The Sub-Prime Primer : Understanding The Sub-Prime Problem

There is a lot of talk about markets tanking because of the “sub-prime” fiasco centered around the US. I have tried to put together some facts and perspectives that seek to explain just what is happening and the extent of the problem, with a minimum of my usual opinions on the insanity of it all.

Disclaimer: This article was written originally in September 2007, so some data points may be outdated.

Saving Deficit

It all started with the old problem. There is a savings glut in Asia, which consists of too many ex-poor people who had freshly come into too much money (i.e. Dollars) and did not know what to do with it. So, predictably, they saved the Dollars and sent them back to the US in the form of “investments” into the Bond markets.

The US has for long been accused of being “savings-deficit”, i.e. net borrowers. In hindsight, we don’t really know which (chicken or egg?) came first, the excess borrowing habit or the excess savings habit. Because of conventional wisdom that savings is good and borrowing is bad, the media set up a harangue, blaming  America for its wayward ways.

I offer the rather contrarian conjecture that all this happened because of an excess of savings, rather than just a bad borrowing habit. Let me point to 2 markets, both of which have evolved post-2000. The first is the “sub-prime market”, which happened after the “prime” market ran out of borrowers. Sub-prime loans made up a quarter of the loans given recently in the first 10 months of the last year. These are “stated income (but not verified, with no underlying documentation)” loans as compared to the hitherto “stated income, verified asset” (SIVA) loans that used to make up the broader market.

What explains such a no-questions-asked lending policy? Mind you, it is not just one individual lender, but an entire marketplace that emerged and grew into the fastest-growing segment of America’s huge financial services industry.

Side by side came the CDO. Put simply, the CDO is a synthetic ‘securitized’ instrument that escrows the cashflows from an underlying asset (mortgage, loan, or bond) into a set of tradeable rights, called the CDO (Collateralized Debt Obligation by the issuer, which confers certain rights on the investor). The market for this grew frenetically as well…….and it developed a symbiotic relationship with the subprime mortgage market, like the twin-headed Hydra that feeds on itself. $100 bn of the $375 bn of CDOs sold in the US in 2006 were sub-prime. By geography, 75% of the CDO sales were in the US, which, through its banks, emerged as the chief “insurer” of credit risks on its own bankrupted population. And why not? After all, they had the monopoly on the right to print the Dollar, didn’t they?

The Meaning of Portfolios

By Mar, ’07, half the CDOs issued had sub-prime debt and these CDOs, on average had up to 45% of their portfolios in sub-prime loans. In other words, roughly 22% of all CDO issuances by value had subprime debt. This was the bond-market equivalent of the famous “portfolio insurance” which created a moral hazard prior to the Oct, ’87 crash on Wall Street. On an issued base of roughly $ 1trn, that works out to $220 bn of CDOs that had sub-prime loans. That is, some $220 bn of the $800 bn of sub-prime loans had been CDO-ed away into a yield-hungry market.

Having exported their insane credit risks to the CDO market, issuers went back to the mortgage market for more. The only focus of these I-banks was the fat fee that comes from issuing RMBS (residential mortgage-backed securities), bundling them into a CDO, and down-selling them to a set of “bigger fools”. Trust me, it is a long story, but in a much smaller way, this is PRECISELY what has been happening in the Indian real estate market. Were it not for an alert regulator (I have always maintained that the RBI is the best Central bank in the world, don’t ask me how they have managed it), we would have been headed for similar trouble in India.

So who are these bigger fools? The people who manage the savings of “widows and orphans”: take a look at this list. The New Mexico Investment Council ($222 mn), the General Retirement System of Detroit ($38.8 mn) the Teachers’ Retirement System of Texas ($ 62.8 mn), Calpers, et al. And why did they buy these CDOs? Because they were getting up to 10% over Libor, more than TWICE the yield on bonds with the SAME rating. The “equity” portions of these bundled CDOs (a.k.a toxic waste) paid 20%. This was a riskless “rating arbitrage” where you get a yield better than junk bonds, but an “investment-grade” rating to show to your Investment Committees. Which crowd of monkeys can resist such a moral hazard?

I can tell you what must have happened in the marketing presentations to these poor sheep. I-banks would have held out that the Rating Agencies ‘oversee’ these CDOs, pretending that the portfolio quality is monitored by the raters. Some 7% of the equity tranches have ended up in the hands of Pension Funds, endowments, and religious organizations, supposedly the most restricted and conservative of investors. These poor sheep look to ratings for their investment appraisal. Yet take a look at this chart below:

I don’t have average yields, but any good trader can see that the vast majority of issuances must have shown deterioration in credit quality far out of proportion to the incremental yield. The only place where a trader could still make money, I suspect would be if he bought in the BBB bucket and sold the A bucket. The incremental yield, I suspect, would be much higher than the incremental probability of default, which is only 12.5% extra. 15% extra yield would compensate for this additional probability, but the actual yield would be much more. How do I know? That is just how markets work….!

