OCC TIMS: Stable-Distribution Scenarios and Portfolio Offsets

Summary

The answer offers a high-level account of the OCC TIMS options risk methodology. It says the system’s statistical calibration was based on a stable distribution, which includes the normal distribution as a special case, while option values were calculated using Black–Scholes or tree models. It describes TIMS as producing tables of underlying price moves, comparable in broad purpose to SPAN, with volatility adjustments.

The explanation also notes that portfolio offsets and correlation recognition involved product-specific rules, including rules related to broad index products. This suggests that portfolio margin methodology combines modeled stress scenarios with tailored offset rules. The account is brief and does not supply a formal methodology, parameter details, calibration procedure, or current documentation. It also flags dividend handling as uncertain, so that point should not be treated as a confirmed description of present-day TIMS behavior.

Key ideas

• The answer describes TIMS statistical calibration as based on a stable distribution that encompasses the normal distribution.
• It distinguishes that calibration from option valuation, which it says used Black–Scholes or tree methods.
• TIMS generated underlying price move scenarios and incorporated volatility adjustments.
• Portfolio offsets relied on product-specific rules, and the answer leaves dividend treatment uncertain.

Tags

# Where can I find OCC TIMS methodology?


# Where can I find OCC TIMS methodology?












I searched the web, but could not find OCC TIMS methodology detailed description anywhere. Any ideas?

OCC TIMS

## Answer by Greg S (score 1)

https://quant.stackexchange.com/a/85189

The TIMS system was based on the stable distribution. As such, it nested the normal distribution. The underlying statistical parameters were calibrated under the assumptions of the stable distribution, but pricing was Black-Scholes or tree based. The dividends were not handled (they still might not be). The system would generate a table of underlying moves similar to SPAN, especially now with the vol adjustments. Because of products like SPX and SPY, there were all sorts of rules about offsets. It was a hodge podge approach to give appropriate offsets and recognize correlations. If RBH is still around that gives a view of what the output of TIMS looked like.

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.