Wednesday, December 12, 2007

Of Quants, Faith and Alcoholics Anonymous

"R&R Consulting, a New York City structured finance [SF] credit metrics firm, celebrates the power of mathematics in [SF] with the relaunch of its website, http://www.creditspectrum.com/. ... 'You don't have to have a Ph.D. in engineering to understand what we do,' says Sylvain Raynes [SR], one of R&R's partners, who has a doctorate in areospace engineering from Princeton. 'But you do have to be serious about the science of finance'. R&R uses heavy-duty modeling techniques taken from engineering to analyze the credit quality of non-recourse debt obligations backed by pools of receivables, like credit cards, residential mortgages, automobile loans, equipment leases, or even corporate debt, known as structured securities. ... In the words of Ann Rutledge [AR], also an R&R partner, 'There's only one set of tools you can use to predict how these pools will tend to behave over time. And that's mathematics--mathematics and statistics. You can't have an honest understanding of risk and return on structured securities without these tools.' ... '[SF] is not taught at any depth in most universities, even in financial engineering programs,' says Raynes. 'Practitioners come to our classes, or read our book, 'The Analysis of Structured Securities,' ... and literally feel joy that a whole world is opening up to them'," http://www.prweb.com/, 17 May 2005.

"'No one really knows how to price asset-backed securities and CDOs and that's a real problem in the market place now,' said [AR], principal of R&R consulting, a [SF] consultancy in New York", WSJ, 10 August 2007.

"'Do you believe? Then place your hands over your investment portfolio and declare your faith in the ones who swear to its goodness', [said] Bob Moon. 'The ratings business , as it stands today, is a quasi-religion that does not have any more to offer than that', [said SR]", at http://www.marketplace.publicradio.org/, 19 September 2007.

"'Who are we going to trust? Are you going to look to JPMorgan or Goldman Sachs? Are they going to rate structured deals? Can we trust them on valuation? We need to take the high road ... because I might not do any better.' ... 'The problem as I see it, as we at R&R see it, is a fundamental lack of recognition of the different basis that exists in [SF] as opposed to corporate finance. ... Valuation ... is the only problem for finance,' [SR]. Raynes then asked how do we move forward from here: 'The first thing I would try to do is to establish a unique valuation theory for structured securities. Does this exist? Well it could if people are listening. They are not. ... '[SF] is a beautiful thing. The men of Wall Street like to fall in love with beautiful things. ... So we don't have to be in love with [SF], we just have to respect it'," http://www.prmia.org/, 25 September 2007.

"No one really knows how ... now", saith AR. Was the knowledge lost? Who ever knew? AR admits she doesn't know. Later, I will offer a suggestion using mutual fund accounting which I understand. "If you give me six lines written by the most honest of men, I will find something in them which will hang him", Cardinal Richelieu (1585-1642). Yes!

"The solution for the subprime-infected credit market seizure that wiped out $66 billion of securities industry capital this year may be found in a bunch of hospital bills. ... [Richard] Field and Deloitte & Touche are a part of a fast-growing cottage industry that includes household names such as Moody's Investors Service trying to shine a light on the world's most opaque markets. ... As much as $2 billion a year could be made by someone with a model that accurately valued asset-backed securities, according to [SR], a principal at R&R Consulting in New York. ... Dan Fuss, vice chairman of Boston-based Loomis Sayles & Co., hasn't found an adequate system to help value CDOs so he refused to include them among the $22 billion of securities he manages. ... Raynes of R&R Consulting was hired two years ago by an affiliate of the Bond Market Association that deals in asset-backed debt to assemble teaching materials on valuing such securities for its members. The trade group paid Raynes half of his fee in advance and failed to pay the remainder after receiving the materials because its members found them too complicated. 'All of the managing directors could not understand any of the formulas so they did not pay us,' he said. 'This is basic bond math, duration and convexity. This is not rocket science. ... Firms promoting new pricing models are 'trying to cash in on fear,' Raynes said. ... We as a market must recognize there is a lack of understanding of what value means in [SF]'," Caroline Salas at http://www.bloomberg,com/, 4 December.

I suspect 73 years from now Graham & Dodd's Security Analysis will outsell SR book. I agree with SR, most ratings are worthless. I disagree with SR, we don't need a "different basis" for SF as opposed to corporate finance valuation. SR belief is pure faith. He believes, I do not. Cash flows are cash flows. Discount rates are discount rates. Consider "arbitage pricing theory" (APT). APT came out in the mid-1970s. I still pick up finance or economics books from time-to-time to keep up.

Are Wall Street MDs as dense as SR thinks? Most went to our better MBA schools: Wharton, Stanford, MIT, Northwestern, Harvard, Columbia and my alma mater, Chicago. Their inability to understand basic bond pricing surprises me. At Chicago, many years ago, I took Rueben Kessel's class on the "Term Structure of Interest Rates", which touched on concepts like duration and convexity. So? Don't MBA schools teach that any more?

