May 20 is the date. After the close.

And what’s funny (not funny) is how calm the options market can look when everyone knows it’s a live wire. Nvidia earnings tends to do that thing where the narrative is “historic numbers” and the stock response is… fine. Slightly disappointing. Then everyone who paid up for premium spends the next morning staring at IV crush like it’s a personal insult.

As I’m writing this, NVDA is hanging around the $200 area, about 8% off the 52-week high (~$216.83). Big cap, big expectations, still not “cheap” on trailing metrics (around 40.7x), but also not priced like a dying hardware company. So the setup isn’t about valuation. It’s about reaction.

Here’s the thing: the market is pretty good at pricing the number. It’s worse at pricing the mood.

Quick rewind. The last fiscal Q4 (reported late February) was ridiculous in the literal sense: $68.1B revenue (up 20% sequentially, 73% YoY). Full fiscal year revenue was $215.9B (up 65%), and non-GAAP EPS for the year was $4.77. That’s not “beat and raise.” That’s a different category.

Then management guided for roughly 77% revenue growth in the current quarter. Consensus is sitting around ~79%, which is basically Wall Street saying: “Cool, but do the thing you always do and beat the guide anyway.” That’s the bar now. It’s a moving bar, and it’s higher than most people admit out loud.

Also: hyperscalers were just over 50% of data center revenue last quarter. The combined capex track for Alphabet/Amazon/Meta/Microsoft has been floating around $700B for the year. That’s the demand pool. NVDA is still the cleanest expression of it, whether we like how crowded the trade is or not.

The part people skip: implied vs what actually happens

NVDA’s average 1-day realized move around earnings over the last five reports is about 6.32%.

And the last four one-day reactions: -5.46%, -3.15%, -0.79%, +3.25%. So if you’ve been buying the “it’s NVDA, it has to swing” story, the tape has been quietly taking money from you. Over and over. Not because the company is weak—because expectations are already inhaled.

Slight tangent, but it matters: remember the February 2025 “DeepSeek kills GPU demand” panic? That was the last time I saw people truly lean into a bearish narrative that felt sticky. The stock dropped hard post-earnings that cycle (about -8.48%). And then… inference needs kept growing. The “demand goes away” trade has had a rough life since.

So going into May 20, the question isn’t “Will they beat?” The question is “How much do you have to beat by to get paid for owning volatility?” Because if IV is elevated into the event (and it usually is), you’re not just betting on direction. You’re betting on a reaction that clears the premium. That’s harder than it sounds.

Where I’m at on structures (not a “checklist”, just how I’m framing it)

If you’re bullish but don’t love paying full freight into an IV spike, I keep coming back to a defined-risk call spread in the first full week after earnings (the May 28 expiration). Something like the $210/$230 call spread is at least the right shape: you’re expressing upside while limiting how much IV drag can ruin your day.

If you think we get the same movie again (strong print, “good but not good enough” reaction), then a defined-risk premium sell is the cleaner expression. Short straddle logic, but with guardrails: an iron condor centered near spot with wings beyond roughly that ~6–7% realized neighborhood. You’re basically saying: “I don’t need to be a hero, I just need implied to be too expensive.”

What matters is… the bear risk isn’t just “macro” or “valuation.” It’s guidance language. Supply. Constraints. Any hint that the bottlenecks migrate from gaming into data center commentary, or any tone shift that makes the next-quarter bar feel even slightly less automatic, and the reaction can get sharp fast. That’s the part I’m not casual about.

• Date check: May 20 after the close (binary event, no ambiguity)
• The day-before habit: price the straddle the morning of May 19 and translate it into % of spot
• IV reality check: if IV percentile is up in the 70s+, I start leaning “sell structures” unless I have a very specific directional view
• Watchlist between now and then: hyperscaler capex chatter + any sudden “spend digestion” talk (that’s usually where the air pocket starts)

And yeah, this is the annoying part: the market can be right about the move and still make most people wrong about the trade.

If you want, I can map the exact implied move math (and what wings would look like around it) once we see where spot and the front-week IV settle—because that’s when it stops being theory and turns into a trade you can actually live with.

May 20 is close enough that the pricing will start to get honest. Or at least… honest-ish. We’ll see.