What thematic research is for

June 2026

Thematic research has a reputation for producing long, interesting documents that change precisely nothing. The version worth doing ends in a decision and a position size. Everything else is a hobby.

I did this for about seven years at Baillie Gifford, latterly as thematic research lead on a multi-asset book of around £4bn, and as a voting member on the rates and currency sleeve. The work was turning research and data into allocation calls. As a voting member I had to back those calls with real money and then live with them, which concentrates the mind.

A theme is only worth anything when it tells you where the money goes. Three questions. What has to be true for the view to pay? What's it worth if it does, against what it costs if it doesn't? And what's the cleanest way to express it, usually the less obvious instrument? Answer fewer than three and all you've got is a thesis. A position needs all three.

The bit everyone skips is uncertainty. Most forecasts are guesses with decimal places stuck on for confidence. The exact number almost never matters. What matters is simple: how wrong can you be and still come out ahead, and what tells you early that you're wrong?

I do thematic and macro research for institutional clients, and advise on capital allocation.

Markets, macro and multi-asset. me@yussefrobinson.uk