Risks of Risk Management

doctorwes · Post by **doctorwes** » Mon Mar 20, 2006 5:02 pm

I've been wondering whether we could see some nasty negative feedback effects arising from credit CPPI deals. I wish I knew more about this market, and whether it is big enough to pose a risk.

Cheng · Post by **Cheng** » Mon Mar 20, 2006 5:02 pm

JabairuStork,

how do you value the mgmt option in a CDO ? Following the classic approach I found it hard to value since it depends on the manager which names are in the pool plus that you would need an idea about future spreads. Putting all this together sounds for me like the problem is too hard to solve.

Re Ford / GM: I think many people (including myself) had the opinion that if one goes the other goes too because of a systemic meltdown. Now the perception has changed and everyone says if one of both goes bust the overproduction problem is solved, so after a few years (given that no technical sheah happens) the other one sould be back on track. Does this make sense ? Because if it does one should also bear in mind that there might be something like negative corr between two companies.

JabairuStork · Post by **JabairuStork** » Mon Mar 20, 2006 7:08 pm

re: the CDO mgmt option

This can not really be valued with standard option valuation techniques because it is a highly path dependent payoff in a very incomplete market. What you can do are assign some bounds given assumption about the forward cross-sectional dispersion of spreads within a given rating class. Assume that the protection seller always opts to reduce portfolio spread as much as possible while obeying constraints on portfolio composition and S&P rating for the tranche, while protection buyer does the opposite. In terms of getting long or short idiosyncratic risk, I believe this option is more valuable to the buyer of protection in almost every case.

re: Ford and GM

One possibility is that one goes bust and the other one picks up the slack, but another possibility is that Toyota or Hyundai or someone else eats their collective lunch. None of the old-line airlines seem to be really benefitting from the woes in that industry, so maybe they all have the same bad business model, and maybe Ford and GM do too.

kr · Post by kr » Tue Mar 21, 2006 12:16 am

This CDOEval thing is really the fundamental guy's margining system, which has been craftily arranged to work well for them. What's amusing is that we're hoping that Mark II of this thing will have a vanilla MTM short basket as well - if it can't be sold to the French, it can't be sold!

On valuing the CDO mgmt option, I actually have some ideas, that came out of internal debates here vis-a-vis valuing CDO equity for clients. This has got to be everybody's favorite topic, no? Ok, rhetorical question. What I told the big boss was: look, if it were on my desk and I wanted a price, I'd figure out just how much it would take to blow the thing up. Only natural. What you do is you buy the portfolio, refinance the tranches like a new CDO, and the net is equity - i.e. plain old balance sheet equation, only I've done the trade enough times to know exactly how it goes. But they said no. "Why are ya gonna stop now? It's unfair! You have to just PV the divs to equity... that's the way everybody does it." Fair enough, but how can the balance sheet equation be wrong... how.... can.... balance... ok, got it. So there is the kernel of an idea there.

Re: Ford and GM, I am just more or less laughing at the latest turn of events... the Board says they will investigate how the clowns in accounting muddled the numbers and jinxed the GMAC sale. Actually, today I am staring at RHJI, which is the publicly traded holdco for a bunch of not-so-attractive Japanese companies. And the thought I had was that even if you try to fix these things by eliminating a bunch of one-time issues which might disperse the cloud over the company (for Ripplewood it looks like they do a lot of pension plan terminations), it does not necessarily solve the problem (as in, half of the companies are having trouble with their debt covenants). As one of our loan guys told me before, yes, you can restructure the debt, which just leaves you with a crappy company that is debt-free. Precisely.

An old Model T can't win at Le Mans. Better to shoot the old horse and get some glue out of the deal.

Cheng · Post by **Cheng** » Tue Mar 21, 2006 8:41 am

re: CDO mgmt option

[b]I believe this option is more valuable to the buyer of protection in almost every case[/b]

I totally agree with you on this point. I was just wondering since some of our trades are self-managed (keep in mind we are prot sellers) that nobody so far seemed to take a margin for the mgmt option. Ok, everytime you do a substitution you PV the difference in deltas plus maybe some dubious cross-gamma effects nobody can verify and agree on a cash payment or adjust the subordination. But again, you play the old "which name gives me the most spread pickup while not changing the rating" game here which might give things a nifty turn if spreads move in your direction (read: tighten). Of course the real fun would start if you were prot buyer. Evil Smile

Thanks for your input kr, I will sit back and think about it.

Re: Ford and GM

Maybe one should really shoot the old horses and sell the remaining cars to North Korea where nobody bothers about things like quality and hey, since they don't have petrol anyway nobody is annoyed when the car gulps 15 l/100 km. The whole mess could be the big chance for Toyota and Co. We'll see wether they seize their chance.

Cheers Beer

IAmEric · Post by **IAmEric** » Tue Mar 21, 2006 4:27 pm

Getting back to the original question...

Do you guys think there is anything to the argument that (market) risk management can actually increase overall risk? Reading Aaron's article on the "Next 10 VaR Disasters", it is clear that the idea is not new, but I also doubt that most risk managers are as sensible as Aaron. In a crisis, do you think they are smart enough to avoid hitting the panic button once VaR numbers start shooting through the roof (an event that is not altogether so unlikely with vols so low now)?

(Aaron, I know you are busy promoting your book, but I am particularly interested in your opinion on the subject. For the record, I have your book, I just haven't picked it up yet :))

Cheers Beer

Eric

doctorwes · Post by **doctorwes** » Tue Mar 21, 2006 4:30 pm

Is it risk managers who hit the panic button, or is it senior management?

IAmEric · Post by **IAmEric** » Tue Mar 21, 2006 4:36 pm

Good question. I don't know. Bueller?

Relative to a risk quant, a risk manager is already pretty senior, but even then, I guess that instructions to panic could easily be handed down from someone senior to the risk managers (who may have even less of a clue than the risk managers).

I could just be spewing nonsense Blush

tripitaka · Post by **tripitaka** » Tue Mar 21, 2006 4:43 pm

[i]...Actually, today I am staring at RHJI, which is the publicly traded holdco for a bunch of not-so-attractive Japanese companies....[/i]

nicely put, I rate this stock A~VOID and figure that, if people smart enough to run Ripplewood figure that it's time to sell this sheah, I'm just about smart enough not to buy it from them... On the plus side, the SEC documentation serves as a good doorstop, or is large enough to sink a battleship if needed.

jungle · Post by **jungle** » Tue Mar 21, 2006 4:54 pm

i'm sure at least one bank pointed out there were no 'VaR feedback' problems in summer '03. sorry i can't be more specific about the bank or the comment. Blush