GOOGL, META, AMZN, CMG all print after the close. I've been running the numbers on this cluster for the past few days. The volatility picture is genuinely interesting because of where the mispricings are, and why they exist.
Earnings are the most violent binary event in options trading. The volatility surface completely distorts. IV inflates on the front expiries, skew steepens, term structure flattens or inverts. The 3D surface looks nothing like a normal trading day. And after the print, all of that unwinds.
Most traders try to guess the direction. That's the wrong game. My edge is in reading the 3D surface. So here's what I see tonight across these four names, and what I'm doing about it.
- GOOGL Earnings Jade Lizard
Sell 310 Put, Sell 350 Call, Buy 355 Call, net credit $610, Probability of Profit 84%.
Let's start with the most important number in this entire cluster: GOOGL's six-quarter realized/implied ratio is 0.49. Average realized move 2.64%, average implied 5.36%. Realized exceeded implied exactly once out of six earnings.
The options market is consistently paying double what the stock actually delivers. What makes this genuinely strange is that the fundamental picture is strong. Cloud growing 61% year-over-year. Search re-accelerating; the last print was a clean beat (EPS 2.82 vs 2.58), but stock ended essentially flat. The market priced a 5%+ move, got nothing.
What's happening is that the AI capex narrative is functioning as a permanent fear tax on GOOGL's volatility surface. The story keeps options elevated. The stock keeps ignoring it. That wedge is the edge. Tonight's implied move is 4.67%; still rich versus a six-quarter realized average of 2.64%.
My Earnings Jade Lizard fits the situation for a specific reason. The real risk in GOOGL tonight is actually to the upside. If the Street decides to finally bless the Cloud backlog and Gemini monetization story, you could see a genuine upside re-rating. The call spread eliminates that completely.
The 50 DTE means I'm not playing the binary itself; I'm using the elevated pre-earnings surface to collect credit and letting theta do the work over the next month.
2. AMZN Micro Calendar (Small Account-Friendly)
Sell 265 Call 5/01 (2d), Buy 265 Call 5/04 (5d), net debit $0.40.
AMZN is a different problem. Realized/implied over six quarters: 0.97. The market prices this stock correctly on earnings day. No structural overpricing to harvest with an outright short volatility position.
But look at the volatility surface tonight and something else becomes visible. The 5/01 expiry (two days out, capturing the earnings event) is massively inflated relative to the 5/04 expiry five days out. This is the classic pre-earnings term structure dislocation. The front expiry holds disproportionate event premium. The second expiry retains more proportional time value.
After the print tomorrow, front-week vol collapses violently. The short leg decays faster than the long leg. That differential is what I'm capturing. Not direction, not outright volatility level, just the shape of the surface normalizing after the event passes.
$40 net debit per spread - that's the total capital at risk if AMZN gaps far from the 265 strike.
This is the trade structure I've covered before in the context of 3D surface trading; you're not betting on what the stock does. You're betting on what the surface does after the binary resolves.
3. META Earnings Calendar
Sell 670 Call 5/01 (2d), Buy 670 Call 5/04 (5d), net debit $1.45, theoretical theta: $317/day.
Same structure as AMZN, but very different underlying situation.
META's realized/implied sits at 1.17, modestly above fair. So the outright short volatility case is marginal here, unlike GOOGL where the structural overpricing is obvious. And the narrative risk tonight is genuinely two-tailed in a way that makes a directional volatility bet uncomfortable: upside if AI ad tools and Advantage+ monetization are seen as justifying the $115-135B 2026 capex; downside if the market refocuses on Reality Labs burn and the multi-year AI infrastructure spend with uncertain payback.
The calendar sidesteps the whole problem. I'm not expressing a view on where META goes. I'm reading the 3D surface and seeing exactly the same front-expiry inflation that I see in AMZN; earnings premium bloating the 5/01 strike relative to 5/04. That dislocation resolves the same way after the print regardless of direction.
The higher theta reflects the larger absolute premium in META. More premium in the surface, more to capture. Stock pins near 670: strong P&L from volatility crush. Capped loss on the net debit.
4. CMG Short Straddle
Sell 32.8 straddle, max profit $466, POP 56%, BP $1,146.
CMG is the structurally different trade in this set, and I want to be honest about why. Realized/implied over six quarters: 1.01. The market prices CMG correctly. I'm not claiming an edge from overpriced options. The surface isn't distorted the same way GOOGL's is.
The absolute premium is large; average earnings move of 7.5% in a stock trading in the low 30s, and 50 DTE gives the straddle enough time to work even if the stock gaps on tonight's binary. Management guided to roughly flat comps for 2026 with margin pressure from input cost inflation. The uncertainty is real, which is why realized and implied are in alignment. But selling fair-value premium with 50 days of theta working for you is still a positive expected value trade, properly sized.
The Through-Line
The AI capex narrative is functioning as a fear tax across the Big Tech volatility surface tonight. But that tax is not equally mispriced everywhere.
At GOOGL, the tax is at maximum distortion; implied is pricing double the realized move, quarter after quarter. That's where I'm taking a more aggressive short-volatility position.
At META and AMZN, the tax is closer to fair, so I'm not selling outright volatility, I'm trading the term structure dislocation that the binary event itself creates on the surface.
At CMG, the surface is correctly priced. I'm getting paid fair value for real jump risk. That's enough (with strict sizing).