The World of 0DTE: When the Delta Hits the Fan

The World of 0DTE

I don't think it's necessarily about the futures being overbought or oversold as much as I think it's options-related; they all play an intricate part of the price action. Truth be told, more than 3 million daily trades in S&P 500 index option contracts expire on the same day, commonly known as 0DTE.

In addition to all the algorithmic, high-frequency, and AI bot programs, the 0DTE options have been a major part of the volatility and quickly became the new fad for option traders that want to make 1-day bets or even the last 20 minutes of the day bets.

We always think the big firms like Goldman Sachs, JPMorgan, B of A, and a host of other big banks and large proprietary trading firms have the better seats and are always on the right side of the trade. Back on August 16th, 2023, Bloomberg reported: “Goldman Sachs Blames 0DTE Options for Fueling S&P 500 Selloff.”

Goldman Managing Director Scott Rubner argued that 0DTE options helped accelerate an S&P 500 decline of roughly 0.4% in about 20 minutes on August 15. I have never forgotten that story, and I wanted to post the original Bloomberg story, but it's still under the paywall, so I had to piece it together. Here is a detailed recap:

Goldman Sachs Managing Director Scott Rubner argued that 0DTE SPX options contributed to a sudden acceleration in the S&P 500 selloff on August 15, 2023, when the index dropped approximately 0.4% in roughly 20 minutes late in the session.

The key contract was the SPXW Aug. 15 4440 put. Nearly 99,000 contracts traded, representing approximately $45 billion in notional value.

Around 3:18 p.m., the option cost roughly $0.70 and had about 10% delta. By approximately 3:40 p.m., the option had exploded to roughly $9, while its delta jumped toward 80%.

That rapid delta change was the problem Goldman highlighted. Market makers on the opposite side needed to rapidly adjust their hedges, potentially requiring substantial selling of equities and futures as the market declined.

Rubner's central warning was:

He was specifically referring to such a dramatic delta change occurring within approximately 20 minutes.

Goldman Also Pointed to Deteriorating Liquidity

Rubner estimated that a measure of market liquidity had fallen 56% over the preceding two weeks.

That matters because the same amount of dealer hedging can move the S&P substantially more when order-book depth is thin.

Other Wall Street firms were seeing similar activity. Nomura and Citigroup noted that 0DTE trading had reached record levels that month, with traders increasingly using puts.

UBS examined July 27, August 4, and August 10 and found that large 0DTE put flows coincided with significant intraday market moves.

Rubner also warned that rising volatility and weakening momentum could trigger additional selling from systematic/rules-based strategies.

The S&P had just closed below its 50-day moving average for the first time since March, and its recent intraday range had expanded substantially compared with July.

His conclusion was notably bearish:

The Important Trading Takeaway

What Goldman was describing is essentially:

0DTE put buying → market declines → put delta explodes → dealers adjust hedges → additional ES/equity selling → market falls further → delta increases again → more hedging.

That's the gamma/delta feedback mechanism that can turn an ordinary decline into a very fast intraday move when liquidity is thin.

There was also an important sequel: Bank of America challenged Goldman's explanation about a week later, arguing that 0DTE hedging flows may not actually have caused the August 15 decline.

The Goldman-vs.-BofA disagreement centered on a critical question that remains relevant today: Can massive 0DTE options activity amplify intraday S&P 500 volatility when dealer hedging collides with thin market liquidity?

I did write about this when the story came out, and I am as awe-struck now as I was then. I just could not get over Goldman complaining about what I believe were losses that day.

I have a simple way of looking at this: when the big banks report record trading profits and they are the ones that help make a lot of the prices people trade off of, you don't hear the smaller shops or traders saying they got screwed... do you?

And when the ES and NQ drop or rally 2% or 3% in a day, or a stock rallies or drops 10% or 15% after a company reports earnings, you don't hear Joe Public saying he got screwed... and you know why? Because no one would listen, that's why!

There has been a lot of talk about September being the weakest month of the year, and I get it. Last week, there was a post on Twitter from a mainly unknown person that got over 750k views, and the data was incorrect. But before we get to September, we still have the last trading day of August to get past.

If you are looking at the last trading day of August, when it falls on a Monday and in a midterm election year, it has a bearish lean and potentially elevated volatility with more down than up, with 1998 as an extreme example, down -6.8%. Out of the last 9 occasions, the ES has been down 5 / up 4, but 2 of the up days were only up +0.01% and +0.04%.

Our lean: Out of the last 13 Mondays, 6 have closed higher, and 7 have been down. That works out to 46.2% higher and 53.8% lower, and if you take the last 6 Mondays, it's 4 down, 2 up, with one only being up 0.02%, or 66.7% bearish and 33.3% bullish.

Yes, we had a nice early rally on Friday, but 1) the futures rejected the 7750 level and 2) the ES closed down on the day.

Like always, I want to remain an optimist, but there just isn't a lot of good news out there. It's the same deal: I don't doubt the ES and NQ can bounce, but I don't think they can hold. I can’t rule out buying some early drops, but I think the money trade is selling the rips.

MOC Recap: Nasdaq Leads a Powerful Buy Program

Friday’s MOC opened with the imbalance surged to $4.9 billion to buy, with $6.9 billion of buy orders against $2.0 billion of sells. The dollar lean jumped to +77.7%, a clear wholesale buy signal, while the symbol lean was only +54.7%, showing that the strength was concentrated in larger-cap names rather than uniformly spread across the market.

