Lesson 7 of 7
Second-level thinking
First-level thinking asks if a company is good. Second-level thinking asks what its price already assumes.
By Javier Sanz
Founder of ValueScreener. Built Ninety Nine, a securities services provider, and sold it to a tier-one global bank; ran operations and internationalization at Alpaca, the brokerage API.
Who writes thisEditorial processLinkedIn (opens in a new tab)X (opens in a new tab)
In one minute
- First-level thinking asks whether a company is good. Second-level thinking asks what its price already assumes.
- A price can be solved backward for the growth rate it requires, which turns a feeling about the price into a checkable number.
- Knowing that a number is priced in says nothing about when, or whether, the market will change its mind.
Second-level thinking asks what a price already assumes, not only whether the business is good. Every purchase requires somebody selling at that price, and many sellers have read the same filings, often for longer. A reason to buy that does not account for the seller's reason is only half an argument.
What the price already assumes
- Value with no growth, after-tax profit ÷ cost of capital500
- Part of the price that is a claim about growth+500
- Enterprise value the current quote implies= 1,000
Two questions that sound the same
First-level thinking
- Asks whether the company is good.
- Stops once the answer is yes.
Second-level thinking
- Asks what the price already assumes about the company.
- Checks whether that assumption is reasonable before buying.
The second question can be answered from the price, the filed profit and a discount rate, before you decide whether the company is good.
Marks's second-level thinking asks what the price already assumes, and whether the consensus is already paid for. A better company can be the worse purchase when its price assumes more than it can deliver.
Solve the price backward
Mauboussin and Rappaport's approach is to stop forecasting and comparing the forecast with the price, and instead solve the model backward for the performance the current price already requires. The difference is procedural, not mathematical. A forecast built first tends to land near the price, because the analyst already knows the price. A number extracted from the price cannot flatter anyone: it was never the analyst's own guess to begin with.
Normalised means an ordinary year, with booms and busts averaged out. Cost of capital is the yearly return investors demand for the risk. The Waterfall above splits the price of the whole business, its enterprise value (its shares at market plus its net debt), implied at 1,000, into two parts: 500 that the business is worth with no growth at all, at an after-tax profit of 50 and a cost of capital of 10 percent, and another 500 that is purely a claim about growth still to come.
The answer moves with the rate: at 9 percent the same price assumes 4 percent growth, and at 11 percent it assumes 6 percent.
Useful for standards, not for timing
Knowing that a price already assumes 5 percent growth forever says nothing about when, or whether, the market will change its mind. The question is useful for setting how demanding your standard should be before you buy, and useless for calling when a price will move. Treating it as a timing signal is how the idea does damage: acting early on a price that looks demanding and acting early on one that looks cheap are the same mistake, made twice.
Check yourself
First-level thinking asks if a company is good. What does second-level thinking add?
- A longer holding period
- Whether the price already assumes that
- A higher discount rate
Show the answer
Answer: B.
Second-level thinking checks the assumption baked into the price, which a good company can still fail to clear.An enterprise value of 1,000 is implied by the quote. A no-growth value of 500 comes from profit divided by the cost of capital. What is the other 500?
- Cash already sitting on the balance sheet
- A claim about growth, not yet delivered
- A pricing error the market will soon correct
Show the answer
Answer: B.
The gap between the no-growth value and the quoted value is the market's claim about future growth, and it converts into a rate.A price assumes 5 percent growth forever. What does that tell you about when the price will fall?
- Soon, because 5 percent is demanding
- Nothing. It sets a standard, not a date
- Never, because the market has already priced it
Show the answer
Answer: B.
The number is a standard the price must clear, not a forecast of when sentiment turns.Sources
- Marks, The Most Important Thing, 2011, on second-level thinking.
- Mauboussin and Rappaport, Expectations Investing, 2001, on solving a price backward for the expectations already inside it.
- Shiller, Irrational Exuberance, 2000, on the cyclically adjusted price to earnings ratio.
What this leaves out
- The arithmetic uses invented round inputs on a perpetuity with no reinvestment (none of the profit is put back into the business), the simplest structure available, and understates the growth a real price implies once reinvestment is added.
- This lesson reports no measure of where any market or company sits today, and names no security.
- A ten-year normalisation here uses the entire fundamentals archive on file, with no earlier cycle to check it against.
- Nothing here is advice about any security or a signal about timing.