Let me start with a truly bizarre quote by Simons:
You cannot make assumptions about how a security is going to perform and rate according to your own assumptions.
Huh? Of course you can. Any rated debt security ever sold, secured or unsecured does that. If I buy a Microsoft unsecured corporate bond, then it has a rating based on some assumptions about how Microsoft is expected to perform in paying back that debt. The idea that you can't evaluate debt based on assumptions is ridiculous.
What he should have done is put the word "bad" before "assumptions." Because that's absolutely true; and that's the mistake the rating agencies made. They made bad -- very bad -- assumptions about the mortgage market.
That raises a question: what if the rating agency and/or Morgan Stanley made better assumptions? Then should the security achieve a AAA-rating? Absolutely.
Now I have no clue what assumptions Morgan Stanley is making, and whether those assumptions are good. But I can provide an example explaining what I mean.
Let's say that those leveraged loans are truly heinous. Imagine they only turn out to be worth 20 cents on the dollar. How in the world could that security get a AAA-rating? Easy. Just include 80 cents on the dollar in a cash reserve to cover losses. That way, any losses from those leveraged loans will be covered by the cash reserve to protect investors' principal.
It's important for people to realize that the credit crunch was not caused by securitization -- it was caused by very poor assumptions used to rate securitizations. In a different world, with smarter rating agencies and investors who did due diligence, things might have turned out better. The future of finance should not exclude securitization. It should continue to be utilized, just with better assumptions.
]]>Now, the rating agencies were paid to rate by the issuers, but the issuers needed the ratings because the risk-based capital formulas for their assets vary by rating by regulatory requirement. My main point is that until an asset class goes through the failure portion of its cycle, likely loss levels are impossible to estimate accurately. Thus the regulators should bar bonds made from new collateral from being held by regulated entities, until we have seen the collateral fail as a group at least once under the disaggregated model where risk bearing, origination, and servicing are done by separate entities.
]]>
The complaint that rating agencies were paid by the issuers is always very exaggerated. In my experience as a former banker having worked with rating agencies on ABS transactions, they didn't really care if you thought their demands for additional collateral were unfair. If they said "jump," we asked "how high."
The problem was, as you mention, they didn't assess the collateral properly. Again, that's a problem with bad assumptions, not with securitization. There are always new lows to be reached in any market, but you can make more reasonable assumptions without eliminating collateral entirely.
1) It broke the link between the lender supplying the money and the mortgage originator who evaluated the risk. Everyone was playing with someone else's money, dumping all the risk on the buyers of CDOs, who had no practical way of evaluating the underlying loans.
2) Bubbles pop when you run out of buyers (the "last fool" theory.) Securitization, by linking California and other bubble markets to the entire world's pool of savings, allowed the bubble to grow much larger than it would have if local mortgages were funded by local banks.
If Morgan Stanley can get this project off the ground and sold, let us not make the same mistakes, force Morgan Stanley to hold in cash reserves half the risk required by the contract. No counterparty sell-off in the form of a Credit Default Swap.
If the bank is forced to keep the potential risk in reserve let's see them confidently make the loan then.
]]>