The readings a macro desk actually prices off — nineteen of them, raw and derived — each scored by how close it sits to the level that has historically mattered, plus the multi-reading signatures no single row can show.
Three patterns, each answering a different question and needing two readings at once. Both conditions are shown whether or not the pattern is armed.
Each era's worst reading scored the same way as today's, averaged across every row with data in that window. 2020 is left out on purpose: COVID came from outside the financial system, so it says little about how economies break.
One reading per sector — the single number that sector turns on. Scored the same way as the board above, but kept out of the overall score: this is where the stress is landing, not how close the system is to breaking.
This is not investment advice. Thresholds are judgement calls, several inputs are quarterly and lag by up to three months, and no set of indicators has ever called a market turn reliably.