Skip to main content Skip to main content
Not investment advice. Educational reading. See Disclaimer.
L.25 · BEGINNER · 4 MIN

Second-Level Thinking: Reading the Consensus

Howard Marks named the discipline of second-level thinking to capture the central analytical move that separates investors who consistently add value from investors who merely react to the same information everyone else has. The framework is simple to describe and unusually hard to apply, because second-level thinking requires holding two views simultaneously — your own view of the future and your honest assessment of the view the rest of the market has already priced in — and then placing capital only when the two views meaningfully diverge.

Quiz · 5 questions ↓
§ 01

First-level thinking asks: what will happen? Second-level thinking asks: what will happen relative to what is already priced in? The investor's return is determined by the gap between actual outcomes and consensus expectations, not by the level of the outcomes themselves.

§ 02
Question being askedFirst-level thinkingSecond-level thinking
Earnings outlookEarnings will be strong next yearEarnings will be stronger than the strong outcome already in the price
Industry assessmentThis is a great industry to be inThis is a great industry that the market has already noticed, so durable excess returns require a thesis the market is missing
Macro viewRates are likely to fallRates are likely to fall less than the steep cuts already implied by the forward curve
Catalyst expectationThe catalyst will happenThe probability the catalyst happens is higher than the probability the market has assigned
Bear-case assessmentThere is a real risk of a recessionThe probability and depth of recession in the price is materially different from my honest probability and depth estimate
§ 03

Variant perception is the active ingredient. Michael Steinhardt's framing was that the entire purpose of analytical work is to develop a view of the future that differs in some specific, defensible way from the view the consensus has already adopted. Without variant perception, the position has no edge — even if the underlying thesis is correct, the return has already been earned by the consensus that arrived at the conclusion before you did.

§ 04

Reading the consensus is a separate analytical skill from forming your own view. The two have to be developed in parallel, because consensus is not a single number — it is a distribution of expectations across investors, analysts, and media, often anchored to recent results and recent commentary. Signs the consensus has moved bullish on a name include compressed short interest, rising sell-side price targets, expanding multiples, broadly favorable media coverage, and visible position accumulation by well-known long-only funds. Signs the consensus has moved bearish include the inverse. None of these signals is decisive on its own, but the constellation gives a usable estimate of what is already in the price.

§ 05
Expected Excess Return ≈ Your View Of The Outcome − Consensus View Of The Outcome
§ 06

Contrarian discipline is not the same as contrarianism for its own sake. Second-level thinking sometimes leads to positioning against consensus; it sometimes leads to agreeing with consensus on the direction while disagreeing on the magnitude; it sometimes leads to declining to take a position because the gap between your view and the consensus view is too small to justify capital. Disagreeing with consensus simply because consensus exists is its own form of first-level thinking — it substitutes a reflex for analysis.

§ 07
Pick one position you currently hold and write down, in one paragraph, the consensus view that is currently embedded in the price — what next year's earnings, growth rate, and business condition the market appears to be expecting. Then write down your own view of the same variables, in a second paragraph. The honest analytical question is not whether your view is correct but whether the two paragraphs differ enough to matter. If they say roughly the same thing, the position has no variant perception and should be sized accordingly.
§ 08
Investor research uncovers that a mid-cap industrial company's free cash flow conversion is likely to improve materially over the next two years because a multi-year capital investment cycle is winding down. On further investigation, three sell-side analysts have already published notes raising the same point in detail, the stock has rallied 28 percent over the last three months, and the consensus 2026 free cash flow estimate is roughly 18 percent above the trailing-twelve-months number — a level that is broadly consistent with the investor's own estimate. What is the most honest interpretation under a second-level-thinking discipline?
§ 09

Second-level thinking is not a personality trait or a contrarian pose. It is a discipline of holding two views in mind at once — yours and the market's — and acting only when the two diverge enough to matter. The hard part is the honesty required to read the consensus accurately, because the alternative is repeatedly being surprised that a correct thesis produced no return.

Five questions · AI feedback

Sit with the ideas.

An investor concludes that a well-followed mega-cap retailer will report strong holiday-quarter results because foot traffic data the investor has been tracking confirms a robust season. The stock has been rising sharply into the print, sell-side analysts have upgraded earnings estimates three times in the last six weeks, and major financial media outlets are running coverage about the company's holiday strength. What is the second-level-thinking question the investor needs to ask before sizing the position into the earnings release?

Why:
See it on a real ticker →