Your 'cyclical adjustment on the underwriting profit', I presume, is just replacing it with something like the 20-year average underwriting profit (itself adjusted for the effects of growth and inflation)? Or just substituting U/W with a percentage of float, you used to use 0.8%, which in my opinion is very conservative.

More like the latter.
I estimate it two ways and average the two months. One is average return on float (1.3%), and the other is average percentage of premiums earned (2.7%, after excluding big retrocession events).

At the moment my formula estimates the current four-quarter rate as $2.255bn/year pretax. The true figure for the last four quarters is $5.517bn--it has been a good year, so my smoothing lowers it.
Conversely the true figure for the four quarters ending 2022-Q2 was -$545m, a bad year, which my cyclical adjustment bumped up to $1.936bn.

The estimated underwriting profit is up almost exactly 50% in the last five years. Maybe Berkshire has stepped away from supercat a bit, but the insurance operation as a whole is still growing strongly.

Jim