The Banana Man's Desk

In 1933, Samuel Zemurray walked into a boardroom in Boston and took control of United Fruit, the largest agricultural enterprise in the world. He owned enough proxies to fire the entire executive committee. He was an immigrant from Bessarabia who had started by buying overripe bananas off the docks in Mobile, Alabama, fruit the big companies discarded because it would spoil before reaching northern markets. He sold them fast and local. He learned the business from the pier upward.

What Rich Cohen captures in The Fish That Ate the Whale is how Zemurray ran the empire once he had it. The reports came in from Honduras, Guatemala, Colombia, Costa Rica. Sales figures. Yields per hectare. Stem counts. The average length of a banana in centimeters. Market rates by port. A typical Boston executive would have a staff process this, summarize it, present it in a Monday meeting with a deck.

Zemurray scanned. Made mental notes. Moved on.

Life magazine, reporting on him in 1951, observed that he so rarely dictated letters that he didn't need a full-time secretary. He would phone division managers across half a dozen countries, hold their reports in his head, correlate the numbers, and decide. No pencil. No legal pad. No committee.

The point isn't that he had a freakish memory, though he probably did. The point is what he had eliminated. The infrastructure of a normal CEO; the analysts, the briefing books, the staff meetings designed to surface what mattered, Zemurray didn't need any of it because he already knew what mattered. He had cut bananas. He had loaded them onto trains. He had bribed Honduran officials in person. He had walked the plantations and felt the soil. When a manager in Tela called to report yields, Zemurray was not receiving new information so much as updating a model he had been carrying in his body for thirty years.

His rivals at United Fruit ran the company from Boston, dressed in linen, summering on the Cape. They commissioned studies. Zemurray's most famous remark, made when consultants told him a swamp could not be drained for new plantations, was four words: I'm going down there. He chartered a boat that night.

The man did not need a meeting because the meeting was already happening inside him.

The Lens

There is a kind of investor who consumes everything. Earnings transcripts, sell-side notes, screens, dashboards, alerts. The information arrives and is logged, summarized, filed. Conviction is supposed to emerge from completeness. If you read enough, surely you will know enough.

Zemurray's posture suggests the opposite arrangement. The work that makes a decision fast and quiet is done years before the decision. By the time the report lands on your desk, you should already know what it is going to say, roughly, and you should already know which deviations would matter and which are noise. If you cannot scan a quarterly and immediately feel where the surprise is, the problem is not the quarterly. It is that you did not understand the business before you opened it.

This shapes how to think about position sizing. The instinct is to size based on apparent conviction, on how much research has been done, on how thick the file is. But thickness is not the same as fluency. A thick file can be a substitute for understanding rather than evidence of it. The question to sit with is whether you could, like Zemurray, hold the operating picture of this business in your head without notes. Not every detail. The shape of it. The pressure points. What a normal quarter looks like, what a strange one would look like, what the manager on the phone would sound like if something were actually wrong.

Most positions in most portfolios fail that test. Mine included.

The system behind these notes — where I track signals, valuations, and whether I'm actually following my own rules — is at enhaq.capital.