Insight-driven value investing
The market only ever quotes a price. It never tells you what something is worth, and it never tells you what to do next — that judgment has to come from somewhere else entirely.
What is Buffett's actual definition of value investing, in practice?
Buffett has reduced the whole discipline to one sentence: find businesses priced low relative to the cash they will generate over their remaining life, discounted back to today. What he has never done, according to Charlie Munger, is build a spreadsheet to prove it — in sixty-plus years, Munger says he has never once watched Buffett run an actual DCF. Buffett does the arithmetic in his head, and if a business does not scream out at him within a few minutes, he moves on to the next idea.
That is worth sitting with, because it cuts against how the framework usually gets taught. The formula is real — intrinsic value is the present value of future free cash flow — but the formula was never the point. The point is training your judgment well enough that you can feel, quickly, whether a price is low relative to a future the business is very likely to deliver. Low risk, high expected return: that is the entire ask. Everything else is method.
If [the value of a company] doesn't just scream out at you, it's too close.
— Warren Buffett
What does Munger mean by opportunity cost in investing?
Every dollar committed to a new idea is a dollar pulled away from something you already own, so the real hurdle for a new position was never “is this good” — it is “is this better than my best existing holding.” Munger called opportunity cost one of the biggest filters in life: compare a new opportunity to the best one you already have, and if it is not clearly better, there is no reason to act.
Applied to a portfolio, that discipline does most of the work of staying out of trouble. You stop evaluating ideas in isolation — is this stock cheap — and start evaluating them against what is already sitting in the portfolio. If a new idea cannot clear that bar, it is not a buy, no matter how interesting the story is. Cash and your existing winners are not a passive backdrop; they are the active hurdle every new idea has to clear.
Intelligent people make decisions based on opportunity costs.
— Charlie Munger
Why does value investing beat momentum and trading strategies over the long run?
The market only ever quotes a price — it does not tell you what a business is worth, and it never tells you what to do next. Momentum treats the price move itself as the signal, which means a momentum investor is, by construction, reacting to something that has already happened. Value investors are pricing the business directly, so they are positioned before the move, not after it.
Trading compounds the problem. Constantly screening, monitoring, and re-entering costs time, and time spent watching prices is time not spent understanding businesses — a cost that never compounds in your favour the way patient capital does. That is probably why almost every investor with a genuinely long, market-beating record runs a value process rather than a trading one. Graham's point still holds: the market is there to serve you with prices, not to instruct you on what those prices mean.
What's the difference between active and long-term value-investing styles?
Active value investors — the ones hunting turnarounds and special situations — buy with a margin of safety and sell once the price reaches fair value, and that requires a real, numeric view of the company's forward growth and margin, because that is exactly what fair value is built from. A longer-term style does not need that precision. What it needs is conviction that the business will keep getting better, which is a qualitative judgment about management, culture, and the durability of the moat rather than a forecast you can put in a spreadsheet.
Both are value investing in the strict sense — priced against intrinsic worth, bought with a buffer — they just draw on different skills. I do not have the time, or honestly the interest, to run the quantitative side that hard, so my own approach has settled firmly into the second camp. Insight has mattered far more to my results than modelling precision ever has.
Where does real investment insight actually come from?
Not from a balance sheet — knowing how to read one is necessary, but nowhere near sufficient. Insight comes from time spent inside or adjacent to an industry, from conversations with people closer to it than you are, from reading widely, and from the slower work of just thinking something through. Writing is one of the better tools for that last part, because it forces a vague hunch into a claim you can actually test.
Nvidia is the cleanest example from my own portfolio. I had been experimenting early with AI coding agents, which put me close to where real demand for compute was actually forming, well before it was a mainstream story. When I ran a reverse-DCF on Nvidia inside Invest Board using nothing but its historical growth and margins, the stock was still priced under fair value. The valuation work was not sophisticated, and it did not need to be — the insight was already there before I opened the model; the model just confirmed the market had not caught up yet.
Alphabet is an older one, from when I was working as a developer at Sony Mobile. The stock had been sold off hard because analysts were worried Alphabet could not compete on mobile advertising and search as the industry moved to smartphones. From where I sat, that fear looked backwards: we had just switched Sony's phones onto Android, and digital advertising was only beginning to take off. I remember a lunch in downtown Lund where a friend and I agreed that a company like Alphabet would simply be much bigger in ten years. The valuation at the time was not demanding at all — the market was pricing in a competitive loss that, from inside the industry, did not look like the likely outcome.
For a holding meant to be owned for years rather than months, what matters most shifts. Management quality, company culture, optionality, and the durability of the moat carry more weight than any single quarter's numbers, and judging them well is more a matter of feel than of formula. That is exactly why staying inside your circle of competence matters so much for this style: conviction built from genuine understanding is what lets you sit through a drawdown, or buy more into one, instead of selling on the first bad print. It also helps that this kind of conviction is easier to build in sectors that are growing quickly, where being early is worth more than being precise.
None of this removes the need for a sell discipline. My own rule is to trim once a position gets meaningfully overvalued, and to sell in full only when the story itself changes — when it is clear the moat has narrowed and a competitor now holds the better hand. The mechanics of that decision, and how to tell a broken thesis from a temporary setback, are their own subject, covered in when to sell a stock and investment kill criteria.
Can AI generate genuine investment insight?
Usually not, and it is worth being honest about why: large language models are trained to reproduce the consensus view, and insight is close to the opposite of that — it is a view the consensus has not priced in yet. AI is genuinely useful for the mechanical half of the work, pulling numbers into a reverse-DCF or summarising a filing, but the read on whether a business is actually improving still has to come from an independent thinker, not from a model averaging what everyone already believes.
