Chip-led selling rotated into defensives rather than out of stocks

What changed

The Oct 16 sell-off stayed inside the stock market: selling hit chips, tech and the Nasdaq 100 hardest while energy and financials closed higher, and our provisional answer is that the day was a rotation from rate-sensitive growth stocks into defensive large caps, not money leaving equities. The catalyst was inflation. September core CPI rose 0.3% on the month against a 0.2% consensus. September producer prices, out a day later, rose 1.3% against 0.5%. On the day stocks fell, WTI crude rose 3.92%.

The chain from oil to growth stocks runs through the rate path. Michigan one-year inflation expectations reached 4.7%. Futures now price 74.1% odds of a quarter-point hike at the December meeting. A higher hike path thins the extra return stocks offer over bonds, and growth stocks, whose value rests most on earnings far in the future, give ground first when that cushion shrinks. The 10-year yield barely moved, so the pressure came from the level yields have already reached rather than from a fresh jump.

A day of broad de-risking would have pulled sectors down together. The equal-weight sector data, which give every company the same weight regardless of size, show the opposite: consumer defensive, industrials and consumer cyclical rose, while technology fell. Volume points the same way. XLK traded 15,304,600 shares against an average of 8,669,950, so the heavy selling was concentrated in the sector that drives the index. Prices splitting this cleanly by sector is the first reason we read the drop as a reweighting within equities rather than a liquidation.

What hasn’t changed

The reweighting within equities stayed contained because three conditions underneath it held:

• Fed target range at 3.75-4.00% — unchanged since the Sep 16 hike. The next decision comes on Oct 28, and futures price 83.9% odds of a hold.
• High-yield spread at 323bp — the extra yield on junk bonds over Treasuries, on FRED’s Oct 15 observation, still below the 350bp line where our credit gauge turns cautious. That reading predates the Oct 16 session, so the official series has not yet caught the day’s moves in high-yield ETFs.
• Stocks and bonds moving together — the 60-day correlation between SPY and TLT is +0.4. On Oct 16 TLT rose only 0.1% while SPY fell.

With a further hike still on the table and long bonds not absorbing equity losses, money coming out of growth stocks had few places to go other than the front of the curve and defensive sectors. The fall in the 2-year yield, the maturity most tied to Fed policy, hints that short-dated Treasuries are starting to work as a hedge again even though the long end is not. Calm official credit gauges are the other reason the tech decline reads as a reshuffle inside equities, which makes the day lighter than the chip losses alone suggest.

Our read

The reshuffle inside equities matches how options are priced. Because positioning is hedged and dealers are short gamma, the more painful move for the index would be upward, not downward: dealers whose hedging amplifies price swings would have to chase a rally. On the surface, Oct 16 looked like a day when tech broke and the whole market shook. Underneath, investors held their stocks and bought insurance against a large decline.

Put simply, investors kept their stocks and paid up for crash protection, and in credit the traded ETF flinched before the official spread did.

The record on heavy hedging supports the same conclusion. CBOE Skew is the price of far out-of-the-money index puts relative to ordinary options, and it now sits above 150.

Returns in the high-Skew band were lower, but drawdowns were barely deeper. That history amounts to only about 11 non-overlapping windows, so the sample is small. Read with that caution, heavily insured markets have tended to cap gains more than they have deepened declines.

Valuation explains why tech was the first thing sold. A 10-year yield of 5.10% now competes almost directly with the S&P 500’s 5.26% earnings yield. That leaves a nominal equity risk premium, stocks’ earnings yield minus the 10-year yield, of just 0.16%. The tech-heavy Nasdaq 100 trades at 23.3 times forward earnings, a richer multiple than the broad index carries, so it has the least room when the reward for owning stocks is this thin. Earnings lean the same way: S&P 500 estimates rose 3.3% over the past month, while Nasdaq estimates were flat. Trimming tech while keeping other large caps is the choice those revisions favor.

The case against a simple rotation is real. QQQ closed at $709.8, below the lower edges of both its Bollinger Bands and its Keltner Channels, two gauges of its normal trading range, and its trailing trend line has flipped down. SPY’s MACD momentum is negative on the weekly, daily and four-hour charts. SPY is also inside a volatility squeeze with no trend in place, and of the eight past squeezes that resolved under similar conditions, five broke lower. That keeps open a path in which weakness that began in tech spreads to the broad index.

On balance, the evidence favors rotation. The heavy volume was concentrated in tech, defensive sectors rose, the VIX stayed below 20 and official credit is quiet. One combination would change that read. The first leg is an SPY close below the $753.5 to $756.2 shelf, where four support levels cluster. The second is VIX/VIX3M, the ratio of one-month to three-month implied volatility, closing above 1.0. Both at once would mean the tech weakness has turned into selling of the index itself.

Protection this heavy also makes a rising index the more uncomfortable direction for most participants. A de-escalation around the Strait of Hormuz or a soft September core PCE on Oct 29 could prompt hedge holders to sell their protection, and short-gamma dealers would then add to the rise. With so many positioned for a decline, the costlier surprise today is a rebound, and that cost falls mostly on positions that bet on further declines.

