"Terrifying data" adds fuel to the AI bull market? U.S. retail sales experience the largest decline in over a year, and interest rate hike expectations are hit hard again.

The U.S. retail sales data, dubbed "horrific data," significantly fell short of market expectations, leading to a major cooling of financial markets' expectations for an interest rate hike by the Federal Reserve before early 2027. The dovish outlook for the Federal Reserve arising from this "horrific data" has further fueled the recent global semiconductor resurgence leading to a new AI super bull market. The latest series of U.S. economic data has clearly shifted the Federal Reserve's monetary policy balance from "must restart the rate hike path quickly" to "the Federal Reserve can remain on hold (i.e., maintain interest rates)." The Fed now faces a reality of "overheating demand + inflation re-accelerating," where inflation is still above target but showing signs of marginal cooling, while consumption and employment begin to lose momentum. In July, U.S. retail sales unexpectedly marked the largest drop in over a year, highlighting that American consumers have reduced their shopping expenditures at online stores and auto dealerships. The data released by the U.S. Census Bureau on Friday showed that the retail purchase amount, unadjusted for inflation, decreased by 0.6% month-on-month, the largest decline since May 2025, significantly below the market consensus expectation of a slight 0.1% monthly growth, and a considerable drop from the previous value of 0.2%. When excluding autos and gasoline, the retail sales data dropped by 0.2%. Following strong consumer spending in the second quarter, July's retail sales fell, particularly with the directly GDP-relevant control group unexpectedly showing weakness. The decline was exacerbated by reduced consumer spending during June's Amazon Prime Day event, although sales in the restaurant sectora key indicator of discretionary spendingremained stable, said Andrew Sacher, a senior economist at Bloomberg Economics. Since around 70% of U.S. GDP components are driven by consumption, the retail sales data plays an important role in guiding investors' assessments of the current state of the U.S. economy and the outlook for monetary policy. It is referred to as "horrific data" because it is crucial for macroeconomic expectations and the Federal Reserve's policy trajectory, often causing significant volatility in financial markets, including the stock market. U.S. consumption has suddenly "hit the brakes"! The soft landing trade is facing a key stress test. This U.S. retail sales data report indicates that after experiencing relatively strong consumption in the first half of 2026, consumers temporarily slowed their spending last month. However, some analysts have warned that these data might be influenced to some degree by a pre-release of consumer spending, mainly because the e-commerce and cloud computing giant Amazon moved its Prime Day promotional event from July last year to June this year. As shown in the chart, U.S. retail sales recorded their largest decline in over a yearparticularly, online store sales saw their steepest drop since early 2025. Among the 13 categories covered in the report, five showed declines, with sales from non-physical retailers like Amazon leading with a drop of 2.2%. Sales at auto and parts dealers decreased by 1.8%. Meanwhile, revenue for restaurants and bars increased by 0.5%; this was the only category in the retail sales report pertaining to the services sector. The so-called "control group sales"an indicator that is included in the governments calculation of goods consumption expenditure for GDPdeclined by 0.4%, the largest drop since early 2025. This metric excludes food services, auto dealers, building material stores, and gas stations. Earlier in 2026, substantial tax refunds had temporarily boosted consumption, and with the personal savings rate dropping to a four-year low in June, economists remain cautious about the future outlook for consumer spending. Data from Bank of America and PNC Financial Services Groups independent card statistics indicate that after sales surged in June due to the Amazon Prime Day promotions and the World Cup, the growth rate of consumer spending slowed in July. However, according to a survey by the Bank of America Institute, consumers' overall financial health appears to remain solid, especially as the independently compiled savings levels are still above pre-pandemic levels, and the proportion of households able to pay off credit card bills in full is increasing. CPI + PPI + non-farm payrolls + retail sales are hitting the hawkish stance of monetary policy from all sides, with interest rate futures markets indicating that the probability of a Fed rate hike in September has fallen below 30%, a sharp decline from over 50% prior to the CPI data release. The latest data combination has clearly shifted the policy balance from "must quickly resume rate hikes" to "the Federal Reserve can conditionally wait." Julys retail sales unexpectedly dropped by 0.6%, far below the market expectation of +0.1%, while the control group sales closest to GDP consumption accounting fell by 0.4%, even as the market originally expected a growth of 0.3%; this coincided with the unexpected decrease of 23,000 non-farm jobs in July, a mere 0.1% rise in CPI month-on-month, a 0.2% increase in core CPI, and zero growth in PPI. This combination of "cooling demand + lukewarm employment + no re-acceleration in inflation" directly diminishes the urgency felt by recent hawkish dissenters like Cleveland Fed President Beth Hammack, who have insisted that "a rate hike must happen now." Approximately five minutes after the retail data release, pricing based on federal funds futures