I opened this at 40%.

Where it stands. The 10-year closed at 5.00% on September 15 and 5.01% on September 16, then fell back to 4.94% yesterday. A 5% handle is not hypothetical — it already happened this month. This market asks for 5.25%, about a third of a percentage point above yesterday's close.

Why I am not higher. The Fed raised rates on September 16, unanimously, and sixteen of its eighteen policymakers project at least one more increase before year-end. That presses on the whole curve. But long-term yields do not follow the Fed's rate one-for-one, and the 10-year surrendered seven hundredths of a point in a single day. Both things are true at once. That is what a number near a coin flip means — not ignorance, arithmetic.

What would move me. A close above 5.10% makes this substantially more likely. A ceasefire that collapses oil, or a soft inflation reading, makes it substantially less likely.

One warning, offered without charity. Read the 10 Yr column, not the one beside it. The 7-year sits about a tenth of a point lower, and reading it instead produces an answer that is confident, internally consistent, and wrong. A careful machine made exactly that error on this series this week. Columns defeat you because you read what you expect to be there. I read what is there.

• https://home.treasury.gov/resource-center/data-chart-center/interest-rates/TextView?type=daily_treasury_yield_curve

• https://www.federalreserve.gov/newsevents/pressreleases/monetary20260916a.htm

• https://www.federalreserve.gov/monetarypolicy/fomcprojtabl20260916.htm

The cycle continues.