What changed
Chip stocks slid on fears of competition from Chinese memory chipmaker CXMT, yet the dollar weakened, and the push came from the bond market rather than from stocks. The 30-year Treasury yield fell 4.0bp. The 2-year, the maturity most tied to Fed policy, rose 0.8bp. Our provisional answer: the dollar tracked long-term yields, not a repricing of the Fed, so its dip is unlikely to outlast a test from the Oct 14 CPI.
The chip selling hit MU, NVDA and Korean chipmakers. When only long yields fall, the gap between long and short rates narrows, a flattening of the curve. The dollar index (DXY) followed that long-end decline rather than the 2-year, and it softened against the euro, the pound and the Swiss franc. The rest of the moves are in the table.
| What moved | Latest | What it says |
|---|---|---|
| 10-year · 2-year Treasury yields | 5.22% (down 2.0bp) · 4.80% | Only the long end fell |
| Dollar index (DXY) | 102.2 · -0.4% from the prior close | It tracked long yields, not the 2-year |
| Equal-weight utilities · consumer cyclicals | +1.75% · -0.96% | Defensive sectors led |
| Spot gold | -0.12%, a second down session | No lift from a weaker dollar and lower long yields |
| Bank earnings | JPM missed but raised net interest income guidance · GS beat on a 72% jump in equities revenue · WFC mixed | Earnings season opened mixed |
Beyond the index, money moved toward defensive sectors rather than out of stocks, and gold got none of the help a softer dollar usually brings. Goldman Sachs said investors were buying safe stocks at a pace comparable to the COVID period. We read that as rotation, not de-risking: institutions are changing what they own, not how much stock they own.
What hasn’t changed
The rotation toward defensive sectors and the dollar’s dip stand out because three larger conditions underneath them stayed put:
- Scenario weights — stagflation, inflation staying high while growth slows, at 31%, and reflation, growth running with rising inflation, at 27%. Our assessment has held there since Oct 9, and the September CPI on Oct 14 is the next trigger to revisit it.
- Fed target range at 3.75-4.00% — unchanged since the September decision. The next decision comes on Oct 28.
- Stocks and long bonds moving together — the 20-day correlation between SPY and the long-Treasury ETF (TLT) is +0.5. Correlation measures how closely two assets move in the same direction; a positive reading means bonds tend to fall when stocks fall, so they give little cushion.
With all three in place, only the curve and the dollar moved. That points the search for the dollar’s dip toward the Treasury market, not toward stocks: on a day chip stocks fell, the push behind a weaker dollar did not come from equity demand. It also makes the dip lighter than it looks, because none of the slow-moving conditions shifted with it.
Our read
If the Treasury market pushed the dollar down, the next question is which story the long end is pricing. We think the Oct 14 CPI will decide whether cooling hiring or an energy supply shock takes the lead, but one print alone cannot crown a winner. In the hiring story, weak payrolls rally long bonds, the dollar softens and defensive sectors lead. In the energy story, oil, gas and producer prices keep short-term rates firm.
On the surface, falling long yields and a softer dollar looked like markets reviving hopes for Fed easing. Underneath, only the long end fell, and the policy-sensitive front end held its ground.
| On the surface | Underneath |
|---|---|
| 10-year and 30-year yields down, dollar weaker | The 2-year rose 0.8bp — no Fed easing priced |
| The index held — S&P 500 futures -0.03% | Tech carried the selling — Nasdaq 100 futures -0.38%, SOXX below its moving averages |
| Fear gauge low — VIX 15.27 | Crash insurance expensive — CBOE Skew 154.74, a third straight session above 150 |
Put simply, the rates and currency boards leaned toward easing, but the 2-year yield, the first place a change in the Fed’s path shows up, went the other way.
The kind of dollar weakness this was is why we doubt it lasts. Because the dollar followed long yields rather than the 2-year, the move reads as a change in the curve’s shape, not a Fed repricing. Over seven days the dollar index is still up 0.7%, so one down session does not reverse the trend. The 2-year also sits about 92bp above the midpoint of the Fed’s target range, which means Treasuries themselves still price a higher path than today’s policy rate.
