Explain differences with Mogilyat et al. (2024)

[memoryfull]

Our replication of a Bank of Russia study of the cost channel of monetary policy in Russia by Mogiliat et al. (2024) revealed vastly different mean interest payments to cost of sales ratios while the medians are the same.

We have no explanation for this discrepancy other than material differences in the outlier filtering procedures between the target study and our replication which we cannot account for here as they are undocumented. These differences do matter, as they double the interest expense to cost of sales ratio, which was found to be insignificant in the original study.

[memoryfull]

The authors of the paper kindly explained that they used a weighted average instead of a simple average. The weight is the cost of goods sold.

Commit ad490d9 accounts for that. Now the results are in alignment:

I am closing this issue.