How Markets Work

the edges empires are built on

from probability to edge

Kris Abdelmessih

Mar 2, 2025 — 5 min read

I cleaned up a thread I wrote before NVDA reported earnings Wednesday afternoon. It conveys what I was thinking as I looked at the option surface for the Friday 2/28 expiry.

Thinking aloud as I'm looking at NVDA options before earnings.

My instincts — given the January sell off in NVDA that the stock probably has some discount in the price going into earnings. In other words, if earnings are a nothingburger I’d expect a small rally. No news is good news.

Narrator: I start by announcing my bias. Whether it’s dumb or not is irrelevant. Before you look at a price, chart, option surface you probably have an expectation of what your eyes will see. I probably had many biases. But that one was the most salient before I looked at the option surface.

continuing…

Translating into option speak -- the stock is probably going up in terms of “hit rate”.

When I look at the option market, it confirms my bias (not necessarily my reason, but I’m not sure I’ve ever known a reason for anything that wasn’t as tautological as “more buyers than sellers”).

What makes me think the market confirms my “it’s probably going higher” bias?

The downside skew for earnings is fat and the 130/135 call spread is expensive. The market communicates via prices because prices offer odds. You disagree with the market by saying “buy” or “sell”.

How does the price of a call spread tell me anything about the odds?

Start with a simple proposition.

If NVDA was 50/50 to go up or down, you'd expect a $5 wide call spread to trade for $2.50 with the stock at $232.50 which is the midpoint of the strikes.

The stock was $129.50 and the 130/135 call spread was priced around $2.20.

Hmm. Antennae is up.

The call spread has a .15 delta. Meaning if we shift the stock $3 lower from $132.5 to the current $129.50 price, I expect the call spread goes from about $2.50 to $2.05 ( a change of $3 x .15)

So the call spread looks about $.15 rich compared to coin flip pricing.

How do we contextualize how if $.15 in terms of edge?

Well, think go back for a moment and think of the 50/50 case. If the stock was at the midpoint you expect the call spread to be worth $2.50. But if it’s $.15 rich or $2.65, if you sell it what kind of odds are you getting?

You are risking $2.35 to make $2.65 or getting 1.13-1 odds. What does a bookie give on an even money bet? 110-100, right. Empires are built on less edge. $.15 is a lot of edge on a $5 wide call spread.

[Personal opinion interlude: I think this was the biggest insight SIG conveyed to us when I started in the biz. Markets were wide when I was in training and you could get that kind of edge on benign things like call spread. SIG had a reputation for trading big. Using their capital, we were told to get as many lots as we could on every trade. You’re job was to fight for large allocations or splits. The edge compared to the risk on these things was much larger than the edge they saw gambling operations compound on so they optimized for market share. Many less-capitalized firms wanted to maximize edge per trade, which meant they were reluctant to tighten for market share, which can be the right move if you’re undercapitalized.

We were trained to be pigs because they understood that the getting was way too good to last so you want to maximize p/l, area under the curve. So you often pissed everyone off. Our nickname on the floor was “evil empire”*. When you start out you’re reassured “oh, it’s because we wear the black smocks”. But you learn that it’s because we acted like Amazon in a world that still had lots of indie bookstores. You can level whatever you want against Amazon, and they’ll just say “it’s better for the consumer”. Of course what they also mean is — “scoreboard”.

E-commerce and trading are bloodsports. As you learn in trading, in 5 years, you’d kick a grandma down the stairs for the thin margins you’re complaining about today. If you’re young and reading this, the cycle I’m describing is evergreen. There is someone willing to physically fight for your job if you are in a seat with an edge. That you don’t see a fist swinging at you doesn’t mean you’re not in a conflict. The zero-sumness of prop trading is the animating force behind its evolution. If that sounds rough perhaps you’d prefer a career in asset management where “solutions” can be customized to a client’s frontier. Also, if what I’m describing makes no sense to you or sounds dramatic…you’re not in trading.]

Back to the NVDA call spread…

What was my thinking and what did I do?

I didn’t trade it.

I did not sell the call spread that looked expensive. The vol lens said it was expensive, but the prior I had, “the stock is probably going a bit higher” was baked into the option market. If my prior no news is bad news, then I would have sold them.

[Note that the prior is playing a real role here and it was based on vibes. This is a complicated matter. On the one hand, I know why I give weights to my prior. It’s an old habit, that was justified. Was.

In a trading seat, you ingest a lot of unstructured flow data. Which brokers bid or offered a particular option in size? Maybe they passed. Maybe they checked another strike. Why that one? How aggressive were they, have I seen the signature of this flow before? You’re always playing this little deduction game and its output becomes the priors that influence positioning.

Today, I don’t get the same context. It makes me more hesitant. It makes me give more weight to implied odds versus my prior. In fact, with a liquid name like NVDA there is so much electronic flow that perhaps the option surface odds simply reflect a risk premium. Maybe I shouldn’t read too much into the implied odds of earnings call spread and just accept it as the price to clear a flow imbalance as opposed to anything deserving predictive weight.]

Finally, I believe the contrarian trade is probably to spec some way topside call or call spread because that's the most discounted part of earnings. Probably for good reason, but that also means it’s the more “destabilizing” move. You only get nitroglycerine if everyone is offsides.

The main lesson:

Unlike myself, many of you have stronger conviction stock opinions, so I hope marrying it with what the options market is offering allows you to better pinpoint how you should express them.

🔗Further reading to that ends