Design a financial stress index for five economies, Aldora, Brevia, Corvin, Dessala and Estmar, from credit growth, the current account balance and short-term external debt to reserves. How do you normalise, weight and rank, and what are the index's weaknesses?

Convert each indicator to a z-score, flip the current account so deficits count as stress, and average: with equal weights Dessala ranks most stressed at +0.89, just ahead of Corvin at +0.72. The ranking is a judgement, not a fact. Put half the weight on short-term debt, the most direct measure of a sudden-stop risk, and Corvin moves to the top at +0.97. The index must disclose its weights, and two of its inputs largely measure the same credit boom.

Measure each country against the group, in units of the group's spread. A class ranks pupils across maths, which is marked out of 100, and an essay marked out of 20, by asking how far above the class average each pupil sits in each subject, not by adding the raw marks. A z-scoreHow many standard deviations a value sits above or below the average of the group. Zero is average; +1 is one spread above it. does the same: subtract the five-country average and divide by the spread. Credit growth averages 16% with a spread of 9.1, so Dessala's 30% scores +1.53. For the current account, multiply by minus one first, because a deficit is the stress; Dessala's -6% then scores +1.45.

Start from what the index is for. The PM asked about a sudden stop in foreign funding, and short-term external debt against reserves is the most direct measure of that: it asks whether a country could pay what falls due in a year if foreign lenders refused to roll it over. Credit growth and the current account describe how the pressure builds; the debt ratio describes whether it breaks. Equal weights give Dessala +0.89, Corvin +0.72 and Aldora +0.45. Put half the weight on the debt ratio and a quarter on each of the others, and Corvin leads at +0.97, well clear of Dessala at +0.59: Dessala's stress is a credit boom financed so far with long-term money, Corvin's is debt that falls due within the year.

Name them before the PM does. First, credit growth and the current account move together here, with a correlation of 0.94 across the five, so equal weights count the same credit boom twice and underweight the one indicator that measures funding risk directly. Second, z-scores against a group of five are unstable: add a sixth country with a 200% debt ratio and every score moves. Scoring each country against its own history is a sensible second view. Third, the inputs are annual and published with a lag, while a sudden stop happens in weeks, so pair the index with market prices such as currency moves and bond spreads.

Close with how you would use it. An index like this does not say when a crisis comes; it says where one would hurt most. A macro fund would use it to decide which currencies to be short against which to be long when global funding tightens, and would test it by asking whether it ranked past stressed countries highly before their trouble, not after.