Explaining the subprime mortgages

Back to the subprime mortgages then.  The Mortgage Bankers Association, the apex body of mortgage lenders, opines that 1% of the 50 mn mortgages outstanding will end in foreclosure, a figure much bandied –about in the media. But this assumes “normal” lending over a long time, not the kind of lending that creates and perpetuates bubbles. Even they have admitted to a 12.56% default rate in the sub-prime segment in Q3, CY’06, up from 10.76% in Q3, CY’05. Mind you, this is after examining 42.6 mn of the 50 mn mortgages outstanding.

Sub-prime mortgages were around 20% of the market in 2005-06, 12% of which are already in some part of the foreclosure process.

However, the more independent Center for Responsible Lending (www.responsiblelending.org) took a sample of 6mn loans issued between 1998 and 2006 and came to a default rate of 4.4%, i.e. 2.2mn mortgages out of the total 50mn. The biggest culprits are the ‘multiple mortgages’, the ‘stated income, no documentation, no verification’, Blacks and Hispanics. Translate this into India and take a walk around Gurgaon….tell me what you see!

Just how bad can things get? Anecdotally, one RMBS quoted on Bloomberg shows a delinquency rate of 54% of loans 60 days overdue and 17% in foreclosure. These bonds would have a market value of up to 50% of Face Value. By number, 65% of the bonds listed in ABX indices (that tracks subprime lending) don’t meet the rating criteria that were in place at the time of their issuance. Some $60bn of “mezzanine” CDOs have up to 70% of their portfolios invested in subprime mortgages. Yet, some 90% of these bonds by value are still rated “investment–grade”. Why? Because the defaults have not yet started, the raters cannot trigger off a scare based on ‘rumor’.

The total size of the problem: 25% of CDOs would be $250 bn, and a 20% default on the $800 bn sub-prime market would be $160 bn. Yes, there may be some hyperbole, so let us take half of that and round it off, say, to about $100 bn. Looked at one way, that is just a quarter’s worth of Fiscal Deficit for America. The Fed could create another bail-out with the injection of funds into the banking system, and life would get back to ‘normal’.

The real problem is the potential collateral damage, not just the couple of hundred billion in subprime. Between the sub-prime and the “prime” market (which was the AAA component, ‘guaranteed’ and partly refinanced by Fannie Mae and Freddie Mac) were the Alt-A loans, less risky than sub-prime, but not prime. This is the part to watch. 30% of prime loans in the last 2 years were option Adjustable Rate Mortgages (ARMs), floating rate loans, where you have the “option” to set your rates for a while, subject to certain conditions of the repayment period. 80% of sub-prime loans had the same (option ARM) clauses. You can imagine what kind of rates such borrowers would have chosen. When Mr. Greenspan had brought rates down to the 1.2% range, he actually recommended publicly that people should take these option ARMs. While these rate structures balloon in later years, the borrower is looking to sell the asset and extract some profit on the marginal or no equity he has put in. $1.8 trn of these loans are coming up for interest rate resets this year, of which 80% of the borrowers were only appraised to pay the minimum (usually 1^st year) installments on these loans. Till 2003, only 0.5% of all home loans were these Option ARMs, and 1% of California loans. Today, 23% of all home loan borrowers have these Option ARMs, with 80% of California loans.

The Piggyback loan

‘Stated Income’ loans, also called ‘Liar Loans’ had 60% of the borrowers overstating incomes by half. ‘Piggyback’ loans were given to fund the equity contributions of borrowers, thereby pushing even older loans back to 100% funding on the house value. ‘Piggyback’, ‘no documentation’, and ‘liar’ loans accounted for 47% of total loans written in 2006, resulting in loan-to-value (LTV) going up to 86.5% in 2006 from 78% in 2000. So if 60% of these liars overstated their income, some 30% of the mortgages done in 2006 were to such people. Can you imagine the oncoming default rate in this segment?

The rates on the bigger chunks of these loans will be getting reset from October, ’07 onwards. Already, the Case-Shiller Index of house prices is down 3% on average, coast-to-coast. This means that in some markets, notably California and Florida, the drop in house prices is running from 5-15%. These loans are already underwater.

