Correcting a Covariance Typo in Thorp’s Example 6.2

Summary

The note describes a replication problem in Thorp’s paper: the reader could not derive the stated correlation between two variables in Example 6.2. The reported resolution is that the paper contains a notation error: the first occurrence of the correlation should be covariance. This correction explains why the derivation seemed inconsistent and points to checking later parts of a paper before trying to reproduce an earlier result.

The document offers no derivation, underlying equations, or independent confirmation of the proposed correction; it points to an external answer instead. It is therefore a narrow lesson in reading and replication rather than a broader treatment of value at risk or covariance estimation. Readers applying the correction should verify it against the paper’s context and notation, since the note alone does not establish the intended result.

Key ideas

• A reported notation error replaces correlation with covariance in Thorp’s Example 6.2.
• A mismatch in a derivation can arise from a source typo rather than an error in the reader’s work.
• Reading ahead may reveal context that helps resolve an apparent inconsistency.
• The note does not provide enough detail to independently verify the correction.

Tags

# Thorp's var caclulation


# Thorp's var caclulation












I have been working through Thorp's paper, and with some guidance have got as far as page 20, but I am now stuck with Thorp's result in Ex 6.2 (on that page) where I cannot get the result for $Cor(X_1, X_2)$, so if anyone is able to show the derivation of this it would be much appreciated. Thanks

## Answer by markm (score 1)

https://quant.stackexchange.com/a/34706

I have found the answer. And it is a very dull one:

There is a typo in the paper, the first $Cor(X_1, X_2)$ is actually meant to be $Cov(X_1, X_2)$. There is a lesson here, and that is to read ahead before trying to replicate the results in a paper.

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.