An edge is a durable reason to expect better than the average. All four are real and all four have paid. Three are races, and every race ends when enough participants catch up.
How an edge ends, soonest first
- Informational, knowing sooner.
- Gone on disclosure.
- Analytical, modelling better.
- Copied once published, or reverse-engineered from the trades.
- Executional, being faster.
- An arms race, and the lead is bought again every year.
- Structural, the shape of the payoff.
- Nothing to leak. Survives being known.
Structural edge is no secret, so there is nothing to find out. It is convexity: downside bounded, upside open, so being wrong costs a little and being right pays many times over. Publish it in full and it still pays, because the barrier was never knowledge.
Six routes to the shape
A structural barrier still has to be crossed, and every route charges. The question is not whether the shape is known but which price you pay, and five of the six keep charging.
- Bought, long options and tail hedges.
- Premium, every calm year.
- Borrowed, leverage.
- Symmetric rather than convex. It magnifies both tails.
- Lucked into, a record that happened to bend.
- Unrepeatable. Survivorship worn as skill.
- Free, the index.
- No floor. Full drawdowns eaten whole.
- Rented, a discretionary expert.
- Drifts, charges richly, and dies with him.
- Built, encoded in the rules.
- Hard to build. Paid once.
FX212 is the sixth. The first five hold a payoff for as long as they keep paying for it. The sixth builds the cause, and the payoff follows.
What was built
We encoded the veteran’s decision logic rather than its output: a mechanism, not a curve fit. The rules act on conditions rather than forecasts, leave winners to run, and withdraw as conditions degrade, so losses stay small and a few large outcomes carry the record.
The shape is never the objective. It is what running the rules leaves behind, which is why it is hard to copy from outside. Read the payoff and you have the effect. Only the rules cause it.