Table of Contents
There is a strange ritual that happens in finance every single day. A young analyst opens Excel, builds a discounted cash flow model with twelve tabs and three hundred assumptions, and produces a number that looks suspiciously like the price the company is already trading at. Their managing director nods. The client nods. Everyone agrees the company is worth what everyone already thought it was worth. And then they bill for it.
This is the dirty secret of valuation work. Most DCF models are not analytical tools. They are elaborate justifications. They are the financial equivalent of a horoscope written after the fact, where the stars somehow always predicted exactly what just happened.
If you want to build a DCF that actually means something, you have to start by understanding what you are really doing when you build one. And what you are really doing is making a series of philosophical claims about the future dressed up in the costume of mathematics.
The Model Is Not the Point
Let us begin with a confession that most finance professors will never make. The DCF is not really about precision. It cannot be. You are projecting cash flows ten years into the future for a business that may not look anything like itself in three years. You are picking a discount rate that pretends to capture risk when risk is, by its nature, the thing you do not know about.
So why do we do it?
We do it because the process of building the model forces us to make our assumptions explicit. That is the entire point. The output number is almost beside the point. The real product of a good DCF is the conversation you had to have with yourself to produce it. What do I actually believe about this company? What has to be true for this to work? Where am I making a heroic assumption and calling it a reasonable one?
Think of it like the difference between a map and a journey. A bad analyst treats the DCF as a destination. A good analyst treats it as a map of their own thinking, useful precisely because it shows where the thinking is weakest.
The Tyranny of False Precision
There is a famous problem in physics called the coastline paradox. The length of a coastline depends entirely on how closely you measure it. Use a ruler the size of a continent and Britain has a tidy perimeter. Use a ruler the size of a grain of sand and the coastline becomes nearly infinite. The closer you look, the less stable your answer becomes.
Valuation has the same problem and most people refuse to acknowledge it. A model with revenue projections down to the third decimal place is not more accurate than one with round numbers. It is just more confident. And confidence without basis is the single most expensive mistake in finance.
When you see a DCF that produces a target price of $147.82, you should be suspicious. Not because the math is wrong, but because the math is pretending to know things it does not know. The honest answer is almost always a range, and that range is almost always uncomfortably wide.
A good rule of thumb is this. If your model produces a single point estimate, you are doing it wrong. If it produces a range so wide that it makes you uncomfortable, you are probably doing it right.
Garbage In, Cathedral Out
The classic complaint about DCFs is garbage in, garbage out. This is true but incomplete. The real problem is worse. It is garbage in, cathedral out. You can feed a DCF model the most absurd assumptions in the world and it will still produce an answer that looks polished, defensible, and intelligent. The model launders bad thinking into respectable looking output.
This is why the most important skill in valuation is not modeling. It is being able to look at your own work with suspicion. Every assumption needs to be interrogated. Why is revenue growth ten percent and not eight? Why does the margin expand? Why does working capital behave the way you assumed it would? If the answer to any of these questions is some version of “because that is what management said” or “because that is what the analyst consensus shows,” you are not doing valuation. You are doing note taking.
The best analysts I have ever read have a particular quality. They sound a little nervous. They hedge. They show their work. They tell you what they do not know. The worst analysts sound like they have figured it out, and they almost always have not.
The Terminal Value Problem
Here is something that should bother you more than it does. In a typical DCF, somewhere between sixty and eighty percent of the total valuation comes from the terminal value. That is the chunk that represents everything beyond your explicit forecast period. Which means that the work you put into projecting the next five or ten years in painstaking detail accounts for maybe a quarter of your answer. The rest comes from a single formula that assumes the company grows at a steady rate forever.
Forever is a long time.
This is the part of valuation that should keep you up at night. We are valuing perpetuity, and we are doing it with a Gordon growth formula that was designed when companies looked very different than they do now. We assume terminal growth rates that hover just below the long term growth rate of the economy, as if every business gracefully matures into a sleepy utility. Most businesses do not. Most businesses die. Some explode. A few become trillion dollar empires that nobody saw coming.
