Built and run by one person.

Why does my annualised rate not match the figure the regulator published?

Because the two of you annualised the same filings on different bases. Both numbers are right, they answer slightly different questions, and nothing in either figure tells you which basis produced it. This is the commonest reason a careful calculation lands a few basis points away from an official one.

Three bases are in common use:

A measured example. US credit union charge-offs for Q2 2026: the NCUA publishes 0.78 per cent, which is the year so far annualised. The same filings, taking the quarter on its own, give 0.76 per cent. Neither is an error. For reading a turn the quarter basis is the more useful of the two, and it carries one honest caveat: annualising a single quarter assumes the rest of the year looks like it, which a seasonal book does not.

Then check the denominator, because it moves the number too. A rate needs a balance to divide by, and the average balance over the period and the balance on the last day of it are different denominators. On a book that is growing, dividing by the period-end balance makes the rate look lower.

What to do about it. Match the basis before comparing two sources, and state the basis wherever you publish a rate. When a figure differs from an official one and you cannot say why, the gap is almost always the basis or the denominator rather than the data.

The credit union series carries all three bases, at overall and at segment level: https://www.tigzig.com/tremor (US Credit, then Credit Unions). Worked through with the Q2 2026 numbers: https://www.tigzig.com/post/credit-union-q2-2026-losses-delinquency-sep2026. Related: are US credit unions in trouble.

Building something like this? How I work covers the rates, the availability and what I take on.

← All Agents FAQ