Penalty cap notebook, 8% target variant (executed outputs included)#17
Open
rolandgem wants to merge 3 commits into
Open
Penalty cap notebook, 8% target variant (executed outputs included)#17rolandgem wants to merge 3 commits into
rolandgem wants to merge 3 commits into
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The 8% target variant of the empirical cap notebook (outputs stripped per repo convention; runs end to end in about 15 minutes, figures posted in the temp channel).
Covers the recent asks: exclusivity windows 26/12/2s side by side in all charts and tables; May-calibrated constant now PER PAIR (each pair's own May quantile, which coincides with the frozen k x vol cap by construction; degenerates to 0 bps on tick-bound pairs at short windows, where more than 8% of windows have no move); revert-rate variation over time per method; correlated pairs ETHBTC and WBETHETH (wstETH/ETH proxy); a three-design Monte Carlo (daily resampling, hourly resampling, and an hourly AR(p) volatility model with the lag order chosen by AIC) addressing the concern that daily resampling favours hourly-calibrated methods; and an appendix split. Realized Ethereum revert rates per asset per day are computed in the last section (needs the penalties dataset CSV in data/).