Skip to content

Research

Macro surprise

Point-in-time macro releases traded as cross-sectional spreads, and the vintage discipline that makes the surprise real rather than hindsight.

A macro surprise is the gap between a release and its expectation, traded as a cross-sectional spread rather than a directional bet. Different companies have different exposure to the same macro variable, so a mild aggregate inflation event can create a larger relative move without requiring a view on the level.

Everything here turns on vintages. Macro series are revised, sometimes substantially, and the revised value is the one every convenient data source serves. A study that uses it is trading on a number that did not exist on the day of the trade, and the resulting effect can be entirely an artifact of the revision process. The point-in-time vintage lake this research runs on exists for that single reason, and the rule used to date each series is published with it.

This is the mechanism behind one of the book's live sleeves, a point-in-time consumer-price-surprise size spread, which makes the cluster small but load-bearing. It is also where one of this book's published corrections lives: a missing release was found and the affected figures were withdrawn and restated in public rather than quietly re-run.

That correction is the most useful document here. Every research process makes errors. The relevant test is whether flattering errors are found and published as readily as unflattering ones. The missing-release note gives that question a concrete answer rather than an assurance.

The macro surprise documents