Serving the Quantitative Finance Community
Log-normal LMM in high vols (JPY)
Hello,I was wondering if anyone had any experience they could share on how to adapt these models to a high vol environment, e.g. JPY. Put aside the debate on log-normal vs. normal for this currency, because we are not always masters of our own work. The most noticeble problem I am seeing is in re-pricing the yield curve instruments, and while methods such as predictor-correctors make a huge difference, they still leave troublesome errors for, say, 20y structures with 20% + vol. Kind Regards.