Its Wildness Lies in Wait
Last week's Bulletin stressed that money and credit excess would likely manifest into significant divergences in relative prices. The analysis did not give proper attention to another key manifestation, extreme financial market volatility. It was certainly a week of extraordinary volatility and extreme price divergences in the U.S. stock market. For the week, the S&P Bank index declined 1% while the Semiconductors surged 17%. The Dow and the S&P500 gained better than 2%. The Transports added 1%, while the Morgan Stanley Cyclical index jumped 4%. The Morgan Stanley Consumer index posted a marginal gain, while the Utilities dropped 3%. The broader market rallied, with the small cap Russell 2000 and the S&P400 Mid-Cap indices gaining 3%. Technology stocks rallied significantly, with the NASDAQ100 surging 10% and the Morgan Stanley High Tech index gaining 8%. The Street.com Internet index jumped 14% and the NASDAQ Telecommunications index gained 10%. Biotech stocks jumped 9%. Financial stocks were mixed, with the AMEX Securities Broker/Dealer index increasing 9%, while the consumer finance stocks came under heavy selling pressure. With bullion dropping almost 7%, the HUI Gold index sank 7%.
It was a bit of a reality check for the U.S. credit market, with yields rising across the board. Two-year Treasury yields jumped 10 basis points to 2.80%, with 5-year yields increasing 15 basis points to 3.87% and 10-year yields 17 basis points to 4.67%. The long-bond saw its yield jump 12 basis points to 5.43%. Benchmark Fannie Mae mortgage-back yields jumped 14 basis points and agency yields generally increased 16 basis points. The benchmark 10-year dollar swap spread widened 3 to 71. The dollar index added about 1% this week, although weakness returned today as financial stocks faltered.
The Mortgage Bankers Association weekly index of applications to refinance jumped 24% last week, to one of the highest levels ever. It has now more than doubled in the past three weeks. Broad money supply declined $14 billion, but remains up more than $170 billion over the past five weeks.
Today from Bloomberg: "Argentina's credit rating was cut by Moody's for the fourth time this year to below that of any other country. The decision to reduce Argentina's foreign and local currency debt ratings to 'Caa3' from 'Caa1,' lower than ratings assigned by Moody's to Cuba, Ecuador and Pakistan, reflects the likelihood the country will default on $132 billion of debt." Moody's also downgraded asset-backed securities exposed to Argentine-based collateral. It now appears that Sunday's elections for the entire Senate and half of the lower house could further weaken the national government, with fear that a strong showing by the opposition could lead to the resignation of Finance Minister Domingo Cavallo and precipitate a government debt default.
Also today from Bloomberg: Moody's "indicated it is considering lowering the credit ratings of California counties on concern a state budget deficit will hurt local finances. Moody's assigned a 'negative' outlook to the counties' (and the view) 'primarily reflects the possibility that the state could address a significant part of any budget shortfall by diverting revenues' from counties and other local governments."
"Fitch-NY-October 8, 2001: Fitch has reviewed the potential short and long term impacts that the events of Sept. 11 are likely to have on all of the U.S. cash flow Collateralized Debt Obligation (CDO) that it rates and is placing certain classes of 26 CDO transactions on Rating Watch Negative."
From PRNewswire: "MasterCard today reported performance results for the first six months of 2001 During the first half of 2001, financial institutions increased their issuance of MasterCard cards around the world at the highest rate in 10 years. At June 30, 2001, member institutions reported that they had issued 475.3 million cards, an 18.6% increase over June 30, 2000, when members reported that 400.9 million MasterCard cards were issued. In the critical U.S. market, the number of cards issued at the end of June jumped 20% -- almost twice the growth rate reported for the same period a year ago -- to 256.2 million."
