Company Deconstruction Series · Part 5. The first four lenses (business, moat, financials, capital allocation) together answered one question: is this a good company? This lens answers a completely independent question: is it, right now, a good price? These two questions must be kept separate.
1. The Most Expensive Confusion: Good Company = Good Investment
This lens starts by correcting the most widespread and most expensive confusion of them all – "good company" and "good investment" are two entirely different questions, and countless people conflate them.
This confusion sounds harmless, even self-evident: a good company – surely buying it is a good investment, right? But it's precisely the most dignified way to lose money. Because – if you buy a great company at too high a price, it remains a terrible investment.
The logic is hard: your return depends not just on how good the company is, but on how high a price you paid for that goodness. If all the "goodness" of a great company (its growth and profits for many years to come) has already been fully and preemptively priced into today's high stock price, then even if the company actually delivers all that goodness, you make no money – because you've already prepaid for it. Worse, if it delivers even slightly less than that "perfect" expectation, you lose money.
I've said this repeatedly in industry research – Nvidia's moat is real, Palantir's moat is real, but their stock prices may already embed the perfect assumption that "the moat will never be eroded and growth will always be high." Stocks priced for perfection leave no room for error; and reality always makes errors.
So, confirming "this is a good company" through the first four lenses by no means implies "buy now." It only means "this company is worth my time for the next lens – valuation." A good company is your ticket in; a good price is the reason you pull the trigger. Separating the two is the starting point of all discipline in this lens.
2. The Weighing Machine and the Voting Machine: What Is Valuation Really Valuing?
So what exactly is valuation trying to estimate? Graham's classic analogy says it all – "In the short run, the market is a voting machine; in the long run, it is a weighing machine."
- Short term, the market is a voting machine: stock prices are decided by emotion, money flows, narratives, and popularity. A sexy story, a wave of money, can make a stock price severely depart from its true value for the short term – overshoot on the upside (euphoria) or downside (panic). Short-term prices reflect "popularity."
- Long term, the market is a weighing machine: no matter how hot the emotion, it will eventually cool; no matter how expensive the bubble, it will eventually revert. Over a long enough time horizon, stock prices will eventually converge toward the company's true "weight" (intrinsic value) – it weighs the cash the company actually earns.
Valuation is the process of estimating the weight on that "weighing machine" – approximately, what is this company's intrinsic value; and is the current market price (the voting machine's result) above or below that weight?
This distinction gives us an incredibly useful perspective: price and value are two different things. Price is what the market quotes you (the noise of the voting machine); value is what the company itself is (the fact of the weighing machine). The core act of investing is to exploit the gap between the two – buy when price is far below value (market panic, mispricing), and be wary or sell when price is far above value (market euphoria, overvaluation). This is the wisdom of Graham's "Mr. Market": he quotes you a price every day; you don't have to heed his mood, you only need to judge whether the price he offers, relative to the value you calculate, is cheap or expensive. You exploit the market's emotions rather than being exploited by them.
3. Decomposing the Three Sources of Return
To make valuation actionable, here's the most practical framework I know – decompose your long-term return from buying a stock into three sources:
Return ≈ Earnings Growth + Valuation Change + Dividends/Buybacks
- Earnings Growth: how much the company's earnings grow each year. This is the "weighing machine" part, determined by the quality of the business (the first four lenses).
- Valuation Change: whether the multiple (P/E, etc.) that the market is willing to pay for those earnings expands or contracts during your holding period. This is the "voting machine" part.
- Dividends/Buybacks: cash returned directly to you (dividends + buyback shareholder yield).
The power of this decomposition is that it makes clear what "buying expensive" really means:
When you buy at a high valuation, you are effectively doing two things – you are betting that "earnings growth" will be strong (the first term), but you are simultaneously turning "valuation change" (the second term) into your enemy. Because the valuation is already high, it is more likely to contract (revert to the mean) in the future than to continue expanding. So even if the company's earnings grow as fast as you hope, the contraction in valuation will eat most of your return. That's what "pricing for perfection" looks like in math: you earn the earnings growth but get offset by valuation contraction.
Conversely, when you buy a good company at a fair or low valuation, you put all three terms on your side – earnings growth makes you money, valuation has room to expand (repair) rather than pressure to contract, and dividends/buybacks provide a floor. This is why "good company + good price" equals a good investment: it makes all three sources of return work for you simultaneously, instead of turning one of them (valuation) into a drag.
So when looking at valuation, don't just stare at the isolated number "what's the P/E?" Look: at this current valuation, is the 'valuation change' term more likely to be my friend or my enemy in the future? At high valuations it's probably your enemy; at fair valuations it could be your friend.
