大陸留學生係海嘯元凶

快樂牛郎

http://phatdat.blogspot.com/2009/02/blog-post_27.html


http://www.wired.com/techbiz/it/magazine/17-03/wp_quant

rainbow-davie

有無咁緊呀

ricrick

哦，隥伯伯果然有遠見，我睇中國就黎可以稱霸全球

soli

當初條友發明個o甘重要既model ﹐ 但又乜recognition都無﹐ 等到D人濫用個model ﹐ 到爆晒煲後就賴人話人係原凶..

jiujai

agree.. every finance / econ / math / stat u grad know in college that all those model has assumptions. i've lot of fds work in finance, exec senior level in those big banks in mess.. Those finance guys are smart people and shouldn't forget that.. more likely "temporary memory lost"

This guy model is obviously assuming the CDS spread is calculated with current data, hence the CDS price following the risk. So any product bases on it should have a mechanism to adjust the risk with latest CDS price. Well.. math and stats can always be twisted from it original form to tick commoner ...

快樂牛郎

原帖由 jiujai 於 2009/3/4 19:00 發表
This guy model is obviously assuming the CDS spread is calculated with current data, hence the CDS price following the risk. So any product bases on it should have a mechanism to adjust the risk with latest CDS price. Well.. math and stats can always be twisted from it original form to tick commone

I haven't looked into his model but I don't think "assuming the wrong price" is where the problem is.  There has to be some kind of distribution function that models the prices of CDS.  Because CDS contracts have a relatively short history, the variances we have seen in the past were quite small, which inevitably drives the estimated variance to be lower than the actual. So when you do stress tests or try to determine the margins, you would believe that there is a very low chance of seeing the troubles we are witnessing today.