JPMorgan's $2 Billion Loss Exposes the Process Illusion

Good process does not guarantee good outcomes, so you should never judge a decision by whether it worked.

This distinction has spawned an entire cottage industry of decision-making frameworks—books, courses, corporate training modules—all built on the assumption that process and outcome are separable variables. You can dominate a soccer match and lose on a handball call in injury time, or play badly and win because the other team's striker misses a penalty.

The argument takes for granted but never examines something crucial. It assumes that there exists a way to evaluate process in the absence of outcomes. You cannot know whether your reasoning was sound without any external check on it.

In 2008, JPMorgan Chase's London office constructed a $20 billion synthetic credit portfolio—a carefully reasoned position built on what the traders insisted was rigorous modeling and sound process. The position imploded inside four months, losing $2 billion, and an external investigation later called the trading process "reckless." Yet the team had followed their own documented procedures and simply bet wrong. JPMorgan's risk committee couldn't tell whether the process was unsound or the outcome was simply unlucky until after the loss materialized.

A physician performs a statistically justified but high-risk surgery and the patient dies. The licensing board must decide whether to discipline the surgeon. The board cannot evaluate the physician's process without reference to the outcome, because in medicine—as in trading, as in any domain with genuine uncertainty—the only available evidence of process quality eventually comes from repeated outcomes over time. The process-versus-outcome split assumes you can peek at the process in isolation, but you cannot. What you can do is ask whether you'd make the same decision ten times over, knowing you'll win some and lose some.