Alphatica

SPX GEX LEVELS: May 6 Yesterday's post said the gamma was decaying and the flip was 13 points from spot. Today the market responded by rebuilding the entire structure in a single session. 🚨🚨🚨Tomorrow's expected range: 7,200 – 7,325. Bias higher toward the 7,300 magnet. Net GEX snapped back 45%, from +$787M to +$1.14B. The flip widened from 13 points to a cluster that tops out at 7,245, still close to spot but no longer sitting on the regime threshold. Call buying dominated the session. 7 of the top 10 volume GEX strikes were positive, concentrated between 7,275 and 7,350. The heaviest flow hit 7,300 (+$88M), 7,290 (+$37M), and 7,350 (+$34M). That's directional conviction above spot. SPX closed at 7,259.23, 41 points from the max magnet at 7,300 which carries +$197M. That's the strongest single-strike reading since the April pinning regime. The magnets above ladder cleanly: 7,350 (+$117M), 7,400 (+$117M), 7,500 (+$71M). Every strike from 7,200 to 7,600 is positive gamma. The path to 7,300 and beyond is paved with stabilizing force. The 7,000 strike flipped negative for the first time in the series. Net GEX contribution is -$6M. Put OI (1.038M) exceeds call OI (993K) by 45K contracts. The level that anchored the March-to-April rally, carried +$308M at the peak, and acted as the gravitational center for three weeks is now working against price. That's the structural migration, the market's center of gravity moved from 7,000 to 7,300 over six weeks. $AMD reported after the close. $1.37 EPS vs $1.29 street ($1.33 Alphatica estimate). $10.25B revenue vs $9.89B street ($10.0B Alphatica estimate). Revenue +38% YoY. Data center +57%. 14 straight quarters without a miss. Our Alphatica Earnings Quality Signal flagged AMD 4/6 BULLISH heading in. The data was telling the truth. The gamma decay schedule is still active, another 15% of the blanket expires by Friday, and monthly OPEX on May 15 takes the structure down further. But today showed that one session of aggressive call buying can rebuild what took three sessions to drain. The structure is more reactive in both directions now. That's the new normal. $SPX $SPY $QQQ

"Low volume rallies are suspicious." No they're not. We tested it. 22 years of SPY and QQQ data. Every day where volume dropped below 50% of the 50-day average — the quietest of the quiet. 268 episodes. Forward 10-day win rate: SPY: 70.6% QQQ: 68.2% Forward 20-day win rate: SPY: 65.4% QQQ: 69.7% 7 out of 10 low-volume days, the market was higher two weeks later. Is the edge better than just being long? No — baseline win rate is 62%, so the difference isn't statistically significant (p = 0.21). But the claim isn't that low volume is bullish. The claim is that "low volume is suspicious" and bearish. The data says the opposite. Low volume days go up at a slightly higher rate than normal days, not lower. The logic is simple: If volume is low and price isn't dropping, sellers aren't showing up. That's not a fake rally. That's a market with no supply. Path of least resistance = up. "35M in volume today. DEAD." Dead volume, alive market. That's what 22 years of data shows. $SPY $QQQ

AMD reports today after the close. 4/6 on our Alphatica Earnings Quality Score. +240% in 52 weeks. 13 consecutive quarters without a miss. The AI GPU story is accelerating and the street knows it. Street EPS: $1.29 (Non-GAAP) Street Revenue: $9.89B Alphatica Estimates: EPS: $1.33 Revenue: $10.0B OUR CALL: BEAT ✅ AMD hasn't missed in 13 straight quarters. Last quarter was a blowout: $1.53 non-GAAP EPS on a $1.32 estimate, a 16% surprise. Revenue hit a record $10.3B. Management guided Q1 at ~$9.8B plus or minus $300M with ~55% non-GAAP gross margin. The street consensus at $9.89B sits inside the guide range but slightly above the midpoint. The question: does the 16% beat pattern repeat, or was Q4 the peak? OUR FOUR SIGNALS: Earnings Quality: 4/6 BULLISH. One of the strongest VA release trends in the semiconductor database. VA% dropped from 45.7% to 24.4% in three years, with an $798M absolute release in FY2025, the largest in AMD history. Revenue +53% in two years. Operating income nearly 10x from $0.40B to $3.69B. OCF $7.71B running 1.78x net income. Two neutrals: profitability needs one more year of sustained high earnings to confirm, and SBC at 4.7% of revenue is elevated but within semiconductor norms. The accounting and the business are pointing in the same direction. Disclosure Signal: NEUTRAL. LOW distance (0.118) with NEUTRAL tone (-0.014). AMD's multi-year tone improvement from -0.019 to -0.012 has stalled. Two consecutive years of slight deterioration back to -0.014. Not enough to alarm but the positive trend is broken. Risk factors barely changed (0.001 cosine distance). Quarterly filings stable. No triggers in either direction. Prediction Model: BEAT (Medium Confidence). The quantitative model is not validated for AMD due to historical EPS volatility around zero. But the directional indicators favor a beat: 13 consecutive non-negative surprises, conservative YoY estimate of $1.33 above the $1.29 consensus, and a 75% beat rate over the last 4 quarters. LOW CONF because the model itself can't statistically confirm, and Q1 is seasonally AMD's weakest quarter. Earnings Risk Signal: ±6.7% NEUTRAL. EVRP at 49.57 indicates options are pricing a significant move relative to recent volatility. ATM IV at 115%. IV spread at +1.87, calls slightly above puts but essentially balanced. Avg PMIE over 3 years is 7.04% with a max of 17.3% (last quarter). This stock can move $23 in either direction tonight. THE AI INFLECTION: Lisa Su laid out the thesis on the Q4 call: data center segment revenue growing 60%+ annually over the next 3-5 years, AI business scaling to tens of billions in 2027, EPS target of $20+ in the strategic timeframe. MI350 series shipments ramping. OpenAI partnership for 6 gigawatts of Instinct GPUs. MI400/Helios launching this year. EPYC server CPUs hitting record share. The only headwind: semi-custom gaming SoC revenue declining "significant double-digit percentage" in FY2026 as console cycle matures. But data center is more than offsetting that drag. THE RISK: +240% in 52 weeks. Q1 is seasonally the weakest quarter with an expected sequential revenue decline of ~5%. The last time AMD's stock was this extended going into earnings (Q3 2024 at $164), it dropped 28% in the subsequent quarter. The AI capex narrative is consensus now. Every hyperscaler is spending, but AMD still needs to prove MI350/MI400 can take meaningful share from NVDA at scale. And Q4's 16% EPS surprise was partly boosted by a $360M inventory reserve reversal and $390M in China MI308 sales, both of which are non-recurring at that magnitude. $356.29 is the price. $9.8B is the guide. $9.89B is the street. 13 straight is the streak. Tonight is the test. Not Investment Advice. $AMD $NVDA $QQQ $SMH