(As an aside, this also ties back to macro – the "valuation change" term is deeply affected by interest rates. Rates are the gravity of valuation; in an environment where the rate regime is shifting higher, overall valuations face contraction pressure. So judging valuation cannot be separated from the interest rate regime.)
4. Valuation Is Not Precision – It's a Range with a Margin of Safety
Having said this, I must dispel a common misconception about valuation – the purpose of valuation is not to calculate a precise "intrinsic value" number, but to judge a "range," and to demand a sufficient margin of safety.
Many people think valuation means building an elaborate DCF model (projecting and discounting decades of future cash flows year by year, producing an intrinsic value precise to the decimal point). But this is a dangerous illusion of precision – future cash flows are fundamentally unknowable (it's hard enough to forecast next year, let alone ten years), and any model that spits out a precise number is a "precise error." Change the assumptions slightly, and the conclusion is wildly different.
The right attitude is Keynes's saying – "It is better to be roughly right than precisely wrong." Valuation doesn't need a precise point; it needs an approximate range (roughly, what is this company worth to what?), and then – only buy when the price is significantly below the lower bound of that range. The gap between the "price" and the "lower bound of the value range" is what Graham called the margin of safety.
Why is margin of safety crucial? Because you will inevitably be wrong – your understanding of the business will have biases, the future will surprise you, your value range estimate will have errors. The margin of safety is the cushion you reserve for these inevitable mistakes. You buy cheap enough that even if your judgment is partly wrong, the discount protects you from big losses; and if your judgment is right, the discount is your excess return.
So my practical approach to valuation is: don't pursue precision (that's an illusion); only pursue a conservative value range plus a sufficiently large margin of safety. I don't need to know if a company is precisely worth 103 or 107; I only need to roughly judge it's worth "a little over 100," and then when the market panics and it drops to 70, I buy with a margin of safety. This kind of "roughly right plus margin of safety" is far more reliable than a "precise DCF," because it is built on an honest acknowledgment of one's own ignorance.
5. Two Traps: Overpaying and Value Traps
This lens of valuation must simultaneously guard against two traps – equally deadly:
Trap One: Paying too much for a good company (growth trap). This is what we discussed – good company + high valuation = mediocre or even terrible investment. Attracted by a great story and beautiful fundamentals, you buy when the valuation has already priced in the future, then get tortured by repeated valuation contraction. A good company is not an excuse for a high price.
Trap Two: Buying a bad business cheap (value trap). This is the opposite trap, equally common. You see something "cheap" (low P/E, low P/B) and buy it, only to find it is a declining bad business – it's cheap for a reason (fundamentals are deteriorating), and it will get cheaper. You think you are catching a bottom, but you're catching a falling knife. Cheapness alone is not a reason to buy; cheapness relative to value is. A business with persistently low ROIC and a crumbling moat – no matter how low the P/E – is likely a trap.
See the pattern? These two traps precisely demonstrate why "business quality" (first four lenses) and "price" (this lens) must be combined: Look only at price and ignore quality, you fall into the value trap (buying a cheap bad business); look only at quality and ignore price, you fall into the growth trap (buying an expensive good business). Only "good business" + "good price" simultaneously satisfied equals a truly good investment. That's why all six lenses are indispensable, and valuation must follow right after quality assessment.
6. Final Thoughts
Valuation is the most disciplined and the most counterintuitive of the six lenses.
It is counterintuitive because it demands that when everyone is euphoric, the good company's story is most compelling, and the price is highest, you say "too expensive, don't buy"; and when everyone is panicked, the good company is being sold off indiscriminately, and the price is lowest, you say "cheap enough, buy." It demands that you calmly separate the excitement about "a great company" from the judgment of "a good price" – acknowledging that a great company may not be a good investment right now.
And its core is a deep humility: admitting that the future cannot be precisely predicted, so you don't pursue precise valuation, only a conservative range and enough margin of safety, using that cushion to protect yourself from the mistakes you will inevitably make. Valuation is not about calculating the future accurately; it's about giving yourself the odds to win even when you can't see the future clearly, by buying cheap enough.
If only one sentence stays:
A good company and a good investment are two separate questions. Your return = earnings growth + valuation change + shareholder returns – when you price in perfection, valuation turns from friend to enemy. So after confirming a good company, calmly and independently ask: does this price give me enough margin of safety? If not, no matter how great the company, wait.
Next up, the finale of this series, and the veto power hanging over all the lenses – circle of competence: for businesses you don't understand, no matter how good or how cheap, why the only correct move is to walk away.
——
Risk disclaimer: This article is a study of company analysis frameworks. The companies mentioned are only used as analysis examples and do not constitute any investment advice. Market risk exists; invest with caution.
On investment analysis, market insight, and asset allocation — not chasing short-term noise, but understanding what truly drives long-term returns.
