<t>It's true a covariance matrix based solely on historical data will be highly unstable (and hence optimal weights will vary wildly). However, there are techniques to try and get a more "well-behaved" matrix. Additionally, I think more advanced quant. allocation theories also question whether varia...

erstwhile - if you know of any good references in the context of CDO pricing, please post. I'd be interested in reviewing.Thanks!

Recent UBS article states "Current Synthetic CDO models are slowly moving away from correlation/copula framework altogether."Obviously there are many flaws with the copula framework, but what's better? Curious to hear from others about what's around the corner for CDO pricing methodologies.

<t>Are you trying to debug your own code or are you working with a third party pricer? Spread for 0-100 tranche should definitely not change with correlation (at least under the simple 1-factor model). I would start by looking at the exp. loss of the reference portfolio (before tranching) and make s...

<t>I haven't looked this over in detail, but I would agree with Lou2007 at least in terms of synthetic CDOs. ALL ELSE EQUAL, as correlation increases, equity spreads should decrease (the probability of 0 losses goes up) and the senior tranches spreads should widen (the probability of large losses go...

vespaGL150 - Thanks for the response. If you find the literature and wouldn't mind posting it, that would be great...

<t>Wow. Several noteworthy points across the responses, but I was hoping to generate a little deeper technical talk. Regarding Gmike's comments, believe me, if these things were AA/AAA returning 10% p.a., they'd be waived in all day long (assuming investment grade corporate credit). Maybe not what o...

<t>Imagine you're an investor who has accumulated a "nice" portfolio of managed synthetic CDO tranches. Suddenly you wake up one day and realize you're portfolio of CDO tranches has a tail risk profile that's not at all like "your father's bond portfolio". How would you go about reducing your tail e...

<t>Thanks for the reference. I actually implemented the model per Gibson, but the itraxx paper looks interesting. I'll give it a solid read. Regarding your question, I think for pricing, the standard is fixed recovery (I can't say I've ever seen anything different in practice). However, for ratings/...

I have a full MC CDO pricing framework in place, but for risk analysis I'm looking to speed up the pricing process with a semi-analytical model implementation. Any opinions/references on the best place to start? Thanks!

<r>Early in the post, sounds like an error was occuring when copying sheets, although now it seems more related to charts. In any case, not sure if the problem is related, but I was having similar troubles when copying sheets programmatically. Turned to be an Excel issue:<URL url="http://support.mic...

<t>I echo acastaldo's comment - thanks for the detailed response Aaron!I think one useful application of ES, although it does not meet the four requirements outlined below (most obviously 3.), is when trying to quantify tail risk in a modeling environment. I work in strucutured credit. Here the loss...

<t>I want to compare the tail risks of a CDO tranche to an "equivalent" holding of corporate bonds. Few questions:1. Any thoughts/opinions on what an appropriate "equivalent" holding might be? 2. Seems like unless you're dealing with large/diverse portfolios, tail risk is largely governed by random ...

How basic are you looking for? Chapter 17 of Fixed Income Securities by Martellini et. al. (2003) gives a decent introduction and has some useful references at the end of the chapter.

<t>What are you trying to do, exactly? If you're looking at a pricing problem, then obviously the market standard of implied correlation is the way to start. However, if you're trying to get at a real-world correlation structure I would suggest either using equity prices or even easier, take a look ...