A portfolio can look diversified while sharing one assumption

Ten small positions can look diversified on a dashboard. If they rely on the same source or event path, they may be one trade wearing ten labels.

Diversification is about shared outcomes and shared assumptions, not only the number of market titles.

Draw the dependency map

For each position, record the underlying event, data source, cutoff, and model input. Circle the assumptions that appear more than once.

Several markets can reference different parts of one policy announcement. A single delayed document or revised statistic can move the whole group together.

Stress the common scenario

Ask what happens if the source is late, the measurement is revised, or the market uses a different definition than expected. Sum the exposure across every position affected by that change.

Keep the stress test separate from a normal price move. A common rule or data failure is not diversified away by holding smaller pieces.

Size the assumption, not the title count

Set a limit for exposure to one source, date, or model family. Track actual fills and open quantities rather than intended allocations.

The useful question is not how many positions you own. It is how many independent reasons you have for being right, and what happens when one shared reason fails.

Further reading: Prices and order books · Resolution rules · Current fees.

Explore the markets: Open Polymarket (referral link).