SR may be wrong; cash flows are cash flows and their "wrapping", in a corporation, an SIV or a tortilla, only changes their tax and legal attributes. As Cassius said to Brutus, "The fault, dear Brutus, is not in our stars, but in ourselves, that we are underlings", Julius Ceaser, I:2. Similarly, the fault dear SR may be: cash flows are cash flows! Further, do not try to be more precise than your data permits. A humbling book decrying excessive precision is: On the Accuracy of Economic Observations by Oskar Morgenstern (OM), 1950. OM's opinion: economics is at most a "two-digit" science. Be warned. By the way, OM was a Princeton economics professor. I wonder if his name is even mentioned on that campus today.

Mutual fund pricing values each position, using current market prices, adding up the value of all positions, then dividing the total by the number of oustanding shares for a unit price. Simple, voila. We get a few "hooks". If we invest in a "load fund", with say, a 5.75% fee, when we invest $10,000, we only see $9,425 ($10,000 - $575) on our statement. Easy enough. What has this to do with CDOs? About 30 years ago, we had "dual funds" (DF). DF had "common" and "preferred" shares. The preferred shares got the underlying investments' dividends and the common shares the rest. DF were interesting since in theory, you could buy an equal number of common and preferred shares to recreate the underlying portfolio on an unlevered basis. Some academic work was done on why common and preferred shares sold at premiums and discounts. That's not germane to my discussion here. What is: the common and preferred shares together, sold at a discount to the underlying portfolio. Why? The costs of running the DF were "impounded" into the shares current market price. Neat, huh? The market rarely is stupid. What has this to do with CDOs? Simple, how many "tranches" you break mortgages into doesn't matter; in total, they are worth the mortgages! Less administrative costs, of course. Just like DFs. The valuation problem is not with CDOs, but with the underlying mortgages! It's on the ledger's asset, not liabilities side! Wasn't that easy? As to valuing the tranches, just use bankruptcy law concepts as to their seniority, bearing in mind, at some point, even a "deep out of the money" tranche, has option value. It may be one dollar on 100,000, but some value. Debits to the window, credits to the door. Class dismissed. For extra credit read, "The Theory of Recapitalizations And The Evidence Of Dual-Purpose Funds", Litzinger & Sosin, The Journal of Finance, Vol. 32, No. 5, (Dec., 1977) 1433-55 which deals with contingent claim valuation. Speculation: what interest rate would Citigroup have charged a person with a $100 million mortgage portfolio to borrow $85 million against it and create his own "toxic waste" dump? Just curious, more or less than the mortgages' interest rate?

"'The first step if you are an alcoholic and you need to stop drinking is to get on the podium and admit to yourself you're an alcoholic.' ... Raynes said", http://www.bloomberg.com/, 4 December. Amen. That's good Bill Wilson advice as millions of AA members can attest. Try this, "My name is Sylvain. I am a structured finance professional. ... " Sylvain, after the meeting we will introduce you to people who have at least a year in the program so you can pick one to be your "sponsor".

$2 billion a year? Is SR looking for a "philosopher's stone" that will turn lead into gold. I have a better idea SR, if you find the "mother of all valuation models": keep it to yourself and use it to select stocks to purchase and short. If the model works, it should enable you to make some real money. Had you known how to value CDOs, etc., a year ago, you could have made a nice profit shorting: Ambac, MBIA, Countrywide, Citigroup, etc.. Tens, maybe hundreds of billions a year could be available to you. See my 23 August, two 27 September and 24 November posts.

4 comments:

Anonymous said...

GW --- Yes, in this field of endeavor every model is a philosopher's stone. The Mortgage Backed Obligations are worth a bit less than the sum of the underlying real estate sold at a fire-sale price less all the costs. You factor in nothing for the creditworthiness of the mortgagee, as they could not qualify for the mortgage on traditional repayment terms. Ask any grey-haired mortgage lender who advances their own money. The problem is that the people involved in these "new vehicles" were not risking their own money, they do not know what business and risk is. They were taught by dilettante academics in schools of enterprise (miss-named business) administration of bureaucracies --- excellent for what it is. They need to stay in their field and not be podiatrists performing heart surgery. HDF

Independent Accountant said...

HDF:
You are more skeptical of "financially engineered" products than I. I agree, the "administrative costs" reduce the value of the package of securities which are supported by the underlying real estate. Just like my dual fund example. A problem with these products is that their originators let their credit standards slip under the theory that "we can sell the paper to someone once we get it rated AAA. Therefore it doesn't matter who the borrower is". If the originators had to hold even 10% of the mortgages, they would not have made many of these loans. We are not far apart.

Independent Accountant said...

"There is innovation here, but only in packaging, nothing in value", Jim Willie at www.kitco.com, 27 December. Precisely.

Anonymous said...

Alcohol dependent people must first admit the fact that he/she is alcoholic. This is true since not all have the courage to admit to themselves that they have alcoholism. A successful road to recovery must first be accepted by the person that he/she needs treatment, and is willing to change the bad habits.