Issue #5

The options anomaly detector that finds mispriced vol in 3 minutes

A 3-prompt workflow that spots vol mispricing. No Greeks PhD required.

AI Finance Brief 2026-04-13 Free issue

The Vol Anomaly Scanner


Step 1: The Baseline Check (60 seconds)


Pull the current IV rank and IV percentile for a name you are watching. Most brokers show this. Paste it along with the last 30 days of daily price moves, then:


You are a volatility analyst. Here is the current options data for [TICKER]:

- Current IV: [X]%
- IV Rank (52-week): [X]%
- IV Percentile: [X]%
- Last 30 days of daily returns: [PASTE]

Analyze this:

1. REALIZED VS IMPLIED: Calculate the 20-day realized volatility from the daily returns. Compare it to the current IV. Is the market overpricing or underpricing vol?
2. IV RANK CONTEXT: An IV Rank of [X]% means current vol is at the [X]th percentile of the last year. Is this consistent with what the stock is actually doing? (A stock grinding higher on low volume should have low IV. A stock whipsawing daily should have high IV.)
3. SKEW CHECK: If the IV is low but the stock just had a 3% move, that is a potential anomaly. Flag it.
4. CATALYST CALENDAR: Are there earnings, FDA decisions, or other known catalysts within the next 30 days that should be keeping IV elevated?

Give me a verdict: is this name's vol CHEAP, FAIR, or EXPENSIVE relative to what it is actually doing?

Step 2: The Historical Pattern (60 seconds)


Now check the vol pattern after similar setups:

- In the past year, every time [TICKER]'s IV Rank was below [X]% while 20-day realized vol was above [Y]%, what happened to the stock over the next 10-20 days?
- After earnings vol crush, how quickly does IV typically rebuild for this name? (Some stocks stay quiet for weeks. Others reprice within 5 days.)
- Is there a seasonal vol pattern? (Some names consistently see higher vol in specific months due to industry cycles.)

Tell me if this is a "normal post-earnings quiet" or "vol is genuinely mispriced."

Step 3: The Trade Structure (60 seconds)


Based on this analysis, if I wanted to express a view that vol is [cheap/expensive]:

1. LONG VOL: What is the most capital-efficient way to get long vol on this name right now? (Straddle, strangle, calendar spread?) What strikes and expiry optimize for the catalyst timing?
2. SHORT VOL: If vol is expensive, what is the highest probability short vol structure? What is the max risk?
3. RISK/REWARD: For each structure, what is the breakeven, max profit, and max loss?
4. SIZING: If I want to risk no more than [X dollars] on this trade, what is the appropriate position size?

Give me the specific strikes and prices based on current data.



The Edge


Most retail traders buy options based on direction — "I think the stock goes up, so I buy calls." Professional options traders think in volatility — "implied vol is cheap relative to what this stock actually does, so I buy options regardless of direction."


The LLM does not know the future. But it is extremely good at comparing the current vol regime to historical patterns and flagging when the options market is mispricing risk. That comparison is the edge.




A Note on Execution


This workflow gives you the analysis, not the trade execution. Always verify strike prices and current premiums in your broker before executing. Options pricing moves fast and the LLM's analysis is based on the data you provide, not live market data.


The value is in the framework — knowing what to look for and where the mispricing is. The execution is still your job.




What is Coming Next Week


Issue #6: The macro regime detector — one prompt that tells you what economic regime you are in and what it means for your portfolio.




AI Finance Brief is written by a team that runs live algorithmic trading systems daily. Every workflow is tested on real market data before it reaches your inbox.


Free every week. Subscribe at aifinancebrief.com

Get the next issue free

Weekly AI workflows for finance professionals. No fluff, just the prompts and systems that work.

Subscribe at aifinancebrief.com