Most stock pages on this site show a discounted cash flow estimate — a read on what the business might be worth based on the cash it should produce, rather than what the market is quoting today. And it doesn't change the score at all.
That sounds like a contradiction. It isn't. It's the most considered decision in the whole model, and it comes down to a simple observation: a DCF is the most powerful idea in valuation and the easiest one to fool yourself with.
The problem with a DCF is that it always gives you an answer
Feed it a growth rate and a discount rate and it will hand back a precise number — a dollar value per share, carried to two decimal places. That precision is the trap. Nudge the growth assumption a few points, shade the discount rate down a little, and the same company is suddenly worth half again as much. Nothing about the business changed. You changed two numbers in a spreadsheet.
That variability is why the DCF sits outside the score. I didn't want something that volatile having a say in the ranking. A metric that can swing fifty percent on assumptions I chose shouldn't be allowed to outvote the fundamentals I measured.
So the question I had to answer wasn't "how do I build a better DCF." It was "how much should I let this thing matter."
Keeping the inputs honest
The first thing I did was refuse to let the assumptions wander into fantasy.
The growth rate is capped at 25% a year. Sustaining 25% over a decade is already a heroic outcome; anything above that isn't a forecast, it's a wish. It's floored too — a shrinking business is modeled as declining, not driven to zero, because businesses in decline usually keep producing cash for a while.
The discount rate is boxed between 6% and 15%. Below six, you're claiming a business is nearly as safe as a Treasury. Above fifteen, the math has stopped describing a real required return. It's built from the 10-year Treasury yield, which the site now reads live rather than carrying as a number I set once and forgot.
Terminal growth — the rate the company grows at forever, once the explicit forecast ends — comes off that same Treasury yield and sits about a point and a half below it. No company outgrows the economy forever, and every DCF that claims otherwise is quietly assuming it. The reason the two are tied together is less obvious, and I got it wrong at first: both are nominal rates. Lifting the discount rate while leaving long-run growth where it was asserts that every move in rates is a real-rate move with no inflation in it. When I measured that assumption it accounted for almost the entire effect — six percent off every fair value on the site, from a decision nobody had actually made.
None of these clamps make the estimate more accurate. They make it harder for me to talk myself into a number I wanted. That's a different goal, and for this purpose it's the more useful one.
Why three time frames instead of one
Most DCFs pick a horizon and defend it. Ten years is conventional. Five is common. The choice is usually arbitrary and then argued for after the fact.
I run three — one year, five years, and ten — and average the result.
The reason isn't mathematical. It's that not every investment thesis lives on the same clock. Some names are a story about the next few quarters: a turnaround landing, a margin recovering, a cycle turning. Others only make sense if you're prepared to hold through most of a decade. If I picked a single horizon, I'd be quietly assuming every company on the site should be judged on the same timeline, and that isn't true.
Running all three and averaging covers all of them, consistently, without me having to guess which kind of thesis a given name is. The short horizon leans heavily on what the business is worth roughly as it stands. The long one gives the growth assumption room to compound before the terminal value takes over. A company that looks cheap across all three is telling you something different from one that only works on the ten-year view.
That's the same principle as everywhere else in the model: apply one rule to every name and let the differences between companies show up in the output, rather than tuning the method per stock until it says what you expect.
Why it stays out of the score
Here is where the caution becomes concrete.
The quality-growth score that ranks every stock on this site doesn't use the DCF at all. The score comes from the quality, growth and value pillars, and nothing the DCF produces feeds into any of them. The only place it still counts is the site's older composite score, a fallback for the few companies the quality-growth screen can't score, where it can move the number by at most two points.
The reasoning is simple. A DCF telling me a company is 40% undervalued is not, to me, evidence of anything — it's inside the range that my own choice of inputs could have produced. A number that sensitive to assumptions I picked shouldn't decide where a stock ranks, even a little.
So the DCF sits beside the score on the stock page, not inside it. It contributes nothing to the ranking. That's the system working as intended.