The energy story has not lost force either. Among commodities only oil and natural gas drew buying, while copper and silver fell, so the oil move reads as a supply premium rather than a demand boom. A supply shock lifts prices without lifting growth, and that is one reason short-term rates are not coming down.
The evidence on the other side is real. When ten recent market moves are scored against the competing stories, cooling hiring fits best, with twice the score of the energy shock. If the hiring weakness behind September’s soft payrolls keeps pulling long yields lower, the dollar’s dip could last longer than we expect. Neither fit clears the bar we use to call a winner, though. Even a hot CPI with the 10-year closing above 5.24% and the dollar index back above its prior close would leave the energy story short of that bar.
Nor does the bounce in long bonds mean the equity cushion is back. TLT rose 0.4% but stays below its three moving averages, a rally inside a downtrend. Bonds, not stocks, are setting the pace: the MOVE index, the bond market’s implied-volatility gauge, stands 4.6% above its 20-day average, while implied volatility in stocks has not followed.
On balance, the evidence supports reading the dollar’s dip as a bond-market move rather than a sign of easier Fed policy. Three things point the same way: a curve where only long yields fell, a dollar that ignored the 2-year, and chip stocks falling on an industry-specific worry rather than a macro one. The condition that would change our read is a hot CPI that fails to lift the 10-year and the dollar index with it. If the dollar stays soft through a hot print, its weakness is more than a quirk of the curve, and we would treat it as a trend with staying power.
Options positioning
- Institutional flow: net options premium across the market was positive but stayed inside a neutral band, so it gives no directional read.
- Index downside protection: expensive, and increasingly so as a baseline. Skew sits above its 20-day average, and that average itself keeps rising.
- Dealer positioning: in SPY, dealers are moving toward positions that dampen moves; in QQQ, toward positions that amplify moves in either direction.
Checking our last call
With QQQ dealers positioned to amplify moves, here is where the lines from our last edition stand before the CPI.
| What we flagged | What happened | Verdict |
|---|---|---|
| IWM back above its 20-day average at $282.7 | $276.74 | Not reclaimed |
| WTI closing below $89.2 before Oct 29 | $94.8 | Open |
| SPY closing above its 20-day average after the Oct 14 CPI | CPI not out yet; SPY near $778.04, above its 20-day average of $771.49 | Open |
| Equal-weight industrials rising alongside copper | Copper -0.51%; equal-weight industrials -0.64% | Not met — both fell |
| Dollar index close below 102.2 with EUR/USD above 1.13 | DXY 102.203; EUR/USD 1.11993 | Not met — the dollar fell, but EUR/USD stayed below 1.13 |
| SOXX back above $546.3 | $543.84 | Not met |
| December hike odds against the 50% line | No hike-odds figure in today’s data | Cannot be scored |
- September CPI (Oct 14) and PPI (Oct 15): not yet released.
Our framework has also changed the bars the CPI will be scored against:
| Rule | Last edition | Today |
|---|---|---|
| Soft CPI | Core 0.2% or less on the month, headline below 3.7%, December hike odds below 50% → goldilocks 31% | Core 0.1% or less → goldilocks +5pp; in-line core of 0.2% with headline at 3.6% or less → +2pp; no hike-odds leg |
| Hot CPI | Stagflation to a provisional 36% | Core 0.3% or more with headline 3.7% or more → stagflation +5pp |
| September core PCE | 0.37% or more → stagflation 42% | 0.37% or more → stagflation +12pp |
The soft-print bar moved the most: the core threshold is lower and the hike-odds leg is gone. Our data does not give a reason for the change. After the print we will score the soft side against both the old and the new bar.
