This is a beautifully written and thoughtful post, I am very happy I found it on Christmas Day lol. I am a first-year undergraduate student, and if you ever come out with a valuation course, I would gladly pay 1000 CAD. Otherwise, I ordered expectations investing and wanted to ask what courses you recommend for some active practice/learning? Also, learning Excel? I wanted to give some pushback on the terminal value here though. What is your opinion of the way Constellation Software does valuation? They do scenario analysis (walking wounded, etc.), and I believe they are thoughtful in analyzing whether incremental investment is hitting their IRRs. Still, if you use 20% hurdle rates and can forecast a couple of years reliably (which can be done when a software company has low churn), then the terminal value is not a big factor either, so discounted PV works. What am I missing here? Why not just use a higher hurdle rate? I would LOVE your thoughts here. You are so COOL, THANK YOU SO MUCH
Hi Smaug, glad you found my post on Christmas Day :)
In terms of courses, I'd recommend basically all of Mauboussin's work (https://www.michaelmauboussin.com/writing), though that'd be quite a lot. Expectations Investing is a fantastic place to start. They also have a website (expectationsinvesting.com) that should be used as you're reading the book.
There are a lot of gems in Mauboussin's writing catalogue which I recommend you go through if you're interested.
I'm not great at Excel myself, so no advice from me there. (I rely more and more on AI for formulas and more complicated models)
Scenario analysis is something I talk about as well. From my understanding, Constellation Software primarily uses higher hurdles because their investments are generally riskier than ours. Investing in smaller, private software companies comes with higher risks than investing in proved, listed companies.
Either way, the hurdle rate is entirely up to you to decide. If you're looking for a 15% annual return, you should set your discount rate at 15%. It really is that simple. I don't believe in overly complicated WACC calculations. Just know that the higher your hurdle rate, the stricter you become and the less opportunities there will be available to you.
I believe checking the percentage of terminal value (TV) relative to the total discounted cash flow (DCF) is a crucial sanity check. One approach is to solve for the discount rate and prioritize investments with higher internal rates of return (IRR), which naturally reduces the reliance on terminal value.
I disagree with Michael Mauboussin on forecasting beyond 10 years. That often leads to the illusion of precision. For example, in 2015, who could have confidently predicted that Microsoft would play well the cloud game and lead the AI revolution? Or how Intel and Nvidia would evolve? History shows that the top 10 companies change dramatically every decade—remaining at the top is the exception, not the rule.
Thanks for your comment and glad you liked it. I agree it's difficult to forecast beyond 10 years, but your Microsoft example may not be valid for all companies. Yes, forecasting in the tech sector is particularly difficult, but forecasting beyond 10 years for a company like Costco is reasonable in my opinion.
I agree. However, it’s not necessarily about predicting what will happen, but rather forecasting what you find reasonable or what you’re comfortable with.
The point is not to predict. It's to make assumptions you're comfortable with. You're never going to be right when forecasting, whether you're forecasting just 5 years or 30 years.
This is a beautifully written and thoughtful post, I am very happy I found it on Christmas Day lol. I am a first-year undergraduate student, and if you ever come out with a valuation course, I would gladly pay 1000 CAD. Otherwise, I ordered expectations investing and wanted to ask what courses you recommend for some active practice/learning? Also, learning Excel? I wanted to give some pushback on the terminal value here though. What is your opinion of the way Constellation Software does valuation? They do scenario analysis (walking wounded, etc.), and I believe they are thoughtful in analyzing whether incremental investment is hitting their IRRs. Still, if you use 20% hurdle rates and can forecast a couple of years reliably (which can be done when a software company has low churn), then the terminal value is not a big factor either, so discounted PV works. What am I missing here? Why not just use a higher hurdle rate? I would LOVE your thoughts here. You are so COOL, THANK YOU SO MUCH
Hi Smaug, glad you found my post on Christmas Day :)
In terms of courses, I'd recommend basically all of Mauboussin's work (https://www.michaelmauboussin.com/writing), though that'd be quite a lot. Expectations Investing is a fantastic place to start. They also have a website (expectationsinvesting.com) that should be used as you're reading the book.
There are a lot of gems in Mauboussin's writing catalogue which I recommend you go through if you're interested.
I'm not great at Excel myself, so no advice from me there. (I rely more and more on AI for formulas and more complicated models)
Scenario analysis is something I talk about as well. From my understanding, Constellation Software primarily uses higher hurdles because their investments are generally riskier than ours. Investing in smaller, private software companies comes with higher risks than investing in proved, listed companies.
Either way, the hurdle rate is entirely up to you to decide. If you're looking for a 15% annual return, you should set your discount rate at 15%. It really is that simple. I don't believe in overly complicated WACC calculations. Just know that the higher your hurdle rate, the stricter you become and the less opportunities there will be available to you.
Happy to continue the discussion!
This was absolutely full of insight, thank you so much.
Keep up the great work my friend!
Thanks, I'm glad you found it useful!
Great article with plenty of solid points!
I believe checking the percentage of terminal value (TV) relative to the total discounted cash flow (DCF) is a crucial sanity check. One approach is to solve for the discount rate and prioritize investments with higher internal rates of return (IRR), which naturally reduces the reliance on terminal value.
I disagree with Michael Mauboussin on forecasting beyond 10 years. That often leads to the illusion of precision. For example, in 2015, who could have confidently predicted that Microsoft would play well the cloud game and lead the AI revolution? Or how Intel and Nvidia would evolve? History shows that the top 10 companies change dramatically every decade—remaining at the top is the exception, not the rule.
Thanks for your comment and glad you liked it. I agree it's difficult to forecast beyond 10 years, but your Microsoft example may not be valid for all companies. Yes, forecasting in the tech sector is particularly difficult, but forecasting beyond 10 years for a company like Costco is reasonable in my opinion.
Yes but still "It's tough to make predictions, especially about the future."
I agree. However, it’s not necessarily about predicting what will happen, but rather forecasting what you find reasonable or what you’re comfortable with.
Also, making those assumptions explicit! Learned so much here
In an uncertain world, predicting 20 years seems exaggerated to me. I can barely handle 3 years of estimates, ha!
I'll repeat my answer to your note:
The point is not to predict. It's to make assumptions you're comfortable with. You're never going to be right when forecasting, whether you're forecasting just 5 years or 30 years.
Tell me, how do you approach valuation?