Running it backward
Where I don't show it at all
There's one set of companies that gets no estimate from this model, and it isn't the cash-burners — those are obvious, since you can't discount a cash flow that doesn't exist yet.
It's banks, insurers and property trusts. For a lender, "free cash flow" measures how fast the balance sheet is growing rather than what an owner could take out: loan originations and repayments run straight through operating cash flow. Discount that number and you get a confident-looking answer with nothing behind it, and it fails in the direction that flatters. Capital One came out at $974 a share against a $205 price, labeled as a margin of safety, on a page that also told the reader the model doesn't apply to banks. Both of those couldn't be true.
So the model declines instead of guessing. Those names get a financial-strength card — book value, return on equity, income cover — and no fair value. I'd rather show nothing than show a number I know is measuring the wrong thing.
There's a second way to use the same model, and it's the one I find more useful.
Instead of assuming a growth rate and asking what the business is worth, you can take the price as given and solve in the other direction: what growth would the market have to be assuming to justify what it's charging today? Same machinery — the same ten-year window, the same fade to the terminal rate, the same discount rate — just run backward until the model's per-share value lands exactly on the current price.
The output isn't a dollar figure. It's a sentence: the price implies free cash flow growing about 22% next year, slowing to 3% by year ten.
That's a question I can actually answer. I have no reliable instinct for whether a business is worth $180 when it's trading at $90 — that gap is exactly the kind of number my own inputs could have manufactured. But I can take a starting rate of 22% and hold it against what the company has actually delivered, what its industry supports, what its best competitor has managed. If a business has grown at 8% for five years and the price is assuming it starts growing at 22%, I don't need a fair-value estimate to understand what I'm being asked to believe.
It also fails more gracefully. A fair-value gap of "121% above fair value" reads as broken even when the arithmetic is perfectly correct — a reader's first thought is that the model is wrong, not that the price is demanding. An implied growth rate never looks broken. It just looks high, and high is something you can have an opinion about.
The search has limits, deliberately. It looks between a 25% annual decline and 60% annual growth, so a price implying more than that reads as "at least 60% a year" rather than pretending to a precision the method doesn't have. If a stock is priced beyond the top of that range, the exact figure stopped mattering some distance back.
What it's actually for
So why show it at all?
Because it's another data point, and it's a nice cherry on top when the metrics and the DCF line up.
That's the honest description of its job. The score leans mostly on measured figures — what the business has earned, how efficiently it turns capital into profit, what it costs against its own history and its peers. The DCF looks forward, with a different method. When those two approaches land in the same place, I feel more confident in that name than either would justify on its own.
That's confluence, and it's worth more than either signal alone precisely because the two methods reach their answers in different ways. When they disagree, that's information too — usually that the market is pricing in a future the fundamentals don't yet support, in one direction or the other.
What it is not is a price target. I don't buy because a DCF says a stock is worth more than it costs. I've seen too many models produce a confident number for a business that then went nowhere for three years. The estimate is a consideration, not a conclusion.
When it doesn't apply
There's one more piece of restraint worth mentioning: for companies burning cash, the DCF returns nothing at all.
A discounted cash flow model on negative free cash flow is meaningless — you'd be discounting a stream of losses and getting a negative valuation, which tells you what you already knew. So for those names, no estimate appears. It isn't a low score or a warning; there's simply no number, because producing one would be false precision on a business the method can't describe.
I'd rather show nothing than show something that looks authoritative and means nothing.
That, in the end, is the whole philosophy here. The forward estimate earned a place beside the score, not in it, because that's the size of the role it can honestly fill. Run backward, the same model earned more of my attention, because "what is this business worth" is a question I have to trust the inputs to answer, and "what is the price already assuming" is a question the model answers whether I trust it or not.
The Bull Rankings is automated fundamentals research for education, not personalized investment advice and not a recommendation to buy or sell any security. Always do your own research.