To The Moon

[Smile]Hi Tigers,

The calm surface of the S&P 500 masks one of the largest rotations in years—funds are flowing from crowded AI trading to cyclical sectors. JPMorgan Private Bank characterizes this shift as a repricing of "AI economics," not a liquidation.

Kriti Gupta, global investment strategist at $JPMorgan Chase(JPM)$ Private Bank, says the path change is not a market crash, but a leadership reshuffle.

She characterizes this shift as "a move from pricing potential to demanding proof of earnings." Evidence is evident in the index divergence: the market capitalization-weighted $S&P 500(.SPX)$ is hovering near all-time highs (around 7,490 points, still below the record of 7,616 points) after a two-day rally, while the equal-weighted index has hit new highs.

Since May 29, the equal-weighted S&P 500 has risen about 3.3%, while the market capitalization-weighted index has fallen 3.5%; the top ten companies by market capitalization lost a combined $658 billion in July.

The $Philadelphia Semiconductor Index(SOX)$ has fallen nearly 24% since the end of June, while the healthcare sector has risen about 9% and the financial sector about 5% over the same period.

Issuers emphasized the causal channel across interest rates. The 10-year real yield has risen 27 basis points in six weeks, a repricing reflected in duration sensitivity: the 120-day correlation between the Philadelphia Semiconductor Index and the 30-year Treasury yield has turned negative since the end of April—to about -0.3—proving that semiconductors are currently trading as a long-duration asset. The July 29 FOMC meeting reinforced the hawkish tone: the Fed voted 9-3 to keep interest rates at 3.50%–3.75%, with Beth Hammark of Cleveland, Neal Kashkari of Minneapolis, and Lori Logan of Dallas voting against a 25-basis-point rate hike. Chairman Kevin Warsh withdrew forward guidance at the second meeting, and after the decision was announced, the 30-year yield hit its highest level since 2007 (about 5.21%), and the 10-year yield approached 4.69%. Technology sector volatility surged (TMT momentum 30-day volatility soared above 45), while S&P 500 volatility remained in the 10–20 range.

Exchange rate factors were secondary but noteworthy: KKR noted that a weakening dollar coupled with rising interest rates tested the Warsh no-guide mechanism. Credit spreads and cryptocurrency transmission are not part of the current rotation's recorded scope, therefore these channels remain unproven and are not active drivers. In equities, leadership has shifted to financials, healthcare, real estate, and energy sectors—energy is currently the best-performing sector in the S&P 500 so far this year, supported by stronger oil prices driven by Middle East tensions and tariff-related cost shocks.

The earnings landscape complicates the "proof of earnings" test. FactSet's Q2 composite growth rate was a staggering 47.4%—the best since Q2 2021—with 61% of companies reporting earnings, 86% of which exceeded expectations, averaging a record +31.4%. However, this headline figure is exaggerated by non-operating income: Alphabet contributed approximately $98 billion, and Amazon contributed approximately $53.4 billion (primarily from Amazon's stake in Anthropic). Excluding these two companies, the hybrid growth rate drops to 28.8%. The forward P/E ratio is 19.6x, compared to a five-year average of 19.9x. Gupta stated that what can stabilize AI trading is measurable business results—profit margin improvements driven by productivity gains—and a return to a state where strong, better-than-expected performance leads to higher stock price returns, rather than a decline indicating risk aversion.

The next observable milestone is the September FOMC meeting (September 15-16, dot plot to be updated), after which FedWatch pricing shows a probability of a rate hike of approximately 45%, down about 20 percentage points from pre-meeting levels. Approximately 39% of S&P 500 companies have not yet released their Q2 earnings, and there are two other date risks on the calendar: the US-China trade truce expires in November, and the US midterm elections. The bearish counter-argument is technical: a break below the support level around 7,309 for the S&P 500—and the swing low of 7,224—would turn rotation into a broad pullback.

The outlook for this trade hinges on a key reliance: whether AI capital expenditures can translate into measurable profit margins in an environment of 27 basis points of real yield repricing. Decisive eyes are the September 16 FOMC meeting and whether long-term real yields stop rising; an earlier signal is whether battered earnings reports are met with buying rather than selling—a clear signal from issuers indicating that repositioning, not liquidation, is complete.

Potential positives:

• "Rotation, not liquidation" scenario: Equal-weighted indices continue to narrow the gap with market capitalization-weighted indices, with continued breadth in the financial, healthcare, and energy sectors, while the semiconductor index stabilizes.

• Positive earnings response returns—stronger-than-expected earnings leading to higher stock price returns—a clear sign from issuers that de-risking is over.

• Excluding non-operating gains from Alphabet and Amazon, the potential profit growth of 28.8% provides a more sustainable fundamental basis than the 47.4% implied by ByteDance.

• The flattening of 10-year real yields and the normalization of the negative correlation between semiconductors and 30-year yields will alleviate duration pressures on the AI ​​sector.

Potential risks

A break below the support level near 7,309 for the S&P 500—and the swing low of 7,224—would translate rotation into a broad-based pullback across all sectors.

A further surge in long-term real yields, especially if Warsh's lack of forward guidance pushes the dollar weaker while interest rates rise, could force a full-scale "liquidation" de-rating of all duration assets.

Continued selling pressure on strong earnings exceeding expectations—marked by issuers as an unresolved de-risking signal—would negate the "earnings proof" stabilization argument.

AI capital expenditures have failed to translate into profit margins and earnings, causing the capital expenditure cycle of hyperscale cloud providers (AWS up 37%, Azure up 42%) to be priced in without sustainable profitability.

FAQ

Q: Why is the tech sector pullback called a repricing rather than a liquidation?

A: JPMorgan Private Bank believes that investors still believe AI will reshape the economy, but are no longer simply paying for spending. In an environment of higher real interest rates, the market is shifting from pricing potential to demanding earnings proof—hence the leadership rotation to cyclical sectors with more defined recent returns, while the broader bull market pattern remains intact.

Q: How much of the S&P 500's 47.4% earnings growth is real?

A: Headline data is beautified by non-operating earnings—Alphabet contributed approximately $98 billion, and Amazon contributed approximately $53.4 billion (primarily through Anthropic). Excluding these two companies, the Q2 hybrid growth rate drops to 28.8%.

Q: What would confirm or break the rotation argument?

A: Confirmation Signal: The gap between the equal-weighted index and the market capitalization-weighted index continues to narrow, cyclical sectors maintain breadth, and strong earnings exceeding expectations lead to higher stock price returns. Break Signal: The S&P 500 index has clearly broken below the support level of approximately 7,309, long-term real yields have surged further, or AI capital expenditures have failed to translate into profit margins.