BEGIN:VCALENDAR
PRODID:-//eluceo/ical//2.0/EN
VERSION:2.0
CALSCALE:GREGORIAN
BEGIN:VEVENT
UID:www.tcs.tifr.res.in/event/792
DTSTAMP:20230914T125938Z
SUMMARY:Credit Risk: Simple Closed Form Approximate Maximum Likelihood Esti
 mator
DESCRIPTION:Speaker: Anand Deo\n\nAbstract: \nDiscrete default intensity ba
 sed or logit type models are commonly used as reduced form models for cond
 itional default probabilities for corporate loans where this default proba
 bility depends upon macroeconomic as well as firm-specific covariates. Typ
 ically\, maximum likelihood (ML) methods are used to estimate the paramete
 rs associated with these models. Since defaults are rare\, a large amount 
 of data is needed for this estimation resulting in a computationally time 
 consuming optimization. In this talk\, we observe that since the defaults 
 are typically rare\, the first order equations from ML estimation suggest 
 a simple\, accurate and intuitively appealing closed form estimator of the
  underlying parameters. We analyze the properties of the proposed estimato
 r as well as the ML estimator in a statistical asymptotic regime where the
  conditional probabilities decrease to zero\, the number of firms as well 
 as the data availability time period increases to infinity. We characteriz
 e the dependence of the mean square error of the estimator on the number o
 f firms as well as time period of available data. Our conclusion\, validat
 ed by numerical analysis\, is that when the underlying model is correctly 
 specified\, the proposed estimator is typically similar or only slightly w
 orse than the ML estimator. Importantly however\, since usually any model 
 is misspecified due to hidden factor(s)\, then both the proposed and the M
 L estimator are equally good or equally bad! The proposed approximations s
 hould also have applications outside finance where logit type models are u
 sed and probabilities of interest are small.\n
URL:https://www.tcs.tifr.res.in/web/events/792
DTSTART;TZID=Asia/Kolkata:20170721T171500
DTEND;TZID=Asia/Kolkata:20170721T181500
LOCATION:A-201 (STCS Seminar Room)
END:VEVENT
END:VCALENDAR