The honest way to deal with terminal value is to acknowledge that you are essentially making a faith based statement about the long term. You can dress it up with exit multiples or perpetuity growth formulas, but underneath the math is a belief. You believe the business survives. You believe it remains relevant. You believe the world that allowed it to thrive continues to exist. Each of those beliefs deserves to be stated out loud, not buried in a formula.
Discount Rates Are Stories
Now we come to the discount rate, which is where most DCFs go quietly insane. The textbook teaches you to calculate a weighted average cost of capital using betas, risk free rates, and equity risk premiums. The textbook makes this sound like physics. It is not physics.
The beta you pull from a financial database is calculated from historical price movements. It tells you how a stock moved relative to the market in the past. It tells you nothing about how risky the business actually is going forward. A company can have a low beta because nobody has noticed it yet, not because it is safe. A company can have a high beta because it had a few volatile quarters, not because its underlying business is uncertain.
The equity risk premium is even worse. There is no agreement on what it should be. Different scholars produce different numbers. Different time periods produce different numbers. You are essentially picking the one you like best and pretending it was handed down from above.
What this means is that your discount rate is, in practice, a knob you can turn to get the answer you want. Want the company to look cheap? Use a lower discount rate. Want it to look expensive? Crank it up. The model will accommodate you. It has no spine.
The mature response to this is not to abandon the discount rate. It is to be honest about what it is. A discount rate is a story you are telling about how risky you think the future is. Tell the story explicitly. Defend it. Show what happens to your valuation when the story changes. Sensitivity tables are not a nice extra. They are the actual output of an honest DCF.
The Qualitative Shadow
Here is the part nobody wants to say out loud. The best valuation work has very little to do with the numbers in the model. It has to do with the quality of the qualitative thinking that produced those numbers.
A model is only as good as your understanding of the business. If you do not understand why customers buy the product, what would make them stop, who the competitors are, how the industry evolves, and what management is actually capable of executing, then no amount of cell formatting will save you. You will be a very organized person producing a very wrong answer.
This is why analysts who come from operating backgrounds can produce better valuations than analysts who came up through pure finance. They have an intuition for what is plausible. They know that doubling sales while shrinking the sales force is not a strategy. They know that margin expansion does not just happen because you put it in cell C47. They have seen what it actually takes to run a business, and that humility shows up in their models as restraint.
The lesson here is that you should spend more time reading about the business and less time formatting the spreadsheet. You should talk to customers if you can. You should read the angry reviews and the love letters. You should understand the supply chain. You should know who the founder is and whether they sleep well at night. All of this seems like a distraction from the model. It is actually the model.
When to Trust It
After all this, you might reasonably wonder whether the DCF is worth doing at all. The answer is yes, but for different reasons than you were taught.
The DCF is worth doing because it forces structure on otherwise mushy thinking. It is worth doing because it makes you specify your assumptions and own them. It is worth doing because the act of building it teaches you the business in a way that no other exercise can. It is worth doing because, when you compare your model to the market price, the gap tells you something interesting. Either the market knows something you do not, or you know something the market does not. Both are worth investigating.
What it is not worth doing is treating the output as truth. The number that pops out of your model is a hypothesis, not a verdict. It is a starting point for argument, not the end of one.
A Final Word on Humility
The greatest investors in the world have a quality that I find genuinely strange. They are confident enough to take large positions, but humble enough to admit they are often wrong. They use models, but they do not worship them. They have strong views, but they hold those views loosely.
The DCF, used well, can be a tool for exactly that kind of thinking. It can clarify what you believe and force you to face what you do not know. Used poorly, it becomes a way to manufacture false confidence in a uncertain world. A lot of DCFs are used poorly.
So if you want to build a valuation model that is not garbage, here is the secret. Spend less time on the spreadsheet and more time on the business. Be suspicious of your own assumptions. Show your ranges and not your point estimates. Acknowledge what your discount rate really is. And remember that the goal is not to produce a number. The goal is to produce understanding.
The number is just the receipt.
