When I first learned how to build a discounted cash flow model, I thought the difficult part would be the math and building the model.
I was overwhelmed making sure the model was perfect and ensuring the math was right, and I had the right colors and presentation correct.
But once you build a few models, you realize the formulas are actually the easy part, and models are commoditized; you can find so many free ones online.
The difficult part is deciding which assumptions to use.
Because as Warren Buffett said: "It is better to be approximately right than precisely wrong."
I think this is one of the most important lessons in valuation.
A spreadsheet might tell you that a stock is worth exactly $173.42 per share, but that does not mean the company is actually worth $173.42.
That number depends on assumptions about:
* Revenue growth
* Profit margins
* Free cash flow
* The discount rate
* The terminal growth rate or exit multiple
* Share dilution
Two investors can use the exact same valuation model and arrive at completely different values without either one making a mathematical mistake.
For example, imagine a company currently generates $10 billion in annual free cash flow.
Investor A assumes:
* 5% annual free-cash-flow growth
* An 18-times terminal multiple
* A 10% required return
Investor B assumes:
* 10% annual free-cash-flow growth
* A 24-times terminal multiple
* A 10% required return
The second valuation could be more than 50% higher, even though both investors are valuing the same company using the same starting financial information.
The difference comes almost entirely from the assumptions.
This is why I do not think a DCF should be treated as a precise answer.
A DCF tells you what a company could be worth **if your assumptions are correct**.
That is an important distinction.
Instead of saying: "This stock is worth exactly $170 per share."
I think it is more useful to say:
"If the company grows free cash flow between 5% and 10% and trades between 18 and 24 times free cash flow in five years, its estimated value falls somewhere between $125 and $200 per share."
That range is less precise, but it is probably more honest.
It also helps you identify which assumptions matter most.
* What would need to happen for the company to grow at 10%?
* Can it maintain its margins?
* How much capital will that growth require?
* Will the business still deserve a 24-times multiple once it becomes larger and more mature?
* What happens if growth slows or the multiple contracts?
This is where the concept of a margin of safety becomes important.
Because your assumptions will never be perfectly accurate, you generally do not want to buy a stock just because it is trading slightly below your estimated value.
If you estimate that a stock is worth $100 and it trades at $95, that may not be enough.
A small change in growth, margins, or the terminal multiple could eliminate the entire discount.
But if your conservative estimate is $100 and the stock trades at $70, you have more protection if your assumptions turn out to be wrong.
That does not guarantee you will make money, though. It simply gives you more room for error.
I like building three scenarios:
1. A bear case
2. A base case
3. A bull case
The bear case shows what could happen if growth slows, margins decline, or the valuation multiple contracts.
The base case reflects what I think is the most reasonable outcome.
The bull case shows what could happen if the company performs better than expected.
If the stock only looks attractive under the bull case, there probably is not much margin of safety.
If it still looks attractive using conservative assumptions, it may be worth researching further.
I personally use a spreadsheet where I can change the ticker and quickly test different growth rates, margins, discount rates, and terminal multiples.
I use Wisesheets to keep the financial data updated automatically, but the same exercise can be done by manually entering numbers from company filings.
The spreadsheet is not the important part.
The important part is understanding what assumptions are required to justify the current stock price and how much room you have to be wrong.
A good valuation model should not create a false sense of certainty.
It should help you be approximately right while avoiding the risk of being precisely wrong.
How much of a margin of safety do you normally require before buying a stock?