Subprime credit card lender Providian (5th largest U.S. credit card issuer) dropped a bombshell today, stating that increased credit losses and the need to write down uncollectible accrued income would force the company to report significantly reduced earnings. Metris, Capital One, Household and MBNA came under heavy selling pressure as well, in what has been a dramatic change in market view for these lending businesses. This does not bode well for the much more important market in these companies' securitizations. Providian ended the first half with assets of $21 billion and deposits exceeding $15 billion, with $36 billion of total managed receivables. Industry leader MBNA reported a 26% (y-o-y) increase in earnings as managed loans increased $2.2 billion during the quarter to $92.6 billion. All eyes now turn to the enormous market in consumer asset-backed securities
Aggressive Subprime auto lender AmeriCredit reported "better than expected earnings," supported by $2 billion of originations for the quarter, up 45% from comparable originations one year ago. Managed receivables have surged 52% during the past year to $11.3 billion. Managed receivables jumped 11% sequentially (over the previous quarter), in line with Wall Street expectations. In the early 1990's, AmeriCredit's (then named U-Car-Co - "We want to be your car company!") aggressive growth strategy was almost fatal. Rising losses led to a restriction of new bank funding, forcing the company to drastically retrench (including selling its car lots) and restrict lending. The company's fortunes, however, changed dramatically with the introduction and rapid growth of a market for high yielding auto asset-backed securities. No longer would the company's aggressive lending to risky credits be stymied by its nervous bankers, not with the market's insatiable appetite for yield. And with such borrowers much more excited with the prospect of obtaining financing for that car they've always wanted than how much it was costing, combined with irrepressible demand for the asset-backed securities and the use of gain on sale accounting, and it was truly the creation of a Wall Street Love Story.
Surely, few arrangements offer more beautiful and alluring "earnings Mo," and this love bug was contagious as companies lined up to lend to risky credits and have Wall Street create the structures and sell the securities. The key to the infatuation lies with what has been unlimited demand for securitizations, as the subprime earnings game (and the viability of the business) always rests upon the necessity of creating enough new loans to stay ahead of steady flow of old credits turning sour. Providian is now losing this game and Wall Street affections, as its growth slows and losses escalate. For now, AmeriCredit's 50% receivable growth ensures that loss ratios appear reasonable (previous losses divided by current loans). But we do see this quarter that while assets expanded 11% sequentially, charge-offs surged 19%. Delinquencies jumped to 10.8% from last quarter's 9.1%, and we will categorically state that any lender that is expanding loans at a 50% annual rate and in the process experiences almost 11% delinquencies is heading for trouble. We read in one Wall Street research report that U.S. subprime auto loan originations have increased to $100 billion annually, supporting the used-car market as well as the general economy. We've never looked at aggressive subprime lending as a viable business over the entire economic cycle, and our expectations are for significant problems in auto, credit card and home mortgage lending.
Risky lenders such as AmeriCredit really get themselves in trouble when they respond to increasing credit problems by lending more aggressively. Yes, such a strategy does get them through the next earnings season, but it is the unmistakable fodder for a terminal exponential explosion of credit losses. So do I have this right: Over the past year economic prospects have taken a dramatic turn for the worst and AmeriCredit has increased subprime originations by 50%?
We also see trouble brewing for MGIC, a key insurer of mortgage loans. Over the past five quarters, loans in default have increased from 2.01%, to 2.17%, to 2.58%, to 2.67%, to 2.75%, and then to this quarter's 3.14%. The actual dollar value of loans in default has surged 70% to $48.8 million during the past year, while actual "losses incurred" during the quarter doubled (y-o-y) to $43.5 million (losses up 20% sequentially). At the same time, "insurance in force" jumped $8 billion, or at a 19% annualized rate. From a Wall Street report: "During the quarter, of the $23.4 billion in new insurance written, $6.7 billion was bulk business with the remainder from the traditional flow business. Of the bulk business, 61% was comprised of jumbo loans with 39% in the subprime area." "Bulk" insurance is generally for policies on "private label" securitization pools of "non-conforming" (non-GSE qualifying) loans, consisting of "jumbo," lower-tier conforming, and subprime mortgages. MGIC has moved more aggressively into the "bulk" insurance business, with the higher premiums available offsetting increased loan losses. This is another troubling response to gathering economic and industry storm clouds that is certain to lead to major problems down the road. It is worth noting that the delinquency rate in MGIC's "bulk" loan insured portfolio has jumped from 5.90% to 8.73% over the past year. MGIC now has "primary insurance in force" of $180 billion, supported by company total assets of $4.4 billion and shareholders' equity of $3 billion.
Industry delinquency and loss data make it clear that the inevitable period of credit problems in the household mortgage sector has commenced. This is particularly disconcerting with the recognition that notable deterioration has developed in the face of unprecedented mortgage credit growth. Things are turning sour despite the aggressive expansion of mortgage insurance and the historic GSE-led mortgage lending boom, that until recently has stoked continued housing inflation and general economic resiliency. When the mortgage insurers found themselves in trouble with arrival of the late 1980's/early 1990's real estate downturn, some began reporting their losses in an interesting presentation of "old book" versus "new book." The "new book" of business was supposedly created with the use of much improved credit scoring systems and more prudent underwriting. As the economy recovered and the great housing price inflation gathered steam, the mortgage insurance business also became a Wall Street growth darling, and a key partner with Fannie Mae, Freddie Mac and Wall Street financial engineers. The combination created an amazingly powerful engine for economic growth, as risky credits were converted to "Triple-A" rated securities and "money."
