Tech • AI • Robotics
Why knowing the future doesn't always help predict markets | The Economist
5/10TL;DR
Even with perfect foresight of economic news, most investors fail to profit because they misjudge market reactions and size their bets poorly.
KEY POINTS
Experiment with “Tomorrow’s News”
Researchers at Elm Wealth simulated a near-perfect information scenario by showing 120 participants future front pages of the Wall Street Journal across 15 trading days between 2008 and 2022. Participants traded stocks and bonds with up to 50x leverage, mimicking hedge fund conditions. Despite access to what appeared to be a “crystal ball,” results were weak.
Minimal Profits and Frequent Losses
Each participant started with $50, yet the average final balance was just $51. Only about half made any profit at all, while one in six lost everything. The findings suggest that even perfect information about upcoming events does not translate into consistent gains.
Markets React to Expectations, Not Just News
A key challenge is that markets move based on how outcomes compare to prior expectations. Strong economic data, such as job growth, may already be priced in—or even disappoint if forecasts were higher. Participants struggled to account for this dynamic, correctly predicting market direction only about 51% of the time, barely above chance.
Bet Sizing Proved the Biggest Weakness
The most significant failure was not predicting direction but determining how much to bet. Participants did not scale their positions according to confidence. They often placed large, leveraged bets when uncertain and failed to increase exposure when signals were strong. This mismatch led directly to large losses and bankruptcies.
Artificial Intelligence Also Struggled
AI models performed slightly better in predicting direction, with about 60% accuracy, but showed similar weaknesses in risk management. Like humans, they failed to adjust position sizes effectively and sometimes incurred losses, highlighting that the problem extends beyond human psychology.
Professional Traders Outperformed Sharply
In contrast, five expert macro traders tested under the same conditions all generated profits, on average more than doubling their money. Their edge was not dramatically better forecasting but superior calibration of bet size to confidence. They frequently chose not to trade when uncertain and used high leverage selectively when conviction was strong.
Confidence Calibration Is Critical
The study underscores that successful investing depends less on identifying the right asset and more on quantifying conviction. Knowing when to bet small, large, or not at all proved decisive. Poor calibration of confidence leads to overexposure in weak ideas and underinvestment in strong ones.
Implications for Market Behavior
The findings suggest markets can appear irrational while still offering profit opportunities. Participants may profit by anticipating how markets will react—even if those reactions misprice reality—rather than by predicting fundamentals alone.
CONCLUSION
The evidence indicates that the central challenge in investing is not access to information but the ability to size bets appropriately based on confidence, a skill that separates professionals from both amateurs and machines.
Explain this