"Low volume rallies are suspicious." No they're not. We tested it. 22 years of SPY and QQQ data. Every day where volume dropped below 50% of the 50-day average — the quietest of the quiet. 268 episodes. Forward 10-day win rate: SPY: 70.6% QQQ: 68.2% Forward 20-day win rate: SPY: 65.4% QQQ: 69.7% 7 out of 10 low-volume days, the market was higher two weeks later. Is the edge better than just being long? No — baseline win rate is 62%, so the difference isn't statistically significant (p = 0.21). But the claim isn't that low volume is bullish. The claim is that "low volume is suspicious" and bearish. The data says the opposite. Low volume days go up at a slightly higher rate than normal days, not lower. The logic is simple: If volume is low and price isn't dropping, sellers aren't showing up. That's not a fake rally. That's a market with no supply. Path of least resistance = up. "35M in volume today. DEAD." Dead volume, alive market. That's what 22 years of data shows. $SPY $QQQ

SPX GEX LEVELS: May 5 The gamma decay we warned about Saturday is happening in real time. 🚨🚨🚨Tomorrow's expected range: 7,100 – 7,275. Wider than last week's ranges because the gamma is thinner. Below 7,188, the regime flips and the accelerators fire. Net GEX dropped 27% in three sessions, from +$1.07B to +$787M. Friday's OPEX drained a quarter of the gamma blanket. Today's put buying thinned it further. The structure that pinned SPX in 100-point ranges for three weeks is losing its grip. The GEX flip tightened to 7,188, 13 points below the close. That's the second time in a week spot has landed on the regime change threshold. Last time was the FOMC knife edge at 7,138 with one point of cushion. The market is walking the same line again, except now the gamma underneath is 27% lighter. Today's volume tells you where the pressure is coming from. The top 4 negative volume strikes were all below spot: 7,150 (-$51M), 7,100 (-$38M), 7,125 (-$37M), 7,120 (-$30M). That's aggressive put layering within 100 points of spot. The 7,000 strike absorbed -$27M in put buying with 80K puts traded against 19K calls. The zone from 7,000 to 7,175 is being loaded with downside exposure. Above spot, call buying held at 7,250 (+$42M), 7,240 (+$35M), and 7,300 (+$28M). The magnets are still there 7,300 at +$136M is the max magnet, 100 points above. But the path to get there runs through a minefield of freshly loaded puts that didn't exist last week. The 7,000 strike is worth noting again. Put OI (1.023M) now exceeds call OI (992K) by 31K contracts. The GEX contribution is +$3M, functionally zero. The anchor of the entire March-April rally has completely neutralized. Here's where we are structurally. Saturday's alert said the path of least resistance shifted lower. The shooting star was invalidated on today's session but the gamma decay and the flow reversal are both playing out. Net GEX dropped 27%, the flip tightened to 13 points, and put buying loaded the zone below spot. The candle didn't confirm but the structure did. Gamma is still positive. Dealers still buy dips above 7,188. The regime hasn't flipped yet. The question is whether tomorrow's session is the one that pushes through. The decay schedule continues: another 15% of the blanket expires by Friday. By monthly OPEX on May 15, 72.6% of last Thursday's gamma is gone. Each day the structure gets lighter and the flip level tightens. Yields aren't helping. The 30-year crossed 5% today and the 10-year hit 4.44%. Rising rates compress equity multiples and that pressure is hitting while the gamma blanket that suppressed every move for a month is 27% thinner. $SPX $SPY