SOXX Bounced Just 0.36% After a 5.63% Crash and Sandisk Fell for a Fifth Straight Session — Is Selling Pressure Still Building, or Is a Base Forming?

Data Calibration

• All prices are September 15, 2026 (Tuesday) US market close, sourced from Tiger terminal.
• Options data as-of September 15 close.
• This article contains no portfolio positions, share counts, dollar amounts, or trade recommendations.

One-Line Takeaway

SOXX bounced just 0.36% following a 5.63% crash — well short of the ≥2% technical oversold bounce typically expected — while Sandisk fell another 1.36% for its fifth consecutive losing session even as the broader sector stabilized. Selling pressure is differentiating: the sector stopped bleeding, but memory hasn't.

Session Data Overview

Table 1: September 15, 2026 Closing Data

VIX: 17.20 (prior 17.10, essentially flat).

Checking the September 14 Pre-Written Test

The four tiers written in advance on September 14:

Conclusion: Scenario A is confirmed — semiconductor selling pressure did not transmit to the broader market. However, SOXX also failed to produce an oversold bounce, remaining in a narrow consolidation range.

Why a 0.36% Bounce Matters

In a normal market, after a 5.63% single-day crash, an ETF typically sees at least a 2-3% technical rebound the next day (short covering + bottom-fishing capital). SOXX's 0.36% bounce tells us:

1. Shorts are not rushing to cover. If this were simple oversold conditions, shorts would lock in profits on Day 2, pushing prices higher.
2. Buyer conviction is weak. Volume ratio at 0.81x (below 20-day average) indicates bottom-fishers are not active.
3. Waiting for a catalyst. The market may be waiting for a fundamental signal (earnings, industry data) before committing to a direction.

Micron vs. Sandisk: Divergence Within the Same Sector

Table 2: Memory Stocks' Relative Spread vs. SMH

Micron and Sandisk moved in opposite directions on the same day. MU managed a small gain roughly in line with the sector, but SNDK continued weakening.

This divergence carries information: if selling pressure were macro-driven or sector-driven, both stocks should move in the same direction with similar magnitude. One up and one down points to individual stock-level factors — potentially differing expectations for NAND vs. DRAM sub-sectors, or differing institutional position-adjustment timelines.

Sandisk's Losing Streak

Table 3: Sandisk Over Five Trading Sessions

From $1,764.17 on September 9 to $1,530.89 on September 15, Sandisk has fallen 13.22% over five trading sessions. The daily decline rate is narrowing (−4.06% → −3.50% → −4.98% → −1.36%), but the direction has not reversed.

Options Market: Expiration-Week Mechanical Effects

Table 4: September 15 Options Data

With September contracts expiring on 09-18, we are now inside the final three days — OI has shrunk further. MU's PCR (OI) jumped to 1.90, SNDK to 2.79 — but this does not indicate a surge in bearish sentiment.

The reason: during expiration week, out-of-the-money calls are closed first (no exercise value), leaving behind mainly in-the-money or near-the-money puts held as hedges. This mechanically inflates the PCR. We need to wait for next month's contracts to accumulate meaningful OI before the options market can provide a reliable directional signal again.

Broad Market Context: S&P on a Two-Day Losing Streak

SPY declined for two consecutive sessions (−0.45%, −0.46%) — modest moves individually, but directionally consistent. This aligns with VIX holding flat (17.10 → 17.20): no panic, but no buying conviction either.

Pre-Written Test for the Next Session

Data sources: Tiger terminal quotes, CBOE VIX Index. All data in this article is publicly available market information and does not constitute investment advice.