indicated that the probability of maintaining the current rate of 3.50%3.75% in September rose to 70.4%, with a 25bp hike only at a 29.6% probability, compared to over 50% hawkish expectations a week prior; however, by December, the probability of current rates remaining unchanged is about 38.1%, with a rate hike once at about 44.0%, indicating that there remains significant disagreement in the interest rate futures market regarding at least one rate hike within the year and the possibility of maintaining the current rate throughout the year. The latest retail data also make Goldman Sachs senior economist Matheus Dibos prediction of the Fed staying on hold for the entire year more persuasive: the current key issue is not that inflation has returned to 2%, but whether earlier shocks from oil prices and tariffs have led to a genuine second-round effect." Dibo believes that housing inflation has room for further decline, the labor market is not overheating, and a wage-price spiral has not formed, allowing the Federal Reserve to wait for more data; the latest CPI/PPI reinforces this judgment. Notably, this is not an entirely isolated counter-view from Goldman Sachsearlier consultations and surveys of economists by Bloomberg Intelligence indicated that the median forecast still expects the Fed to keep rates unchanged for the remainder of 2026. In contrast, Hammack, Kashkari, and Logan, three members of the FOMC, voted 9 to 3 for a 25bp hike in July and still maintain that the policies have not formed sufficient restrictions and should take "action now." "Bad news" is once again becoming fuel for the AI bull market! With the largest drop in retail sales in over a year and a steep decline in expected Fed rate hikes, is the global AI computing power theme about to enter a new round of short squeeze? For the recent resurgence of the "AI super bull market" in the stock market, this retail data indeed adds fuel to the fire, but primarily affects the "valuation and liquidity" sidethe denominator of the DCF valuation model rather than the profit indicators in the numeratorthis does not imply that these economic data prove that AI demand has suddenly become stronger, but rather that it is constructing the "soft yet resilient" macro combination that the stock market favors. CPI, non-farm payrolls, and retail data are gradually serving as "valuation catalysts" defined by the market for the AI bull market, as long as cooling consumer spending does not transform into a recession, and oil prices do not reignite inflation, a continued wait-and-see approach by the Federal Reserve might actually represent the current most ideal macro environment for AI assets. If the economy merely cools from overheating to moderate growth, the Fed should not need to increase the risk-free rate further, which means that stocks focused on AI computing infrastructure and the semiconductor sectorassets with extremely long valuation durations and heavy capital expenditureswill experience a decrease in discount rates and risk premiums on financing costs, while the expected trajectory of AI CapEx (capital expenditures) from hyperscalers (cloud computing giants), as well as the deployment scopes of GPU/ASIC, HBM/DRAM/NAND, and data center optical interconnection infrastructure, are strongly driven by structural investments in AI computing power. Therefore, a "soft yet resilient" macro environment stands as a significant positive catalyst for AI computing infrastructure stocks and semiconductors, both of which have recently had a major impact on the global stock market, second only to actual AI computing power demand. The Philadelphia Semiconductor Index plummeted nearly 29% from its historical high on June 22 to its low on July 29, but has since rebounded by about 20% from the low as of August 13, suggesting that the sell-off in July looks increasingly like a "crowded AI position + extreme leverage position liquidation storm," rather than a reversal of the strong fundamental profit trend in the AI computing industry; the Korean stock market KOSPI index, seen as a barometer for AI computing investments, rallied from 5,593.56 points on July 30 to 6,977.94 points on August 14, cumulatively rebounding about 24.7%, with an 11.5% surge this weekSamsung Electronics and SK Hynix jumped by 19% and 16%, respectively, with foreign capital massively returning within the week. In other words, recent data from a series of supply chain indicators suggest that the fundamentals of AI computing have not only remained intact but are continuing to strengthen. The current market investment phase is increasingly factoring in the valuation catalyst of "lower risks of Fed rate hikes"earnings revisions and the easing of interest rate pressures are beginning to form a double impact. CPI, PPI, non-farm payrolls, and retail sales together have lowered the tail risk on interest rates, strong earnings for the semiconductor sector, coupled with the robust demand for computing power revealed at the AI computing industry level, strengthens the fundamental base, thereby forcing previously low-positioned funds to re-engage. However, this also explains why the current stock market sentiment is increasingly approaching a "FOMO + re-leveraging" scenario centered around the AI bull market, rather than viewing it as a simple inverse trade. Data from Citadel Securities shows that on August 4, the volume of call options for the S&P 500 index reached an all-time high, nearly double the average daily level over the past year, about 10% higher than the previous record; from July 30 to August 5, there was a record volume of call options traded in a five-day period, and around 35% of S&P 500 constituents showed inverted call skew for three-month callsinvestors are actively paying premiums for "missing out" risks.