Options positioning

• Institutional flow: net options premium sits in a neutral band, so it gives no directional read, although the largest unusual trades were institutional in size.
• Index downside protection: demand is high without panic — investors are insuring holdings rather than selling them.
• Dealer positioning: SPY and QQQ dealers have been short gamma in 7 of the last 10 sessions, which amplifies moves in either direction once price leaves its range; the provenance of this dealer data is disputed, so we give it limited weight.

Checking our last call

The rotation reading also explains how the lines from our last edition resolved.

A correction to our Oct 16 edition: we gave PPI excluding food and energy as 0.3%, and that figure was wrong. It rose 0.7% on the month, far above the 0.3% consensus, so both conditions on that line were met, not only the headline one. Under the rule we set, stagflation should have gained 5pp; the move we reported was 1pp. Our scenario work gave the PPI release between no weight and a 1pp addition to stagflation, and today’s data do not say why the full 5pp was not applied.

• September core PCE (Oct 29): not yet released. The hot outcome now lifts stagflation to 47%, against 48% in our last edition. The soft line moved to 0.24% or less, which with energy relief would lift Goldilocks to about 22%.
• Fed Chair Warsh’s speech, the import price release and the Graham Act sanctions deadline from our last calendar: their outcomes are not in today’s data, so we do not score them yet.

Since our last edition, stagflation, slowing growth with inflation staying high, slipped to 36% from 38%. Reflation, growth running alongside rising inflation, rose to 32% from 27%. The lead narrowed to 4pp from 11pp. Our macro data alone put reflation slightly ahead, 35% to 33%. We moved 3pp to stagflation because small caps are lagging, stocks and bonds are moving together, and QQQ is trading below its lower band. A 4pp lead sits inside the error band of both our macro and technical readings, so we treat the ranking as unsettled until Oct 29. Beyond that adjustment, today’s data give no separate reason for the change since our last edition.

The first line matters most. Tech stocks fell while the 10-year yield and the dollar index stayed below the lines we flagged. A stock decline that does not drag rates and the dollar up with it looks less like an inflation scare spreading across markets and more like a reweighting inside equities, which is today’s read.

Key dates

Whether the reweighting inside equities stays contained will be tested by a run of housing, energy and labor data before the Oct 28 Fed decision and the Oct 29 inflation release.

• What to watch: whether SPY closes below $756.2

While SPY holds above that line, the tech decline reads as a shift in weight between sectors. A close below it, with the small-cap and equal-weight ratios also under their averages, would complete the conditions that break the reflation reading. Today only the small-cap ratio has crossed; the other two have not.

What it means for investors

How much the shift in weight between sectors matters depends on what a portfolio holds. Today’s data carry no six-axis alignment table, our tally of how many of six signals (macro, valuation, flows, chart, sentiment and events) point the same way. SPY scores 1 of 6 and QQQ was not scored, both far from the five-of-six level we treat as high confidence. Our overall risk grade is watchful (YELLOW).

• Chip- and Nasdaq-heavy portfolios — the risk that changed is concentration: the selling and the heavy volume landed in this group, and QQQ now trades below both of its volatility bands. QQQ back above its weekly 20-period average near $713.4 would be the first sign that the selling inside US tech has stopped.
• Energy holders — energy shares rose with oil, but USO, the crude oil fund, trades above its upper Bollinger Band, and geopolitical odds put a Hormuz reopening at 60%. The same de-escalation that would lift the index would hit this holding. Another large draw in EIA crude inventories on Oct 21 would show that physical supply, not only geopolitical fear, is holding prices up.
• Balanced portfolios that lean on long Treasuries — the assumption that long bonds cushion a stock decline did not hold on Oct 16, when only the 2-year rallied. A 20-day SPY-TLT correlation, now +0.2, falling below zero would be the first sign that the hedge is working again.
• Dollar-cost-averaging index investors — the index fell far less than tech, but SPY’s short-term momentum has turned down and its squeeze is unresolved. An SPY close back above its 20-day average near $770.6 would show the index has absorbed the tech selling.

The thin valuation cushion frames the equity decision for all of them. With the index’s earnings yield only 0.16% above the 10-year yield, staying in the index is a bet that earnings keep growing rather than that valuations expand, and the current estimate revisions favor the broad large-cap index over the Nasdaq.

If yields stay where they are, portfolios heavy in long-duration growth keep absorbing the valuation pressure first, while defensive large caps carry the weight the market moved toward them.

For investors, the open question is whether the rotation into defensives stays a rotation: as long as SPY holds its support shelf and VIX/VIX3M stays below 1.0, Oct 16 marks a change in which stocks the market wants to own, not a move away from owning stocks.

Glossary

• Rotation — money shifting between sectors within the stock market instead of leaving it.
• Short gamma — options dealers whose hedging buys into rallies and sells into declines, which enlarges moves.
• Volatility squeeze — Bollinger Bands contracting inside Keltner Channels, a sign that price swings have compressed before a larger move.