Use a Paired Test for Matched Market and Model Prices
Summary
The document addresses whether market prices and modeled prices should be analyzed with a dependent or independent statistical test. Its answer is that corresponding observed and modeled prices form paired, or dependent, observations when each model value is calculated using assumptions aligned with that market observation, such as the underlying price and time to maturity.
The key distinction is the matching structure: each market price has a specific model price for the same conditions, so the comparison concerns within-pair differences rather than unrelated samples. This is a concise methodological note and does not describe a particular significance test, its assumptions, or any price data. The guidance applies when observations are genuinely matched by corresponding conditions; it does not establish that a given pricing model is accurate or that the differences are statistically significant.
Key ideas
- Market and model prices are paired when each model value corresponds to a specific market observation.
- A paired design accounts for shared conditions such as underlying price and time to maturity.
- The choice between paired and independent tests depends on how observations are matched.
- The document gives no test results or evidence about model accuracy.
Tags
Full text
# Nature of the sample # Nature of the sample When we calculate a difference between the market price and modelled price. so to test the significance whether we apply dependent test or independent test. ## Answer by nbbo2 (score 1) https://quant.stackexchange.com/a/30320 When you compare each market price to a model price calculated under corresponding assumptions (regarding underlying price, time to maturity, etc.) you have a paired sample (also called dependent sample).
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.