Expensive Isn't a Veto

On July 20th, OptionBench flagged a pre-earnings long strangle on HIMS. Twenty past cycles, eighteen of them profitable — a 90% win rate on a setup with more than five years of history behind it. Worth a look.

So I looked. And my own entry gauges told me to pass.

The relative value of the position ranked 6th most expensive out of 20 past entries at that point in the cycle. Implied volatility sat at 112%, against an average of roughly 89% across previous cycles at the same T-15 mark. Both readings pointed the same way: you are paying up.

I would not have taken the trade.

Eight sessions later, the position was marked 42% above the entry price.

What actually happened

HIMS fell 23.5% between July 20th and July 29th. Not on earnings — earnings were still ten days out. Somewhere in that window the stock simply dropped.

A long strangle is long gamma. It does not care where movement comes from. The scanner models a specific mechanism — implied volatility ramping into the earnings date, with the position closed before the announcement — but the structure gets paid for movement of any origin. This trade won by its own design, through a channel the backtest does not describe.

That distinction matters, and it is the first thing worth being honest about: the setup worked, but not for the reason the historical numbers describe.

The high IV wasn't a warning — it was the point

This is where the reading gets uncomfortable if you treat "expensive" as a red flag.

Implied volatility of 112% over roughly three weeks to expiry prices a move of about ±25%. The stock delivered 23.5%. The move that made the trade profitable was almost exactly the move the implied volatility was announcing.

You were not overpaying for nothing. You were paying a high price because the market expected violence, and violence arrived.

High IV on a long-volatility position is not a mispricing signal. It is a statement about the size of the distribution you are buying into. Whether it is too high is a separate question — and one the platform answers elsewhere, on the Earnings move tile, which compares the move currently implied against what the stock has historically delivered. When implied sits well above realised history, you are paying for movement that tends not to show up. That is the reading that should give you pause. Not the price on its own.

You didn't need the big move anyway

There is a detail in this setup that reframes the whole thing. The take-profit is set at +10% on the premium paid.

The position peaked at +42% mark-to-market, but nobody following the strategy would have collected that. The trade would have closed at its target well before the stock finished falling. You needed a modest move, early — or a pop in volatility — not a collapse.

That lowers the bar considerably, and it explains how a setup can carry a 90% win rate without requiring anything dramatic to happen. Most cycles do not produce a 23% drop. Most cycles produce enough.

What a cheapness gauge actually does

Here is the thing I had to sit with: what is the use of an entry gauge that tells you to skip a winner?

A gauge shifts a distribution. It does not identify outcomes.

Entering in the expensive half of past entries is, on average across many cycles, worse than entering in the cheap half. That is a statement about a population. It says almost nothing about any single trade. A 90% win rate means two cycles in twenty lose — and no gauge in the world tells you which two in advance.

Reading a single result backwards to judge the tool is exactly the mistake the tool exists to prevent. It is the same error as concluding that a high probability of profit means a good trade: a conclusion drawn from one number, in one instance, about a process that only means anything in aggregate.

What the platform tells you to do with it

The Playbook on that same card does not say "cheap, go" or "expensive, stop". It says four things:

• On timing — where this entry sits relative to past entries at this point in the cycle. Context, not a verdict.
• On execution — confirm the fill before trading. Closing quotes overstate the real cost; spreads tighten after the open; send a limit near mid. On this setup the worst leg carried an 11% spread, which is the kind of friction that quietly eats a 10% target.
• On the thesis — the trade is about the earnings move, not a price view. No directional call is required, and the win rate already reflects the historical implied-versus-realised gap.
• On the exit — defined risk, the most you can lose is the debit paid, take profit at the target or let it run to expiration.

None of that is a signal. All of it is the difference between a number on a screen and a position in an account.

Where I landed

I still would not have taken that trade, and I do not think that was a mistake.

Skipping a winner is not an error when the process was sound. The error would be either of the two conclusions that feel natural after seeing the outcome: that the gauge is useless, or that expensive entries are fine. Neither follows from one cycle.

What does follow is narrower and more useful. Expensive is a headwind, not a veto. High implied volatility on a long-vol position is the price of admission to a wide distribution, not evidence of a bad trade. And the reading that should actually stop you is not the premium — it is paying for a move the underlying does not historically