That concentration was most obvious in the Nasdaq. The Nasdaq carried a $4.8 billion net buy with a +92.8% dollar lean and +79.2% symbol lean, a powerful, broad-based wholesale buy. The S&P 500 was also heavily bought at $4.8 billion with a +79.3% dollar lean, although its +57.9% symbol lean showed more rotation beneath the surface. The NYSE was nearly balanced, with both dollar and symbol leans close to 50.0%.

Information Technology dominated the sector flow, posting a $3.6 billion net buy with +92.9% dollar and +68.3% symbol leans. Consumer Staples also reached wholesale-buy territory at +73.7% dollars and +66.7% symbols, while Consumer Discretionary showed a strong +85.4% dollar lean. Materials stood out on the sell side with an -81.7% dollar lean, though its -53.3% symbol reading suggested the selling was concentrated rather than sector-wide.

The biggest buy imbalances were NVDA at $818.7 million, MSFT $418.1 million, AAPL $327.6 million, MU $278.8 million, AMZN $246.3 million, and AMAT $221.6 million. Major sells included MSTR, BRK.B, CRM, SHW, APD, FANG, PG, GOOGL, PFE, HLT, and TJX.

After the 3:51 spike, the imbalance steadily faded, briefly flipping to a $557.0 million sell at 3:55 before recovering. It finished at 4:00 with a $720.0 million buy and a +75.0% dollar lean, confirming that the close retained a strong buy bias even after substantial late rotation.

Over the last 10 trading days, the Daily Direction of Imbalance shows a clear pattern of net accumulation. There were 7 buy-imbalance days versus just 3 sell-imbalance days, producing an approximate cumulative net buy of $5.2 billion across the period. The strongest shift came late in the window: after modest and mixed flows from August 17 through August 21, buying accelerated sharply with roughly $1.6 billion on August 24, $1.2 billion on August 25, and $1.7 billion on August 26. Although August 27 interrupted the streak with about a $1.0 billion sell imbalance, August 28 reversed that quickly with the largest buy imbalance of the 10-day period at roughly $2.2 billion. In total, four of the final five sessions were buy days, generating approximately $5.6 billion of net buying during that stretch. The pattern points to sustained institutional accumulation rather than simple day-to-day rotation, with increasingly larger buy programs dominating the back half of the 10-day period.

How to read the MIM

How the NYSE and Nasdaq Auctions Work
How the NYSE and Nasdaq Auctions Work

BTS Levels - (Premium Only)

Today’s Important Economic Events

Polaris Trading Group Summary Friday, August 28, 2026

Friday was a Cycle Day 1 session shaped by Fed Chair Warsh’s Jackson Hole remarks, an early upside auction, and a later reversal back toward the key 7730 Line in the Sand. The day became especially valuable for BLT recognition, identifying failed continuation, and understanding acceptance versus rejection at important levels.

Pre-Market Game Plan

• David identified 7730 as the key Line in the Sand.

• Overnight price was auctioning higher toward the 7745–7755 target zone.

• Additional upside targets were identified at 7765 through the 7771 Cycle Day 1 objective.

• The room was also aware that Warsh’s Jackson Hole speech could create added volatility.

Morning Price Action

• The market initially followed the bullish roadmap and continued its auction higher.

• As price extended, the focus shifted from continuation to signs of rejection and exhaustion.

• Several traders began identifying short-side opportunities as momentum weakened.

• A short around 7740.25 was reported during the morning session.

• The changing market structure reinforced the importance of reacting to price rather than remaining committed to the initial directional bias.

BLT Opportunities

• BLT setups became the dominant educational theme of the session.

• Multiple BLTs appeared throughout the morning.

• Slatitude39 reported a BLT short near 7776 that produced a quick move in roughly two minutes.

• David later highlighted another BLT forming around a double-top structure.

• The repeated BLT opportunities led David to strongly reinforce the pattern as one traders should learn to recognize instinctively.

• The session provided several useful examples of BLTs developing near exhaustion and failed continuation areas.

Positive Trades and Room Highlights

• The early roadmap successfully identified important upside zones at 7745–7755 and 7765–7771.

• Traders who recognized rejection after the morning extension found quality short opportunities.

• The BLT short near 7776 was one of the cleaner examples discussed in the room.

• DanV reported a DLMB “grand slam” during the downside move.

• Other members were congratulated for successfully participating in the sharp reversal.

• John was also recognized for doing well with his BLT setups.

Return to the Line in the Sand

• By lunchtime, the market shifted into a consolidation phase.

• David noted that price was hovering around the original 7730 Line in the Sand.

• The return to 7730 was significant because the market had traveled well above that level earlier in the session.

• This demonstrated how an important reference level can remain relevant even after a substantial directional move.

Afternoon Structure

• David highlighted the importance of understanding the best and worst times of day to trade.

• Bear flags became a key structural observation during the afternoon.

• David questioned whether the session might finish with a “Warsh Flush-out.”

• The final trading day for August T+1 fund activity added another element to the afternoon environment.

• The later session favored patience and selectivity rather than forcing additional trades.

Lessons Learned

• Recognition is more important than prediction.

• The market initially followed the bullish roadmap, but traders needed to recognize when continuation began failing.

• A directional bias should never prevent a trader from responding to changing market structure.

• Acceptance and rejection at specific levels remain critical concepts.

• A price level by itself is not a trade; the market’s behavior around that level determines whether continuation or reversal is more likely.

• BLTs are especially valuable when they appear near exhaustion, double tops, and failed continuation areas.

• Time of day matters. The morning produced stronger directional opportunities, while lunchtime brought consolidation and the afternoon required greater selectivity.

• Knowing when not to trade is an important part of maintaining discipline.