Journal ·
Friday, 28 August 2026
Regime NeutralMarket regime
High yield tightened 7bps on the week to 2.63%, the largest single move anywhere among the readings we track and the one that carries the most information: credit is the channel that typically deteriorates ahead of equity, and it moved the other way. The 10Y matched that magnitude, and the decomposition again sits almost entirely in the real leg — 6 of the 7bps came out of the real 10Y against 1bp at the breakeven, easing the discount rate without marking down nominal growth compensation. The front end lagged at 4bps, taking another 3bps out of the 10Y–2Y spread; the 2Y at 4.20% still prints above the 3.63% Fed Funds level stamped 1 July 2026, so the week's easing lives in duration rather than in any repricing of the near-term policy path. Unlike the prior block, every rate and credit observation here carries the same 27 August 2026 stamp, so the cross-asset comparison is same-day rather than assembled across two dates. On the growth side there is one fresh number: initial claims at 203K, down 4K, stamped 22 August 2026. Unemployment, payrolls, Fed Funds and housing starts all still carry 1 July 2026 dates.
Key numbers
| Metric | Level | Read |
|---|---|---|
| Regime | Neutral | Model's authoritative read for 28 August 2026 |
| VIX | 14.43 | Calm band |
| Breadth >50-day EMA | 54.2% (530/977) | Mixed |
| SPY close | 769.28 | +7.7% vs 200-EMA (714.56) |
| 10Y Treasury | 4.67% | week over week −7bps (as of 27 August 2026) |
| 2Y Treasury | 4.20% | week over week −4bps (as of 27 August 2026) |
| 10Y–2Y spread | 0.47% | week over week −3bps (as of 27 August 2026) |
| 10Y breakeven | 2.33% | week over week −1bp (as of 27 August 2026) |
| Real 10Y rate | 2.34% | week over week −6bps (as of 27 August 2026) |
| HY credit spread | 2.63% | week over week −7bps (as of 27 August 2026) |
| Fed Funds | 3.63% | as of 1 July 2026 |
| Initial claims | 203K | week over week −4K (as of 22 August 2026) |
| Unemployment | 4.1% | as of 1 July 2026 |
| Nonfarm payrolls | 158.9M | as of 1 July 2026 |
| Housing starts | 1,239K | as of 1 July 2026 |
Regime assessment
The rate and credit channels both eased over the week, and every classified input except breadth sits inside the band the model treats as healthy. When the non-equity channels ease in the same direction and the same magnitude, they stop being an independent test of the read. Two inputs agreeing tells you less than two inputs disagreeing would, and the rate and credit block contains no disagreement to work with.
That concentrates the load. Breadth is the only input outside the healthy band, which makes participation the place the classification is most exposed — and it has slipped into the mixed band with volatility and credit both undisturbed. Anyone weighing how durable this read is is weighing that flag and little else.
The character of the easing also bounds what it can support. A falling real yield addresses what future cash flows are discounted at; it says nothing about the cash flows themselves. The only evidence on the sheet that speaks to those is a weekly claims series and a credit spread, sitting against a labor and housing block frozen at 1 July 2026. The read is currently resting on market-priced inputs because the fundamental ones cannot update until their next monthly cadence — that dependence, not any level in the table, is the honest limit here.
What would change the call
- Breadth staying out of the healthy band. 54.2% (530/977) sits in the mixed band. Remaining there keeps participation the first equity input to fail without help from any other channel, and is the most direct refutation this block permits.
- HY reversing at comparable speed. 2.63% tightened 7bps this week. A widening of similar magnitude across consecutive prints restores credit as a dissenting input and removes the strongest observation in today's data.
- VIX exiting the calm band. 14.43 sits inside it. A sustained exit removes the volatility condition underpinning the classification; a single spike does not.
- A real-rate-led back-up in yields. The real 10Y fell 6bps to 2.34% against 1bp at the breakeven. Rising nominals driven by the real component reverse the discount-rate channel described above and would be the specific rate-side condition to watch.
- Continued flattening at 0.47%. The 10Y–2Y gave up another 3bps. Progress toward inversion converts the curve into a dissenting input; a second week of the front end lagging the long end does not on its own establish that.
- Compression toward the 200-EMA. SPY at 769.28 stands 7.7% above 714.56. Sustained erosion of that gap alongside weakening breadth would pressure the read on equity inputs alone.
What to watch
- Next weekly claims print. 203K, stamped 22 August 2026 and down 4K, is the only high-frequency labor observation available. A directional reversal sustained across two or three prints is the earliest evidence that could contradict the growth side of this read.
- Refresh of the 1 July 2026 block. Fed Funds at 3.63%, unemployment at 4.1%, payrolls at 158.9M and housing starts at 1,239K all update on monthly cadence. Until they print, none can arbitrate a live growth question.
- The next breadth reading. With every other input in band and easing, the participation figure is the only one positioned to change the classification in the near term — either by recovering from its current 54.2% into the healthy band or by weakening further within the mixed band.
Status
Neutral for 28 August 2026 — first neutral read in the journal, a flip from risk-on; prior level 27 August 2026, risk-on.