The Most Important Thing — Marks: COMPLETE

Marks built Oaktree Capital on the memo culture: letters to clients in which he thought aloud about risk, cycles and psychology. The Most Important Thing is the distillate of twenty years of memos, and the book's form is a loop — each chapter answers »the most important thing is X« and then shows why X is not enough without Y.

The first X is the difference between first-level and second-level thinking.

The first level is simple and intuitive: »The company has good products and fine growth — buy!« The second level is, as Marks writes, »deep, complex and convoluted« — deep, complex and intricate. The equivalent reads: »The company is good, but everyone knows it — the stock trades at a P/E of 45 that prices in ten years of perfect execution.

It may be a sell.« The second level is not about finding better companies, but about understanding what the price already expects.

The logic is economic, not philosophical. The market portfolio is the sum of everyone's opinions and positions; whoever thinks like the market gets the market's return — before costs.

Marks puts it plainly: »You cannot do the same things others do and expect better results.« Outperformance must be bought with divergence.

But divergence in itself is not enough. Non-consensus must also be right — that is the brutal half that gets forgotten. Being different and wrong is just a more expensive way of losing.

The second-level thinker's question is therefore always double: how does my picture of the future differ from the market's, and why am I reasonably right in precisely that difference?

In its strong form, the efficient market hypothesis says that prices always reflect all information and that no one can beat the market. Marks accepts half of the claim: prices are hard to beat, because thousands of hungry and competent players chase the same advantages and information spreads instantly.

But he rejects the other half: hard to beat is not the same as always right. »Efficient« does not mean that prices are correct — only that mispricings are hard to find.

Why do mispricings arise then? Because prices are set by humans, not machines. When greed and fear crowd out analysis, people buy too much and sell too cheaply. A mispriced asset requires that someone makes a mistake — and the mistakes are made for emotional reasons, not for lack of information.

Behavior is therefore the most promising hunting ground.

Marks counts three sources of edge. Informational edge: knowing something others do not — rarely realistic and often illegal. Analytical edge: calculating or interpreting better — possible but demanding. Psychological edge: bearing emotions better than the market — patience, discipline, resistance to the group.

The third is the only one open to most investors, and it is trained not in Excel but in self-knowledge.

Practical consequence: look where few look. Large, hot and well-covered assets are almost always the most efficiently priced. Small companies, convoluted capital structures, out-of-favor sectors and incomprehensible situations are sparsely covered — there, second-level thinking can actually pay.

In the AKM1 ecosystem, this is where the work on, for instance, source criticism (V12, reporting quality) in abandoned names finds its rationale: you search where no one else bothers.