Misread How Wall Street Makes Money, and Your Own Strategy Breaks Too
Investing Mindset, Part 3: high-speed trading and long-term investing are not the same business, and mixing up their numbers leads to bad comparisons.
"Wall Street doesn't call the direction right, it just collects fees." It's a catchy line, but taking only half of it is dangerous. What matters isn't just what a firm buys and sells, but what service it provides and what risk it carries. Only then can you tell whether its results are even comparable to your own investing.
Same Name, Different Source of Profit
Citadel Securities states plainly, in its own official description, that it is a separate financial company from Citadel, the hedge fund. The former is a market maker, acting as the counterparty to both buyers and sellers. If you take the fund returns from the latter, the asset manager, and drop them into a comparison chart claiming "individuals can earn 20% a year through market making too," you're mixing numbers from two different businesses. source
| Participant | Reward Sought | Risk Taken On |
|---|---|---|
| Market maker | Bid-ask spread, among other sources | Inventory price risk, information disadvantage, system risk |
| Directional trader | An edge in predicting price moves | Failed predictions, execution costs |
| Diversified long-term investor | Compensation for corporate earnings and risk-bearing | Market downturns, prolonged underperformance |
A market maker quotes both sides of the market to facilitate trading. Profit comes from connecting buyers and sellers at a favorable spread, but the inventory it buys first can also drop in price. Citadel Securities' own explanation says as much: the spread is compensation for bearing market risk. It is not automatic, risk-free profit. source
Even Algorithms Have Accidents
According to the SEC, Knight Capital's order system malfunctioned in 2012, building an unwanted, massive position over 45 minutes and ultimately costing the firm more than $460 million. That was not a Citadel incident. It's a counterexample showing that high-speed trading is a business with operational risk of its own. source
Statistical arbitrage often relies on the assumption that the price relationship between two assets will revert to what's expected. If that relationship widens further, or funding dries up, losses can grow. The word "statistical" is not a certificate that the risk has been removed.
The Claim That All Charts Are Useless Also Goes Too Far
Researchers at AQR examined strategies that use historical price trends across multiple assets as trading rules, drawing on long-run historical data. That makes it hard to dismiss every price-based strategy as pure superstition. Still, this is a historical simulation study from an asset manager, with its own constraints around cost estimates and model choices. It does not validate an individual's arbitrary trendlines or AI-recommended trades. source
There is evidence on the other side too. A study of Taiwanese day traders from 1992 to 2006 found that fewer than 1% of traders achieved positive abnormal returns, after costs, that were both predictable and reliable. That is not simply a win rate, and not a "three-year survival rate." Skill differences did exist, but a persistent edge was hard to find. source
Three Checks Before You Borrow a Strategy
- What is the actual source of the return?
- What risk is being compensated for?
- Does it hold up under your own conditions?
These are the comparison points to check before a company's name or its past returns.</markdown>