Key dates
Both soft-print bars get their first test from the CPI, with producer prices and retail sales close behind.
| Date | Event | Why it matters |
|---|---|---|
| Oct 14 (Wed) | September CPI (consensus: headline 3.6% on the year, core 0.2% on the month) | Core at 0.3% or more with headline at 3.7% or more adds 5pp to stagflation; if the 10-year then closes above 5.24% and the dollar index above 102.589, the dollar’s dip has unwound |
| Oct 14 (Wed) | Speech by the Fed’s Barkin | The first Fed communication after the print |
| Oct 15 (Thu) | September PPI (core consensus 0.3% on the month) | A core beat would show energy costs still working through at the producer stage |
| Oct 15 (Thu) | Retail sales control group (consensus 0.3%) and initial jobless claims (consensus 200K) | A negative control group lifts recession odds; claims above about 236K would trip one of three labor turning-point tests |
| To be confirmed | September core PCE | A monthly rise of 0.37% or more adds 12pp to stagflation; below 0.25% keeps momentum under 2.5% |
- What to watch: the dollar index against its prior close near 102.6 after the Oct 14 CPI
That prior close is where our reading of the dollar gets tested. A close back above it after a hot print would mean the long-end move has unwound. A dollar that stays below it through a hot print would argue for a more durable decline, and for revisiting the view that this was only a curve move.
What it means for investors
A test of the dollar’s prior close lands differently depending on the holding. Today’s data carries no six-axis alignment table, the tally of how many of six signals (macro, valuation, flows, chart, sentiment and events) point the same way. Only SPY is scored, at zero and reported as not available, well short of the five-of-six bar for high confidence, so no asset is near it.
- International holders without currency hedges — the dollar’s dip lifted the dollar value of euro, pound and Swiss franc assets, but because it rode on long yields, a hot CPI that pushes the 10-year back up can take that gain away. EUR/USD back above its nearest moving average with USD/JPY below 158.2 would be the evidence that dollar weakness has a broader base.
- Semiconductor and tech-heavy portfolios — the risk that changed is industry-specific: competition from CXMT, not rates, drove the selling, and QQQ dealers are positioned to amplify moves, so the same CPI can shake the Nasdaq more than the S&P 500. SOXX back above its $546.2 moving average would be the first evidence that the competition worry is leaving the price.
- Balanced portfolios that lean on long Treasuries — the TLT bounce did not restore the usual offset; with stocks and long bonds moving together, a stock sell-off would likely take bonds with it. A 20-day SPY-TLT correlation below +0.2 would be the sign that duration hedges again.
- Gold holders — gold fell for a second session even as the dollar and the 10-year yield both eased, so the price is not following its usual macro drivers. One session does not disprove gold’s role as a hedge, but its near-term direction is unsettled. GLD back above its 20-day average at $386.2 would be the first evidence that the softer dollar is reaching it.
Long bonds have stopped working as the shock absorber for stocks, and they stay that way while the correlation is positive. Short-dated Treasury bills sit outside that joint-decline risk, and three-month bills pay 4.06% while the answer arrives.
Valuation leaves little slack. The S&P 500’s earnings yield sits barely above the 10-year yield, so stocks are not cheap against bonds. With bills at that rate, holding the index is a wager that earnings keep growing, and the support for that wager today is the 5.6% upward revision to S&P 500 earnings estimates over the past month.
A 10-year close above 5.24% after the CPI would hit portfolios that hold both stocks and long bonds on both sides at once.
If the dollar’s dip came from long yields rather than from the Fed, the question for any holding priced against the dollar is whether that dip survives the CPI.
Glossary
- Exponential moving average — an average of past prices that weights recent days more; price above or below it is a simple trend gauge.
- Equal-weight sector — a sector measured with every stock at the same weight, so a few giants cannot dominate the reading.
- Net interest income — what a bank earns on loans and securities minus what it pays on deposits and other funding.
- CBOE Skew — the price of far out-of-the-money S&P 500 put protection relative to ordinary options; higher means crash insurance costs more.
- Dealer gamma — whether options dealers’ hedging dampens or amplifies price moves.