Summary of Key Points
This article uses Keynes' "beauty contest" theory to explain the phenomenon in financial markets where asset prices fluctuate dramatically with market trends: prices are not determined by the true value of the assets themselves, but by higher-order expectations—i.e., how people predict what others will think. The article verifies this logic through an experiment designed to estimate the average value and then extends the discussion to the "trend narratives" in the fields of tech finance and venture capital. It also explores whether artificial intelligence (AI) will make markets more rational, concluding that it may not necessarily do so; instead, it could change the structure of these expectations. The article suggests that an active government should understand the mechanisms behind price formation and intervene in a timely manner to prevent systemic risks.
Detailed Analysis
#### 1. Keynes' Beauty Contest: Prices as a Game of "Guessing What Others Think"
Keynes compared financial markets to a beauty contest where the winner is not determined by who thinks the face is the most beautiful, but by whose judgment others will agree with. For example, when buying stocks, investors do not focus on the fundamental quality of the company but on whether others will buy them and what they think others will do. This "expectation of expectations" is at the core of price fluctuations: even if the company remains unchanged, prices rise as long as everyone believes others will continue to buy; once doubts arise about future sales, prices fall. This explains why, under the right market conditions, even mediocre assets can seem highly valuable.
#### 2. The Experiment on Estimating the Average Value
The article demonstrates this logic with an experiment where participants were asked to choose a number between 0 and 100 that was closest to two-thirds of the average value. Logically, everyone should have chosen 0 (since the equilibrium would be 0 after multiple rounds of selection), but the actual winning numbers were around 12-13, indicating that people's thinking varied by cognitive level. Some only considered the simple average (50 and two-thirds equals 33), while others anticipated what others would choose (e.g., 22). The winners were not the "rationalists" who chose 0 but those who guessed the majority's opinion, showing that prices are the result of interactions among different levels of understanding.
#### 3. The Power of "Trend Narratives" in Tech Finance
Tech companies (such as those in AI or renewable energy) do not have stable cash flows, so their valuations rely entirely on future prospects. Investors care less about whether the technology is viable than whether others will continue to invest and whether they can secure funding for the next round of financing. For instance, a surge in an AI sector is often driven by market consensus rather than sudden technological breakthroughs; once doubts arise, valuations plummet. This highlights how "narratives" (i.e., the stories companies tell about their potential) shape investor expectations and drive capital flows.
#### 4. Will AI Make Markets More Rational?
Many believe that AI can eliminate irrationality due to its fast processing speed and lack of emotional biases. However, the article argues that AI may complicate these expectations:
- Some investors will follow AI-driven trends, while others will act against them (e.g., through contrarian strategies).
- AI itself will also try to predict how other systems will respond to its actions.
For example, if AI knows that everyone is using it to estimate a certain value, it might choose a higher number to anticipate different expectations. As a result, the introduction of AI may not stabilize prices but instead make them more volatile, as investors focus more on what others think about AI.
#### 5. The Role of an Active Government
The article emphasizes that governments should not try to replace market pricing mechanisms but distinguish between two scenarios:
- If prices are driven by fundamental factors (e.g., a company's profitability), there is no need for intervention.
- If prices are influenced by "expectations of expectations" (e.g., speculative bubbles), the government should intervene appropriately (e.g., through regulatory measures).
The 2015 stock market bubble was a result of converging expectations, and government intervention was necessary to prevent systemic risks. Such regulation is not about controlling the market but about acting rationally based on an understanding of price formation.
Conclusion
Financial markets are not mere tools for rational calculation but arenas where expectations interact. Price fluctuations are driven by what people think others will do, and technologies like AI can only add complexity to these dynamics. The role of governments is to identify when expectations become too convergent and take action to prevent risks, rather than blindly assuming that prices always reflect reality. Understanding this logic helps explain why some stocks with average fundamentals soar or why market trends suddenly collapse—ultimately, it's all about how people's expectations change.