Hacker News new | past | comments | ask | show | jobs | submit login

There’s nothing wrong inferently wrong with MBSs. They are a little complicated, but then again, so are lots of things. The problem that occured in 07-08 were two-fold: corrupt/lazy rating agencies and a lack of understanding about the long-tail correlation between mortgages. The copulas banks were using to join the probability density functions of the mortgages were incorrect and spit out much lower levels of correlation than was accurate.

In reality, MBSs are a truly beautiful idea that have a lot of potential to help a lot of people. The banks just needed to have better risk management and I think a lot of lessons have been learned. Despite what the public thinks, banks, especially Goldman, (one of the only institutions to not get hit very much by the collapse) take risk management very seriously. And the guys that didn’t have gone out of business (i.e Lehman).

And you’re right, this is the expected (good) outcome. Learn some lessons and start selling a useful product again gaining some wisdom. Nothing wrong with that.




Lehman had a C-level risk officer, which surely means they took risk seriously at the time they established it. In 2007 they didn't (they fired the risk officer!), but that just makes me worry whether it's possible for banks in general to durably learn risk-related lessons. Some banks are responsible and will manage risk properly on their own - to protect the others, maybe some risky but useful asset classes need to be restricted.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: