The Corridor Lost Its Floor. The Front End Did It.
What moved
The long end stopped. The front end did not. Over the week the 2yr added 13 basis points to 4.76%, the 10yr added 5 to 5.01%, and the 30yr took off 1 to 5.34%.
The 20yr was unchanged at 5.38%. Everything else on the curve moved higher: the 3-month by 7 basis points to 4.14%, the 5yr by 8 to 4.86%, the 7yr by 6 to 4.93%.
The 5yr and the 10yr now sit at the 100th percentile of their 3yr ranges. The 7yr is at the 99.3rd, the 20yr at the 98.6th, the 30yr at the 97.9th. The 2yr is only at the 83rd, and the 3-month at the 27th.
Away from rates: USD/JPY added 3.60 to 157.06, and gold added $32.60 to $4,391.70.
The composite headline is +0.02, regime NEUTRAL, on 5 of 5 sub-indices. Rates & Policy is the dominant sub-index at +1.16; Liquidity is the drag at -0.86.
The floor, revisited
Last Monday this column put the corridor at 40 to 60 basis points and said that below 25 the question stops being about its width and starts being about its sign. 2s10s is 23.8 basis points.
It did not get there by steepening. It got there by the front end climbing into a long end that refused to follow. The spread compressed 10.7 basis points on the week, from 34.5.
The corridor was never a rule, it was a description. It has been rewritten rather than broken. What was the lower edge is now the middle.
Why this is still not an inversion story
3-month money yields 4.14%. 2yr money yields 4.76%. The market is not pricing cuts across the next 2 years; it is pricing a policy path that stays above cash and charges a premium for the risk the next move is up rather than down.
The funding market is not forcing it. The overnight complex is 3.85% secured against 3.90% administered (as of 21 Sep), a spread of minus 5 basis points. RRP take-up is $0.6bn (as of 18 Sep) and reserves are $3,013.8bn (as of 16 Sep). Nothing at the very front is tight.
Credit is not objecting either. High-yield spreads are 2.70% and investment-grade 0.78% (both as of 17 Sep). VIX is 14.81, in the 7th percentile of its range, and MOVE is 80.64. A market pricing a policy mistake does not look like this.
What the real yield is doing
The 10yr nominal is 5.01% and the 10yr breakeven is 2.33% (as of 18 Sep). That leaves a 10yr real yield of 2.67% (as of 18 Sep). The breakeven is not the variable moving here. The compensation for holding duration is.
That is the transmission to everything downstream. BTC is $81,616, up 7.6% on the week. MSTR is $153.92, up 17.4% over the same window - roughly 2.3 times the move in the asset it holds. Its preferreds are calmer: STRF $104.00, up 2.57 on the week; STRK $75.42, up 1.43; STRC $98.51, up 0.37.
A 2.67% real 10yr is the discount rate on every long-duration claim, and it is being set by the front end while the long end holds. That is the constraint the whole BTC-treasury complex is trading against, and it has not loosened.
What would change the read
2s10s back above 35 basis points with the 2yr falling. That is a bull steepener, and it would mean the front end has stopped leading.
The other break is the long end. If the 30yr clears 5.45% while the 2yr stalls, the spine of this move changes from a policy story to a term-premium story. Those are different trades with different hedges.
Everything between a 23.8 basis point spread and those two lines is noise around a front end that is still doing the work.
| TENOR | LEVEL | 1D Δ | Z (3Y) | RANGE %ILE |
|---|---|---|---|---|
| 2Y | 4.610 | -0.060 | +1.39 | 78 |
| 5Y | 4.780 | -0.070 | +2.24 | 86 |
| 10Y | 4.980 | -0.070 | +2.51 | 84 |
| 30Y | 5.320 | -0.050 | +1.99 | 83 |