1) Did sarbox have an impact on the growing market? (postulation) No. The CDS market's initial growth predates serb-ox and is a consequence of removing risk from bank's trading books to increase capacity. Serb-ox applies mostly to non-finance firms as standard financial intermediaries and asset managers have always been under more onerous regulatory framework. 2) Widening credit spreads on genuinely great AAA will drive bond managers out of treasuries, and thus all yields will go up. Um.... this is actually a debatable point. But I don't think so.... in order to make up the ground already lost - it's more likely that managers are going to dive into AAA assets to yield. The players in this space are rating constrained - so if they can buy stuff that doesn't take up their low-quality allocation at a higher yield, they might.
I again misused the word 'risk'. Lets start over since I understand it, just have misused words in the strictest context and restate so we are on the same page. 1) Buying the spread (betting it will go up) is in the industry 'Short the CDS'. I referred to it as opposite intuitively. 2) Selling the spread (collecting it) is in the industry 'Long the CDS'. Again I referred to it as 'short CDS' incorrectly initially. 3) Risk is obviously in selling the spread if the spread widens. Risk is also in buying the spread if the spread narrows. Those who have been in *risk* are ones who have 'sold' the spread, or formally have gone 'long the CDS' according to you, or synthetically are betting that the quality of the bond's outlook will stay high. Let me ask you though.. Does it jive with you that involvement in the CDS market by so many more participants now vs 10 yrs ago has possibly resulted in a tightening of credit spreads to abormally low levels over the last 5 yrs? I know there is a buyer and seller for every transaction, but in a way it equates to an increased presence of volatility seller hedge funds in options selling down the VIX to help keep vol low. Because that concept is pivotal to my whole thesis .. If AIG et al were selling spreads down, there existed a 'bear market' in credit spreads (that is, if you were betting on the spread increasing in number = my terminology) and the prevalent trend created an environment where buyers of insurance EXPECTED lower spreads. Now is an opposite situation I imagine, where players have been unwinding those positions. Whats left is a market that needs to rebalance itself. The sarbox read was just again something intuitive -- thats why I posed it as a question. Thanks.
In analogy to the GSEs who bought mortgages from banks to enable originators to do more business. So any rhyme of reason why the CDS market ballooned these last 5 yrs? Sarbox is 2002 -- perfect timing. One thing that I am curious about is the dynamic ... For every contract there are two sides. Why has this 'unwind' only had an implication of widening credit spreads? Don't these banks hold any 'short CDS' (as you call) bets on widening spreads of meaningful size ?? If this were merely a hedging instrument with a minority of speculative positionholders (betting on widening spreads), these positions wouldn't theoretically matter, even considering leverage. Why have we not heard of even 1 bank that is raking in billions right now? Is it only hedge funds who are betting on spread widening net? (versus banks only betting on spread widening to hedge risk)
I'm going to put aside the terminology issue. You are right that there are two sides. What you are missing is what pushed spreads to all time (and very unreasonable in retrospect) tights in 2006: synthetic CDO issuances. Massive amount of those papers will issued to investors and to create them, banks sold protections and pushed spreads to very low levels. When these trades came undone in 2007 and early 2008, banks had to buy back massive amount of protection and push spreads out. This is what is referred to as unwinding. It's mostly done at this point. The next potential bottleneck is in counterparty risk. But the force that tightened the spread has already played out in the opposite direction.
Your analogy with GSE is exactly right - except in this case it's synthetic issuances. It's the exact, corporate form of mortgage securitization (although lending standards were obviously nowhere nearly as bad). Sarbox was just something that happened around the same time.
So AIG for example, what do you think of their book was likely made up of that has been blowing up as of recent? still CDO CDS? or plain old corporate/other CDS?
They are famous for being loaded with lots CDOs.... single name CDS.... personally I don't think so. But I could be wrong that point. In anycase, what's the point for AIG to unwind their CDS book at this point if they have one.
If indeed the overall position of that 45T in notional exposure of the CDS market is only in real estate tied CDOs, and not corporate debt, then the negativity should subside once the markdowns are appropriate to the liquidation price of all of the underlying assets in the real estate market. Are there any accurate statistics out there at least reflecting what the net exposures per sector are? (ie 20T in CDO related CDS, 10T agency, 10T corporate, 5T muni, etc)
99.9% of CDS are on corporates and sovereigns.... there are almost no real estate tied CDS (unless you count CDS on mortgage CDOs... which sort of exist, but not significant)