FEATURED

Retail Options Hit $7 Billion a Day

Retail investors traded approximately $6.8 billion in options premium per day in June 2026. Not monthly. Per day. That number was $5.8 billion in May, which was already a record. Nine of the ten largest retail trading days ever recorded on Citadel Securities’ platform occurred in just the past month, including seven in the first half of June alone.

These are not abstractions. This is a structural change in how markets work.

And the part that deserves more attention than it’s getting: nearly half of all retail options volume is now in contracts expiring the same day they’re traded. Zero-days-to-expiration, up from 30% in 2025 and just 13% in 2021. The average time to expiry on the Citadel Securities retail platform is under three days.

What’s Actually Driving This

Semiconductors. Almost entirely.

In June alone, retail traders allocated approximately $1.9 billion in semiconductor options premium per day — six times the historical average — with roughly 75% of that activity concentrated in call options. For context on how fast this moved: semiconductor daily options premium was $1.6 billion in May, which itself more than doubled April levels and surpassed the prior record by over 25%. The move from April to June happened in eight weeks.

The individual stock gains have been historically extreme. Year-to-date through the first half of 2026, Micron is up roughly 325%. SanDisk has delivered close to 900% gains in the same period. Those are returns most investors see over decades, not months. On May 26, Micron became the latest U.S. company to reach a $1 trillion market capitalization, driven by surging demand for its high-bandwidth memory chips.

What’s different about this cycle is that retail isn’t chasing speculative corners of the market. Today’s retail investor is concentrated in the same companies driving institutional positioning and benchmark performance. That’s new. And it creates a feedback loop that didn’t exist when retail was piling into meme stocks or SPACs.

The Leverage Layer

Leveraged ETF assets reached a record $218 billion, more than 4.5 times their levels from June 2020. Since the end of March alone, assets increased by roughly $82 billion. Technology-linked leverage grew 136% over that period, and semiconductor-linked leverage nearly tripled, up 175%.

This is mechanically significant. Leveraged ETFs rebalance daily. When semiconductor stocks are up, the rebalancing creates forced buying near the close. When they’re down, forced selling. At $218 billion in assets concentrated in the same names retail is already calling, the daily rebalancing effect is no longer a footnote. It’s a structural amplifier operating in both directions.

Worth noting as a slight tangent: the Roundhill Memory ETF (DRAM), launched in early April 2026, became the fastest ETF in history to reach $10 billion in assets under management, doing so in just 43 days. That tells you something about the appetite.

The Hidden Risk

Here’s what’s being underestimated.

A retail investor who was 5% allocated to semiconductors six months ago might now be sitting at 15% to 20%, purely from price appreciation and additional buying. That concentration isn’t intentional — it happened passively. And it means any meaningful correction in the sector doesn’t just hit speculative positions. It hits broad household wealth in a way that previous tech selloffs didn’t, because the ownership is deeper and more diffuse than ever before.

Semiconductor companies now represent a record 19.7% of the S&P 500 — almost four times their weighting of around 5% in June 2020, according to Citadel Securities. Before the dot-com crash, semiconductors represented just over 8% of the index, less than half their current share. The ten largest companies in the index account for nearly 40% of it. Roughly 18 cents of every dollar allocated to the S&P 500 flows into semiconductor companies.

The average 3-month implied volatility of the ten largest semiconductor companies has more than doubled over the past decade, rising from 32% in 2016 to nearly 72% today. The sector trades like a duration asset now, because so much of the cash flow is back-loaded into the AI capital spending cycle. If there’s a real hawkish surprise from the Federal Reserve or a credit event in private AI infrastructure, the sector moves fast — not because the demand is wrong, but because the multiple is sensitive to the discount rate.

The Corporate Bid Underneath It All

There’s a counterforce worth acknowledging. Year-to-date, U.S. corporates have authorized more than $925 billion of share repurchases — the strongest pace ever recorded through this point in the year. Technology and financials account for roughly 57% of those buybacks. That’s a persistent bid underneath the same names retail and passive funds are already crowded into.

But buybacks are not infinite. And a portion of the headline EPS growth investors are using to justify current multiples is mechanically buyback-driven, not organic. Investors reading total EPS growth as evidence of business momentum may be overestimating the underlying revenue and operating leverage.

What Comes Next

Q2 earnings season begins mid-July. It is the first real stress test for a sector that has seen extraordinary gains in the first half of 2026. Expectations are already high — J.P. Morgan’s asset management team noted that 2026 earnings-per-share estimates for U.S. semiconductor names have been revised up 22% since January. Any disappointment in guidance — particularly on hyperscaler capex continuation or memory pricing — lands in a market structure where 0DTE options amplify moves and leveraged ETF rebalancing accelerates them.

Retail participation historically remains elevated through July — the second most active month of the year by net capital deployment, according to Citadel Securities data. That’s a tailwind if earnings confirm the growth. It’s a crowded exit if they don’t.

The question isn’t whether the AI semiconductor story is real. It is. The question is whether the current price already reflects everything that’s real, and then some.

Watch what happens to implied volatility after the first major earnings miss in the sector. That will tell you more about the next six months than any macro forecast.