Why Option Delta Alone Cannot Describe Exposure Profiles
Summary
The document examines whether an option’s delta can be used to infer its expected positive or negative exposure profile. Delta measures sensitivity to changes in the underlying, and a regulatory exposure formula cited in one answer links exposure to delta, underlying value, and contract terms. That relationship is a limited approximation rather than a complete description of exposure over time.
The key caveat is that derivative exposure can respond to risk factors beyond the underlying price. A portfolio combining a long dated call and put may have near-zero net delta across a range of underlying moves, while still carrying material volatility exposure through vega. The size and shape of that component depend on changes in implied volatility, and the answer says it diminishes to zero at expiration. The discussion is conceptual and gives no calibrated exposure estimates or simulation results. It therefore supports using delta as a partial sensitivity measure, not as a full forecast of a structured product’s exposure distribution.
Key ideas
- Delta captures exposure associated with movements in the underlying, not every source of derivative risk.
- A delta-based exposure formula depends on contract size and underlying value as well as delta.
- A delta-neutral portfolio can retain exposure through option vega.
- Volatility-related exposure depends on implied volatility behavior and falls away at expiration.
Tags
Full text
# Deduce expected exposure profile from option/structure delta? # Deduce expected exposure profile from option/structure delta? I am thinking about whether there exists a relationship between the delta of an option (or any structured derivative) and it's expected positive/negative exposure? An intuitive question would be the following: A Foward has a Delta of 1 and given the above exposure profile and the Delta of an Option with the same underlying, can I deduce that the exposure profile of the Option equals Delta * Forward_Exposure? However, after running some simulations I see that this is not the case, part of the reason being (I think) that for exposure generation one simulates values for all relevant risk parameters and not just the one which corresponds to the Delta/sensitivity. If there are any questions on Definitions of terms I used, I am happy to clarify. Image taken from Jon Gregory's book on CVA. ## Answer by owner (score 1) https://quant.stackexchange.com/a/21973 Based on the `UCITS directives`: `E = n * c * UL * delta` where `E` denotes `Exposure`, `n = contract size`, `c= contract sie`, `UL= underlying price`. As you're probably aware from `BS`model, `call` has `>0 delta` `vs` `<0` for `puts`. Hope the explanation merely helps you to grasp the direct correlation between `E` and `delta` in a `UCITS` framework. ## Answer by RiskyScientist (score 1) https://quant.stackexchange.com/a/27800 The assumption of 100% delta for an option would give a good upper estimate for the exposure due only to the part of the option exposure that comes from the movements in the underlying price. But for example, imagine you had a portfolio which is long a long dated call and long a long dated put, such that the portfolio is overall delta neutral over a reasonably wide range of underlying price moves. In this case it is clear that the approximation will not be good enough, as the delta is zero: option vega has been ignored. For long dated options this is a larger effect, so would have a differently shaped curve from the one you show above. The vega portion of the exposure would depend on the "historical volatility of implied volatility" for the option in question, and the exposure goes to zero at expiration.
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.