Skip to content
All library documents

Managing Near-Expiry Option Gamma with Scenario Repricing

Article Quant Q&A · Author: Rodolfo Oviedo

Summary

At-the-money calls and puts can show very large gamma as expiration approaches, even though the option’s delta changes by a finite amount across any chosen spot move. The discussion asks whether a finite-difference gamma, calculated over a practical underlying-price interval, could cap the closed-form gamma. The replies favor viewing delta or portfolio Greeks across a range of market scenarios instead of relying on one gamma number.

One suggested workflow is to reprice the full portfolio under a grid of spot moves and inspect how delta and gamma change. Breaking the portfolio’s gamma down by expiration can help identify whether a near-expiry option is driving an unusual aggregate reading; it can then be monitored separately. The third-order Greek called speed describes how gamma changes with spot, but the answer notes that traders often find scenario repricing easier to interpret. The discussion gives practical guidance rather than a formal rule for choosing scenario sizes or a universal risk limit.

Key ideas

  • Near-expiry at-the-money options can produce very large gamma readings.
  • A portfolio’s delta across spot scenarios can be more actionable than a single gamma value.
  • Scenario analysis can reprice the full portfolio across a range of underlying moves.
  • Separating gamma by expiration helps locate near-expiry positions driving the total.

Tags

Full text
# Answer by dm63 (score 5, accepted)


# How to compute gamma for at-the-money regular calls and puts when they approach expiration to avoid explosion of portfolio's gamma?












When and at-the-money regular call or put approaches expiration, gamma tends to infinity. However, for practical purposes, there is only a finite change in delta. The problem is that if any of the options in your portfolio gets a crazy high number, this ruins the usefulness of the gamma of the whole portfolio.

I guess that that for practical applications you can choose a fixed change dx in the value of the underlying and compute gamma numerically using the pricing function f(x):

```
= ( (f(x+dx)-f(x)/dx - (f(x)-f(x-dx)/dx )/dx
= ( f(x+dx) - 2*(f(x) - f(x-dx)) / dx^2
```

The absolute value of this numerically computed gamma could be used as a ceiling for the absolute value of the closed-form formula-computed gamma.

Ideally, dx could be something like the standard deviation of daily price changes, or any number (of this order of magnitude) that the traders find easy to mentally manage.

I wrote the above just looking at the crystal ball. How do traders and financial engineers tackle this problem in practice?

## Answer by dm63 (score 5, accepted)

https://quant.stackexchange.com/a/43178

Many traders build a spreadsheet of how their delta changes across a range of market moves (-10%, -8%, ,....+8%, +10%) for example. That is a lot more useful than a single gamma number. It also means that the answers are finite and useful.

## Answer by nbbo2 (score 1)

https://quant.stackexchange.com/a/43180

In addition to computing Gamma for the overall portfolio, I also compute Gammas for each separate expiration date. When I look at it I see immediately that the "funny" gamma is in the subportfolio that is very close to expiration. I can then identify the option that is near ATM and treat it differently than others if I want (such as taking it out of the portfolio and managing it separately from the others, for the short time until expiration).

[But changing the way Gamma is defined I think is not a good idea].

## Answer by Lliane (score 0)

https://quant.stackexchange.com/a/43143

There is a greek for $\frac{\partial \Gamma}{\partial S}$, it's called speed.

> How do traders and financial engineers tackle this problem in practice?

As a third order greek, speed is not easy to manage mentally. Traders usually do a full repricing of their whole portfolio and scenario analysis where they see their delta/gamma at different levels of spot, rather than using speed.

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.