Friday, May 14, 2010

Great discussion on correlation assumptions in CDOs

The post is short and so are the comments.  Basically, the discussants are reviewing the formulas used to estimate the risk involved in a CDO.  Fascinating, especially in light of Tuesday's class.
Posted by at

10 comments:

  1. MattRMay 14, 2010 at 4:39 PM

    I think this quote pretty much sums it up:

    "Then they assumed that the CDS market was efficient (schizzo-finance) so they could estimate the correlations of the latent variables."

    Hopefully this crisis will show us once and for all that efficient market hypothesis isn't true.

    ReplyDelete
  2. KelseyMay 14, 2010 at 10:11 PM

    I honestly question whether we will ever be part of a society that understands we are not living in a predictable world. How many times does crisis have to happen before we realize this?

    ReplyDelete
  3. Chloe ReevesMay 15, 2010 at 2:40 PM

    I think it's human nature to want to have things explained by neat and simple formulas and unfortunately many of these fail in the real world. I hope we can learn from these mistakes and make more prudent financial decisions in the future.

    ReplyDelete
  4. Noah OesterleMay 15, 2010 at 9:21 PM

    Many of our assumptions and formulas we use in economics are usually accurate in predicting things 98% of the time. These fail when out lier events happen, like at the ends of the Gaussian curve we looked at in class last week. We need to be much more prepared for these events.

    ReplyDelete
  5. tcmtMay 17, 2010 at 4:35 PM

    To me the problem in all of financial modeling is that there are always extreme assumptions that, often times, seem reasonable or even realistic. However, when these assumptions blow up, things go very very wrong. In certain kinds of modeling (chaotic deterministic modeling) a change of .000001 causes a cataclysmic change in the model, and yet we make assumptions much larger than that without a peep. In my opinion, Waldmann is overly abusive to Tabarrok. We all make these assumptions, all modelers make these assumptions however Waldmann's point is an incredibly important one.

    ReplyDelete
  6. Jared AndrzejewskiMay 17, 2010 at 4:42 PM

    This post demonstrates how assumptions are the basis on our theories but they can be very wrong. Tommy points this out using a very good example with chaotic deterministic modeling. The thing is though if we don't have assumptions we won't be able to make any model and get it right some of the time, so it is necessary to assume. Hopefully as we get more advance we can start turning the assumptions into fact.

    ReplyDelete
  7. Ben MinkoffMay 17, 2010 at 6:23 PM

    We need to keep in mind these are all assumptions on other theories, and as Jared said, they can be very wrong. But these assumptions need to happen though because then we won't be able to prove theories wrong.

    ReplyDelete
  8. k06rl01May 18, 2010 at 9:36 AM

    Like Minkoff and Jared have stated these assumptions based on other theories. Like anything else I guess you could look at as different paradigms of the the financial industry.

    ReplyDelete
  9. JMalyMay 18, 2010 at 10:22 AM

    I have to agree with Tommy on this. While I can see where these formulas and equations can be used to create a lot of good and help create models that are very useful, but if we become to reliant on them and they end up making an assumption that was way off we tend to suffer severe consequences.

    ReplyDelete
  10. KelseyMay 18, 2010 at 2:26 PM

    We are never prepared for any outlier events. If only there was a way to add these into our models. I just don't understand how we always get sucked into thinking society can fit perfectly into a model.

    ReplyDelete

Subscribe to: Post Comments (Atom)