Setups Don’t Fail. Regimes Do.

The Same Setup, Two Markets

Take a sweep-and-reclaim at a level you have marked for a week. Price runs the low, aggressive selling hits into a bid-heavy book, the level holds and the book flips. On a day when order-flow imbalance is firing across half the crypto universe and funding is flat, that setup runs. On a day when funding is stretched across the venue and the equity books have thinned to a third of their usual depth, the same setup gets sold into within three candles. Nothing about the pattern changed. The market it happened in did. Traders who keep a journal know this feeling well: the losers were not bad setups, they were good setups in the wrong market.

Why Nobody Shows You This

Trading tools split into two families. One goes deep on a single market — a liquidity heatmap, a footprint chart, a DOM — and is blind to the other three hundred markets on the venue. The other lists a hundred metrics for every market and leaves you to infer a state from a table. Neither draws the picture that decides the trade: what the whole venue is doing, right now, on the same clock as your setup. Having looked hard for anyone teaching that picture, we found footprint tutorials on one side and academic papers on the other, and almost nothing in between. So this is us drawing it.

What A Regime Is, In Order-Flow Terms

Strip the word of its mystique. A regime is one metric that every market already carries, reduced across a slice of the venue into a single number per candle. Three choices define it completely.

• The metric: funding, open-interest change, spread, order-flow imbalance, CVD, or any formula you write.
• The reduction: the median, the mean, or the share of markets meeting a condition (breadth).
• The slice: the whole venue, an asset class, a sector or theme, a watchlist, what is on screen, or only the markets where your signals are firing.

Examples That Earn A Row

Each row is a measured number you can read against its own recent range. None of them is a label. None of them is a prediction. They are the answer to “what kind of market is this?”, updated every candle.

• Median funding, all markets — is leverage paying to be long or short across the venue?
• Share of crypto with OFI firing — is order-flow pressure broad, or three names carrying the board?
• Median spread, tokenized equities — have the equity books stepped back from the touch?
• Breadth of open-interest surges, memes — is new money entering the most speculative corner first?

Put It On The Same Clock

The mistake most context tools make is putting the market overview on a different page from the setup. The overview becomes something you glance at in the morning and forget by the third trade. A regime belongs above the board, painted on the same timeline as the setups beneath it, so the question in front of you changes from “is this setup clean?” to “does the market agree with it?”. That is what an A+ setup means here: a moment where flow, positioning and regime agree — and the band is the only place the third part is visible.

Two Rows, One Venue

Scope one row to memes and another to L1s. Scope one to crypto and another to tokenized equities. Now the band is a comparison: which cohort is leading, which is being sold into, and the moment they diverge. A rally where the L1 row is climbing on rising open interest while the meme row climbs on falling open interest is two different markets wearing the same green candles. One is positioning; the other is a short squeeze. The chart does not know the difference. The band does.

Give The Same Rows To Your Analyst And Your AI

A regime read as a badge — “risk-on” — teaches an AI nothing. A regime read as a series does. The desk analyst reads every row over its last 100 bars before it decides whether a setup on your desk deserves a card, and over the MCP connector your own ChatGPT or Claude gets the identical series. An AI told “median funding is +3 basis points and has risen for forty bars while equity spreads doubled” reasons about a squeeze and thin books; an AI told “bullish” guesses. Context is what turns a language model into a co-trader instead of a headline reader.

The Habit

Before the setup, read the band. Most days it will tell you the market does not agree with what you want to do, and the honest response is to wait. A tool that says “not now” more often than “go” is not a weak tool; it is the one that keeps the journal from filling with good setups in the wrong market.

Related reading

Keep going on context and flow: