Narrating An Option Trade

I’ve been narrating my small GME trade this week through this substack and twitter.

On Friday I rolled my short June 20 calls to the 25 strike. I narrated my thinking on twitter but I’ll re-print it here. It’s a combination of real-time thinking and some meta-thoughts about trading as well.

Sharing my monkey thoughts as i mess around in GME… I’m long that 20 lot of June 20/30 call.

Rolling the position

Despite the stock being down today, the 30s are eroding as vol is declining so the spread actually upticked in value.

I’m also looking at the june 20/25 call spread:

The spread value has increased a lot. The vega on these options is small but not totally negligible. Look at the IV spread…it’s fallen 14 points today on a spread with a penny of vega – that’s a 14c rally in the spread on a delta deutral basis! Since the spread is only .23 delta, the vol change has kept the spread little changed despite the stock move.

I decided to roll my short 20s into short 25s.

So I sold the June 20/25 cs. I was filled at $3.71

I was long the June 20/30 cs from $2.08 but collected $3.71 on the 25/30 cs.

=> On balance, I’m left long the 25/30 cs for a $1.63 credit.

Thinking behind the roll

The roll was driven by a sense that the risk-reward on the 20/25 cs at $3.71 is not as great as owning the 25/30 for a m-t-m level of $1.60.

I’m synthetically “selling the 20/25/30 fly” in my reasoning.

This is a mix of seeing the strike vol changes today and feel. This may sound woo woo. If you require higher standards of trade discretion I can understand that but for the most part this is kinda what trading looks like for all the nerdom that gets bandied about.

GME is a name that doesn’t lend itself to data analysis or cross-sectional triangulation.

[Also, these are microscopic stakes —the original trade only risked about $4,200 and was only 1/5 of the size I wanted to scale into. Unfortunately, the call spread went straight up in value after I was filled and I was too anchored to chase]

In a professional setting, you will be more plugged into flow which tightens the reasoning and timing plus better execution tools/costs, but the mental progressions of a MM very much rhyme with mine especially in idio situations.

The value of tracking

After selling the 25/30 call spread, I stored it in my watchlist on a delta-neutral basis vs the stock price reference I sold it against to track its performance and my fill quality.

The market on a delta-neutral basis is $3.52-$3.89, I sold at $3.71.

I noticed as the stock went down my fill marked better and vice versa.

This can be an artifact of market widths in the legs esp since both calls are ITM but if you can rule that out by tracking the counterpart put spread instead (CS + PS = Box so you can translate the 3.71 cs as a 1.29 PS) you can get a fingertip feel for how the skew moves as the stock changes by watching a delta neutral price move. I used to do this for lots of structures. They’d be in my window for weeks so I can see how certain large trades worked or not.

Experience is a repo of unstructured data

Overall, these little habits accumulate as a big unstructured data repo otherwise known as experience. Discretionary trading uses “science” for measuring. models, dashboards, stat studies. But the trading is a boardgame sitting on top of that.

Systematizing

Systematic trading is different. I sat next to people running large systematic strategies. It’s piloting & auto-piloting a more diversified strategy, less chunky risk but it’s not an alien approach. It’s also inspirational because you can port measures and ideas from it

I never really figured out how to systematize my trading from end to end. I like to say discretionary trading is just trading without ignoring unstructured data. The unavoidable pitfall is part of that “unstructured data” is bias however.

You can fight that with a team of people conscious of behavioral biases because while it’s hard to debug yourself it’s easy to call out others’ blindspots. You can help each other. (see Trading Is A Team Sport)

moontower.ai

It’s funny as I’m sandboxing analytics to develop for moontower.ai, I’m always thinking about how to pre-chew ideas for users.

There’s always a balance between legibility and user effort. If investors just hand you money to manage for them, it requires no effort for them but they have no control over the trades or process.

At the other extreme, you can give a user a Bloomberg terminal where they have full control but need to make all the decisions.

For a retail user this trade-off requires tremendous thought. You could just feed someone a signal without teaching them how to fish. When the signals don’t work they’ll churn. Or you can teach them how to approach their goals methodically understanding that there’s a learning curve. It takes more effort for the user but for the person who wants to learn to fish it’s the only way.

moontower.ai balances being opinionated but not signal-driven because it inherits the exact trade prospecting funnels I used as a discretionary PM.

We can see the seeds of signals in our “axe list” concept. It’s just a roadmap idea at this point. But some of the data play I’m doing for this coming week’s paid post looks like it might be a source for another analytic regarding skew.

As a reminder, if you’d like to signup for moontower.ai as a paid user, a moontower.substack paid sub is included.

We also have a fully automated affiliate program. If your audience would be interested in option analytics with a point of view…you can sign up to be an affiliate and get $100 for users who sign up.

$ Become A moontower affiliate

Extra

Arb in ADBE options? (twitter thread)

A Twitter friend thought he might have found an arb in the ADBE options that could be exploited by a trade known as a ‘conversion’. It was a false alarm but if you scroll through the thread you can learn a lot about computing implied rates and the details of how to execute the conversion arbitrages. They don’t sit around in plain sight but this is the masochism section so if you want to learn the thread is a real-life process of noticing something that looks off, getting to the bottom of it, and learning a lot about options in the process.