Let us go back to the rated bond market. In 1999, there were only 9 cos with AAA ratings, including GE, Berkshire Hathaway, Johnson & Johnson, Exxon, and Toyota (you can see the health of these cos 8 years later in 2007). By giving these RMBS the same nomenclature, the raters confused the bond investors, especially the immature Asian/ European bond investors, into thinking that these mortgage bonds were ‘equivalent’ and could command the same rates. Now raters are dropping these ratings within 2-3 years of issue, leaving bond investors feeling cheated. June ’07 issuances have dropped to $3 bn; the sub-prime component is down to zero. That shows the liquidity crunch in the CDO market. Of the 2006 pool, Moody’s has already downgraded 19% of these issues, and another 30% are on watch. The Investment Banks who created and sold these CDOs made 5% on fees and 5% on selling commissions. That is 10% of the CDOs issued, over what the investor makes. Plus they get the yield on the ‘toxic waste’, which by the way, was often less than the fee the Banks have already made on the CDOs.

Let us now turn to the actual damage already recorded. Bankruptcies in the mortgage lending sector have already trebled, from 24 to 78. Old Hill Partners, Wharton Asset Management, and United Capital ($500 mn in assets) have stopped investors from withdrawing capital. Braddock Financial is closing its sub-prime mortgage-laden Galena Street Fund, till it can sell its $300 mn holdings. In London, Caliber Global Investment is closing after losing 53% of its $ 2bn value. And Boston-based Sowood Capital announced a 50% loss on its $3 bn asset base. Countrywide Financial, a leading sub-prime lender has reported that 30-day-or-more-overdues on its sub-prime portfolio have jumped to 23.7% from 15.3% yoy. Even on prime home loans, overdue have gone to 4.6% from 1.8%. Net Income slid 33% and Bank of America had to step up with $ 2bn.

Now let us examine the prospects for employment. 43% of the jobs created in the last 5 years, since the IT Bust 2000, have come from the credit-funded real estate sector. And 40% of these real estate jobs have gone to realtors, i.e. people who do the booking for housing sales, called ‘property dealers’ in India. Against a 5-year employment growth rate of 1.2% annualized, the job growth rate of home mortgage lending was 21%, home construction jobs grew 23%, and realtors 65%.  What happens to these people when the buying and selling of houses drops by 25%?  That is why US jobs data came in at (-)4000 jobs, against an expected +110,000.

The impact on GDP and American Personal Consumption

And look at the impact on GDP and American personal consumption. Between 1997 and 2004, the value of residential construction amounted to 16.4% of American GDP, twice the average of 6-8% seen over 1968-2005.  70% of the rise in household Net Worth since 2001 has come from rising home prices, neutralizing the drop in equity market valuations. An increase in housing wealth boosts spending by 9%, while a similar increase in equity wealth pushes spending by only 4% (because of the inherent volatility).

By a number of people too, more people are exposed to real estate than equities. 69% of American households live in their own houses, while stock ownership varies between 45-50%, but even then, equities are a much smaller proportion, being their second or third largest asset. So the wealth effect of home price appreciation is greater and more widely distributed across the American population.

The ATM effect has also been wider. Mortgage Equity Withdrawal (MEW, which is nothing but the increase in leverage outlined above in piggyback loan data) has contributed to 25% of GDP growth in the 80’s being at 1% of disposable incomes during that period. It rose to 4% of disposable income (i.e. 400% growth rate) over the period 1997 to mid-’02. At $150 bn, this was at 6% of disposable income by 2003, peaking at $250 bn (10% of disposable income) by 2004, contributing a whopping 75% of GDP growth. This has now dropped 50% to $113 bn by Q3, ’06. The ‘engine’ of growth has run out of steam…

So to summarise, what have I said so far? That bubbles were blowing across America, the big one being in housing and mortgages. These created artificial jobs, holding up America in the last years of Greenspan. The chickens are now coming home to roost. And from now onwards, it will be about as much fun as going for a root canal with your in-laws…

We turn to the collateral damage expected in equities. We have already seen how the blowout in the mortgage market has affected the entire bond market, particularly the subprime and high-yield/ junk segments. And the junk securities markets have propelled stock buybacks. In Q1, ’07, non-financial corporates bought back $128 bn in stocks and issued $130 bn in debt. Coincidence?

This same money was flowing into the high-yield bond market in Emerging Markets and the ECB markets that I talked about in my column last month. And we don’t know how much has been funding those PNs that are going into frontline (and mid-cap) Indian stocks.