A macro view is not a trade. This is the process I use to translate a top-down read on the economy and policy into a specific, sized, risk-defined position — and the questions that have to be answered before any capital is committed.
Every position I take begins one or two levels above the instrument itself. Before I think about what to buy or sell, I want a clear read on the prevailing regime: the direction of growth, the path of policy, the behaviour of liquidity, and where the market's consensus is currently positioned. The trade is downstream of that read.
The most useful question at this stage is not "what do I think will happen" but "what is already priced." A view only has value where it diverges from consensus. If the market and I agree, there is little edge — the move has largely happened. The opportunities worth sizing into are the ones where the data, the policy path, or market structure point somewhere the crowd is not yet positioned.
A single macro view can usually be expressed a dozen ways, and most of them are dirty — they carry exposure to things that have nothing to do with the thesis. The discipline is to find the instrument that captures the view with the least unrelated risk.
If the read is on real rates, an outright equity position drags in earnings, sector rotation and idiosyncratic risk that dilute the bet. A rates instrument, a relative-value pair, or an options structure may express the same view far more cleanly. I spend more time on instrument selection than most people expect, because a good thesis in the wrong vehicle is how you can be right and still lose.
Conviction tells you whether to take a trade. It does not tell you how large to make it. Size is a function of one thing: the distance between entry and the level that proves the thesis wrong. I fix the dollar risk per trade first, then let the invalidation distance dictate the position size — never the other way around.
This is where the tools on this site come from. ATR-calibrated position sizing, R-multiple planning, and portfolio-weighted contribution are not academic — they are the mechanics that keep a single idea from being able to do outsized damage. A high-conviction trade with a wide stop gets a smaller position than a moderate-conviction trade with a tight one. That feels counterintuitive until you have watched conviction-sized positions blow holes in a book.
The single most important sentence in any trade I take is the one that describes how I will know I am wrong. It has to be written before the position goes on, because afterwards the mind is very good at moving the goalposts. Invalidation can be a price level, a data print, a policy outcome, or a structural break — but it must be specific and it must be decided in advance.
A trade without a pre-defined invalidation is not a trade, it is exposure with a story attached. The discipline of writing it down first is what separates a process from a series of hopeful guesses.
No position exists in isolation. Before adding anything, I look at what it does to the portfolio as a whole: does it concentrate an existing exposure, hedge it, or introduce something genuinely new? Three "different" trades that all depend on falling real yields are one trade in three costumes, and the book is far more fragile than it looks.
Managing correlation across the book is, over time, more important than the merit of any individual idea. The goal is a set of positions whose risks are as independent as possible, so that being wrong on one does not mean being wrong on all of them at once.
This is an evergreen methodology note, not financial advice or a specific market call. Dated trade theses and live positioning appear on the Trade Log.
The Desk Note — periodic macro reads and the thinking behind live positioning. Free, no spam.