Original market research, tested against the data.
SMH closed 21.8% below its June record on October 22, 2026, its second break since July, while the S&P 500 sat 2% from its own. Since 1994, chips 20% below their high with the index near its high happened 13 times. The S&P 500 was higher a year later in 11, by a median 19.3%. The two exceptions, 2000 and 2007, were the only times the index took years to get back to its high, and at no bear-market peak since 1973 were chips already 20% down. Chips themselves usually fell further.
The Nasdaq A-D line has diverged for 63 months, longer than the NYSE before 2000. The case for and against another warning, measured, plus the AI trade.
S&P 500 record on Oct 6, 2026 with 69 NYSE lows vs 37 highs. 47 were income funds and REITs; stock lows were led by utilities and consumer names.
The diesel crack is at a record. After 9 spikes since 1989, the S&P 500 had a weak first year; from six months in, the next year was close to normal.
Gold is 23% below its record. Since 2004, three 20% declines recovered within 18 months; one became the 2013 bear market. The case for each, measured.
The S&P 500 rose in the year after all 19 midterms since 1950, by 16.6% on average. Before 1950 it fell twice, and years starting near a high did less.
43 sessions without a 1% drop. Since 1929, long calms that broke near a high led to a 10% decline 8% of the time, against 11% for other 1% drops there.
Today vs the 2000 top: profits are growing and broader than the headlines, but valuation rests on a record profit share. What could break it, measured.
S&P 500 near a record with 26% of stocks above their 50-day: since 1927 that setup came before the 1973 and 2000 tops, and 2023. Most tops were broad.
XLF is 6% below its high with SPY near a record. Prior episodes returned 2.57% over a quarter versus 2.65% for the baseline. See every case and the limits.
QQQ set a record on Sep 22, 2026 with 37% of Nasdaq stocks on buy signals, more new lows than highs and 47 of 100 members above their 50-day. 8 prior cases.
BPI at 30.6%, just above the oversold line of 30, with SPY 2% from a record. No dependable edge: seven prior cases look weak, five spaced ones ordinary.
As of Sep 14, US 10-year yields rose 91 bp since March. The refinancing bill, limits of Fed cuts, and what higher yields mean for the economy and dollar.
SPY's daily range is at its 5th percentile since 1993, yet the VIX sits at its 58th. Prior calm spells brought ordinary returns and ended within 58 sessions.
August 2026 review: stocks rose, breadth narrowed, metals and crypto surged, business accelerated, and household and energy strains deepened.
AI capex more than doubled, yet Nasdaq-100 revenue, operating profit and free cash flow still rose. The fundamental stress test says no crack yet.
Nonfinancial corporate margins rose to 15.2%, third-highest since 1947, while consumer sentiment set a 47-year low. What drove the gap and what stocks did next.
Across 18 comparable Q2 13F filings, Alphabet led current buying while TSM and Amazon showed the strongest multi-quarter accumulation.
XLE at a 28-year record with Brent 31% below its April high. Refiners did the work on a 99.9th-percentile crack spread. The last time, November 2022, the margin halved and XLE lost 12%.
Existing-home sales at 4.06M, a 2010-trough pace; prices +1% nominal, −2.8% real; owners at 72% equity, delinquency 1.9%. Frozen and slowly deflating, not 2008.
Copper hit a record $13,552/tonne. We ranked all 1,901 US-listed $2B+ stocks by 15 years of copper correlation: the top miners, the diversified route, the indirect plays — and what the 2011 bust did to each.
WEC fell 11 straight sessions and every large utility dropped 3–10% while SPY set records — and the 10-year barely moved. What 43 long utility streaks since 2010 did next.
Tech beat energy by 10.6 points in five sessions, but most sectors rose. After 12 similar rallies near highs, SPY’s next quarter matched its baseline.
45 of 101 Nasdaq-100 members are 20%+ below their own highs while QQQ sits 3.5% from a record. This shape has 22 precedents in 15 years — 20 of them this year.
The VXN−VIX spread hit 11.1 pts — its widest since 2003, a level seen only in the dot-com unwind — while VIX sits at a calm 16.6. What followed similar readings.
Headline PCE hit 4.1% and the real fed funds rate just went negative (−0.4%). Trump hired Warsh to cut; his Fed is signaling hikes. Is the Fed cornered? Mostly — here’s the nuance.
US sentiment hit a record low, yet real spending is up 2.1% and delinquencies are falling. The risk is thin savings, not vibes.
A record $8 trillion sits in money market funds, $2.25T of it retail. Will it fuel stocks? The data says it depends on one thing — and it isn’t the cash.
Leveraged funds hold a record net short in S&P 500 futures, yet the index is at all-time highs. Why the basis trade — not bearishness — explains it, with our own data.
Household equity allocation reached 45.8% after a record 47.1% — above the 2000 peak. Why it points to lower long-run stock returns and says nothing about an imminent crash.
CPI accelerated from 2.4% to 4.2% in four months while PPI hit 13.1% — a 97th-percentile gap. What the five prior episodes since 1948 did to consumer prices.