Second-Level Thinking
To outperform, you must think differently AND better than the consensus
Howard Marks distinguishes between first-level and second-level thinking as the fundamental separator between average and exceptional decision-making. First-level thinking is simplistic and surface-level: 'This is a good company. Let us buy the stock.' Second-level thinking is deeper and more contrarian: 'This is a good company, but everyone thinks it is a great company, and it is not. So it is overrated and overpriced. Let us sell.' The critical nuance is that outperforming requires not just being different from the consensus but being different AND correct. Being different without being better produces poor results. Being better without being different produces average results — because if you agree with everyone else, the consensus is already reflected in the price. Only when you are both different and correct do you outperform. This framework extends far beyond investing. In any competitive domain — business strategy, product development, career planning — consensus thinking produces consensus results. To achieve exceptional outcomes, you must develop independent views based on superior analysis or information, and you must have the courage to act on those views even when they contradict what everyone else believes. Marks emphasizes that second-level thinking is inherently uncomfortable because it requires standing apart from the crowd, which triggers deep social instincts to conform.
- First-level thinking says: this is a good company, let us buy. Second-level thinking says: everyone thinks it is great, so it is overpriced, let us sell.
- To outperform, you must think differently AND better than the consensus.
- Being different without being better produces poor results. Being better without being different produces average results.
- In competitive domains, consensus thinking produces consensus results by definition.
- Identify the Consensus ViewBefore forming your own opinion on any important decision, deliberately map out what the consensus view is. What does the majority believe? What is priced into the market, the job offer, or the competitive landscape? The consensus view is the default expectation — it is what everyone already agrees on and what is already reflected in current conditions. You need to understand it clearly before you can meaningfully disagree with it. In investing, the consensus is reflected in the current price. In business strategy, it is reflected in what your competitors are all doing. In career planning, it is reflected in the conventional career path in your industry.Pro tipRead widely and listen to expert predictions not to follow them, but to map the consensus. The consensus view is your baseline — outperformance is measured against it.WarningDo not dismiss the consensus reflexively. The consensus is often correct. Contrarianism for its own sake is just as dangerous as conformism.
- Develop an Independent View Based on Superior AnalysisOnce you understand the consensus, develop your own view through deeper analysis, better data, or a different analytical framework. Your independent view must be based on something the consensus is missing — not just a contrarian hunch. Ask: what do I see that others do not? What data am I weighing differently? What second-order effects has the consensus failed to consider? If you cannot articulate a specific reason why the consensus is wrong, you do not have a second-level view — you just have a different opinion, which is not the same thing.Pro tipThe best second-level views come from understanding the psychology driving the consensus. When the consensus is driven by fear (everyone is selling), the asset may be undervalued. When driven by greed (everyone is buying), it may be overvalued.WarningHaving a different view is easy. Having a different view that is also correct is extremely difficult. Most contrarian bets fail. Second-level thinking is about increasing your hit rate, not guaranteeing it.
- Act on Your Independent View with Appropriate SizingWhen you have genuine conviction in a second-level view — one based on specific evidence or analysis that the consensus is missing — act on it. But size your position appropriately to the confidence level. High conviction based on strong evidence warrants larger positions. Moderate conviction warrants smaller, exploratory positions. The key is that you must act on your independent thinking, because second-level thinking that never translates into action produces zero results. At the same time, you must manage the risk that you are wrong, because being different and wrong is the worst outcome.Pro tipMarks recommends asking: 'What is the range of possible outcomes, and how am I positioned across that range?' This prevents you from betting everything on a single scenario.WarningNever bet so heavily on a contrarian view that being wrong would be catastrophic. Second-level thinking improves your odds over many decisions but does not guarantee any individual one.
In 1978, Howard Marks founded the first high-yield bond fund at Citibank. The consensus view was that high-yield bonds (then called 'junk bonds') were too risky to hold. First-level thinking said: these are bad companies, stay away. Marks' second-level analysis showed that while the companies were indeed weaker, the bonds were priced at such steep discounts that the expected return (accounting for default probability) was actually superior to investment-grade bonds. He was different (he bought what everyone else avoided) and he was correct (the returns proved exceptional).
Howard Marks developed the concept of second-level thinking through over 50 years of investing experience, starting at Citibank in 1968 and later co-founding Oaktree Capital Management, which grew to manage over $120 billion in assets. He codified the concept in his landmark book The Most Important Thing (2011) and has elaborated on it in his famous investor memos, which he has written since 1990. The concept emerged from Marks' observation that in public markets, the consensus view is already reflected in prices. Therefore, agreeing with the consensus cannot produce above-average returns — you simply get the return that the consensus expects. Only when you disagree with the consensus AND turn out to be right do you earn excess returns. This insight applies to his specialty, high-yield bonds, where he discovered early in his career that 'the worst companies in America' trading at steep discounts often produced superior returns precisely because everyone else had written them off.