Retail investors sell tech ETFs and buy semiconductor stocks despite weak inflows.

Written by: Rita

Retail investors sold tech ETFs but bought individual semiconductor stocks. J.P. Morgan’s Retail Radar shows retail net inflows of $3.4 billion this week, below the 12-month average of $6.7 billion. Tech ETFs dropped from the 73rd percentile last week to the 4th percentile, while semiconductor ETFs saw outflows of 2.1 standard deviations. At the individual stock level, NVIDIA received $942 million in buys.

JPMorgan's Retail Radar report, released on October 14, 2026, noted that retail flows on Wednesday were at the 7.5th percentile and for the full week at the 5th percentile, indicating weakness. Over the next seven days, 62 companies reporting earnings will account for 14% of the S&P 500’s market capitalization, with a concentration in financials and industrials. Retail investors have lost confidence in tech ETFs, but their buying of individual semiconductor stocks contrasts inversely with institutional positioning.

Retail investors saw a net inflow of $3.4 billion this week, below the 12-month average of $6.7 billion. ETF inflows totaled $3.2 billion, while individual stock inflows amounted to only $200 million. Wednesday’s flow ranked at the 7.5th percentile, and the weekly flow at the 5th percentile, indicating brief retail enthusiasm.

The S&P 500 fell on Wednesday, with the Nasdaq 100 dropping 2% and the VIX rising to 16.9. Core CPI rose 0.33% month-over-month and 2.6% year-over-year, exceeding expectations. J.P. Morgan analysts project core PCE to increase 0.31% month-over-month and 3.1% year-over-year—sufficient to bring the October FOMC meeting into consideration—but still expect the next rate hike in December.

Retail investors shifted to net sellers in tech ETFs, selling 4%, dropping from the 73rd percentile last week to the 4th percentile. The selling was driven by SOXL and SOXX. Momentum ETFs saw outflows of 1.0 standard deviation, with MTUM experiencing outflows of 0.8 standard deviation. The momentum factor weakened this month, as the long side heavily weighted in semiconductors underperformed the short side heavily weighted in software.

The semiconductor sector fell 9% this week, as retail investors did not sustain last week’s enthusiasm for tech buying. Tech ETFs saw outflows of 1.8 standard deviations, dropping from the 73rd to the 4th percentile. SOXL experienced outflows of 0.9 standard deviations, while SOXX saw outflows of 2.1 standard deviations.

JPMorgan noted that retail selling of tech ETFs contrasts with institutional positioning. The congestion indicator for semiconductors relative to software shows semiconductors at the 96.1st percentile and software at the 44.1st percentile. Retail investors are buying semiconductors and selling software, contrary to institutional positioning.

Fixed-income ETFs saw moderate inflows, with significant inflows into high-yield, investment-grade corporate bonds, and multi-sector credit bonds. Utilities have been consistently sold since mid-to-late August. The industrials sector was sold off following the start of the earnings season.

At the individual stock level, retail investors concentrated their buying in technology and communication services, while selling all other sectors. Buying was primarily driven by the Mag7 and SpaceX, with demand for the Mag7 matching the one-year average. Retail investors sold software and bought semiconductors and hardware; NVIDIA received $942 million in buys, SanDisk received $238 million, and Broadcom received $139 million.

Retail investors sold $123 million in Micron, $119 million in Palantir, $70 million in AMD, $64 million in Moderna, and $55 million in CrowdStrike. Buying was concentrated in NVIDIA and SanDisk, while selling was focused on Micron and AMD, indicating a divergence within the semiconductor sector.

JPMorgan noted that retail investors overallocated to semiconductors and memory in May, then shifted toward hyperscale manufacturers; despite recent losses, retail tech portfolios outperformed the market. Retail buying in semiconductor stocks contrasted with institutional positioning, and congestion indicators show semiconductors at historical highs.

The relative congestion indicator for semiconductors versus software shows semiconductors at the 96.1st percentile and software at the 44.1st percentile. Retail investors are buying semiconductors and selling software, contrary to institutional positioning. J.P. Morgan believes this contrarian contrast is noteworthy.

The momentum factor congestion level is at the 94.4th percentile, while the lag factor congestion level is at the 27.2nd percentile. The momentum factor is biased toward semiconductors, and the lag factor is biased toward software. J.P. Morgan notes that the long side of the momentum factor has underperformed the short side, and the momentum factor weakened this month.

Retail investors have continued to sell off utilities since mid-to-late August, with no change in this trend. The industrial sector has been sold off since the start of the earnings season. The financial sector is now entering its earnings season, with companies such as Wells Fargo, Citigroup, Morgan Stanley, Goldman Sachs, BlackRock, and Johnson & Johnson set to report earnings.

Over the next seven days, 62 companies in the S&P 500 will release earnings, accounting for 14% of market capitalization, with a concentration in the financial and industrial sectors. J.P. Morgan notes that financials and industrials are key focus areas as earnings season begins. Retail investor positioning changes in the financial sector are worth monitoring.

JPMorgan believes that retail investor flows remain weak, technology ETFs are being sold off, but individual semiconductor stocks are still being bought. This contrasts with institutional positioning, as semiconductor congestion is at historical highs. The earnings season is focused on financials and industrials, with retail positioning changes serving as a key point for ongoing observation.

Disclaimer

This article is a compilation and interpretation by Chaoxiang Research of a third-party brokerage research report (J.P. Morgan, October 14, 2026), combined with publicly available market information. The ratings, price targets, earnings forecasts, and related judgments cited herein reflect the views of the brokerage’s analysts and represent only the position of their respective institution; they do not reflect the views of Chaoxiang Research nor constitute any investment advice.

The market carries risks; make decisions independently. This article should not be used as a basis for buying or selling any securities.