 years, multiplying it by 10 is indeed what I do.
When you want to do the calculations with a rerating within 5 years, your multiplier should be 0.2x
Follow up question. In the Watchlist sheet. Sometimes you use the 10 Year Average Forward PE as the Fair Exit PE. Other times you adjust. For example Adobe:
Fair Exit Multiple 25
10 Year Average Forward PE 32.
A -21% downward adjustment.
For some you just round it it seems, and others you use the exact 10 year average.
What causes you to adjust? Caution?
Correct. The better I know the company, the better I can make calculations myself.
When I’m not sure yet, I use the 10-year average. :)
It's a nice article, no doubt. But as far as I can find, it does not explain *why*.
Translating that into a simple question: *why* does it make sense to *sum* the dividend per stock price with the change rate (!) of the earnings per share?
Hoping to get an answer but for some reason my question dropped almost to the bottom.
It's a serious question, I do hope to understand the logic of the model.
Or, if the model is just to get a quick but valuable insight and it's not about strict logic, please tell me 😅
I replied to it but for some reason it disappeared. Didn’t you get a notification of my reply? In that case I’ll write it again. :)
Nope :( Got an email about this response, but nothing previous.
Meanwhile I have an intuitive understanding, but I'd still appreciate to understand the actual logic :)
So here’s a visual explaining the earnings growth model: https://www.google.com/url?sa=i&url=https%3A%2F%2Fwww.compoundingquality.net%2Fp%2Fearnings-growth-model&psig=AOvVaw0hfJRUHcsSHBA3uAJAGgfl&ust=1748006232666000&source=images&cd=vfe&opi=89978449&ved=0CBQQjRxqFwoTCIDxgeaUt40DFQAAAAAdAAAAABAE
In this model, I assume the Forward PE goes back to ‘normal’ (reversion to the mean) within 10 years. 1/10 = 0.1 —> That’s why a factor of 0.1x is used.
It’s a smart idea to also do the calculations under the assumption that the valuation goes back to normal in 5 years. In that case, you need to use 0.2x as a multiplier (1/5). You’ll see that some high-growth companies will be less attractively valued when you do this.
Does this make sense?
Ok i see it now! Sorry! You are dividing the multiple compression over the 10 years to smooth it out. Thank you for explaining.
Exactly!
So then if it’s 9 years it’s 0.09? Sorry for the questions?!
Than it would be 100/9 :)
When it would be 5 years, you multiply it by 0.2x