Stock Price vs Intrinsic Value

How long a gap can stay open

The useful version of this idea is not that price converges to value. It is that price and value are linked by a rope, not a rod, and the rope is long. Stocks can trade at twice a defensible estimate of value for years, and at half of it for just as long.

That has a practical consequence. A conviction that something is mispriced is not a trade until something connects it to a timeframe: an earnings recovery, a catalyst, a rate cycle turning, a forced seller finishing. A position sized for a three-month thesis that takes four years will be closed at a loss long before the thesis is tested, and that outcome has nothing to do with whether the valuation work was correct.

The assumption doing all the work

In almost every valuation model, one input dominates. In a growth company it is the terminal growth rate, which usually accounts for most of the computed value, and which is the input nobody can defend beyond ten years out. In a cyclical it is the margin assumed at mid-cycle. In a bank it is the loss rate on the loan book.

The discipline is to find that input and state it as a claim. Not "the shares are worth $40", but "the shares are worth $40 if this business can compound cash flow at 4% for two decades and the market keeps demanding 9%". Written that way, the estimate becomes testable, and it becomes obvious what would have to change for it to be wrong.

What a falling price does and does not mean

A rising price does not confirm that value is rising. It confirms that buyers were more motivated than sellers. A falling price does not confirm that value is falling either, but it is also not neutral: markets are sometimes right early, and a stock that halves is occasionally telling you about something in the accounts that you have not found yet.

The trap in both directions is using the price to update the value estimate without new information. That is how a thesis quietly turns into a rationalisation. The honest question after a 30% fall is whether any input in the model has changed, and if none has, whether you are certain enough to size up or merely unwilling to be wrong.

Using an estimate as a boundary rather than a target

The most durable use of intrinsic value is as a boundary. Rather than one number, produce a range under conservative and aggressive assumptions, and refuse to buy above the top of it. That turns valuation from a prediction into a filter, and a filter is a far more reliable thing to own.

Then attach the risk side separately. Position size from the stop distance and the account, not from conviction, because conviction is exactly the variable a mispricing argument inflates. The position size calculator and the risk management guide keep those two decisions apart, which is the point. The glossary defines the valuation terms used here.