We have, of course, often protested and expounded the danger that lurks as the thinly capitalized GSE's come to completely dominate the residential mortgage market (along with a limited number of also thinly capitalized mortgage credit insurers providing the GSEs with the guise of reasonable credit protection on their massive mortgage portfolios). The obvious (at least to us "Credit Bubble analysts") problem with such an arrangement is that when the housing boom begins to show strain, the key players in the Bubble will not be afforded the opportunity to respond to a deteriorating environment by tempering lending and moving to reduce risk. Quite the contrary, in fact, as their role becomes only more supreme as smaller players with the option of reducing exposure back away from additional lending and may even look to unload loans and mortgage paper. But any effort by the GSEs or related key players to temper lending would have immediate negative consequences for the vulnerable housing Bubble, hence their massive loan portfolios and credit exposure. Accordingly, MGIC and the other mortgage insurers are placed very precariously at the frontline of what will surely develop into a mortgage credit bloodbath. The explosion of their insurance has been a key variable in drawing the marginal borrower into the housing market, while Bubble-feedback has inflated the profitability and at the same time masked the growing risks of such credit insurance. Today, any effort to reduce the volume of new insurance or impose more stringent underwriting standards would be immediately problematic, with the key mortgage players forced to continue to expand aggressively (in an environment fraught with significantly accentuated risk) to maintain the inflated value of the collateral supporting their insurance and loans. Their financial largesse is also the key support for an increasingly fragile economy.
The other key issue, particularly for the credit insurers, is that the Fed and GSE-induced mortgage refi boom continues to feed a dangerous feedback loop of credit excess, housing inflation, housing "equity" extraction, and only greater excess throughout mortgage and consumer finance. The credit insurers are left to write increasing volumes of policies on mortgages backed by inflated home prices, as well as the additional factor of unprecedented equity extraction. This insurance boom runs unabated within the general backdrop of extreme mortgage credit creation, unprecedented general consumer indebtedness, faltering economic growth, and rising consumer credit problems. There is absolutely no doubt that writing mortgage insurance at this stage of this extraordinary cycle is a most risky proposition. But despite the dramatic changing risk profile of this type of insurance, these institutions frenetically write policies like it's business as usual.
We have, of course, expended considerable ink on the subject of derivatives and credit insurance. We have argued strongly, and I hope convincingly, that the sanguine view of derivatives as predominantly instruments of "risk management" strategies is misplaced. The proliferation of derivatives, playing a momentous role in fostering unprecedented financial system leveraging, endemic speculation, and the related expansion of availability of credit throughout the economy during this protracted boom cycle, has played a critical role in creating today's extreme systemic fragility (widely available and inexpensive flood insurance has fostered a self-reinforcing building boom along the river). And despite all the hoopla over the role of computer processing power and sophisticated software and modeling, the overriding issue is really the establishment and rapid growth of new insurance markets. Insurance has been around for a long, long time, but it's just the popularity of insuring the risks of faltering stock prices, rising interest rates, and credit losses that has taken the world by storm during this cycle. Traditional insurance companies have for years provided homeowners protection against loss from fire, but why do they to this day refuse to insure against a decline in home prices? Do they know something Wall Street doesn't?
Well, insuring against falling home values is not an "insurable event." To make a certain type of insurance coverage viable, it must be possible to pool a large number of random and independent risks, with actuarial probabilities capable of predicting the future incidence of loss. It must also be possible to charge enough that sufficient policy premiums are available to settle future claims. An insurance company can pool fire protection policies throughout the country specifically because the incidence of fire is random and independent. Coverage for auto, casualty, health and life insurance are similar. An insurance company, however, would certainly not write coverage against a decline in home prices because such a risk is clearly not random (changing home values are cyclical!) or independent (as we have witnessed with the strong correlation of price gains nationally over this boom). The risk of catastrophic losses in the event of a general decline in home prices is too great.
Losses in the financial markets should not be considered "insurable events." Price movements are not random and financial stress in one market is immediately contagious to other markets. There is the further important issue of the "dynamic hedging" aspect of financial derivatives, where losses are not paid from a pool of previously collected premiums, but cash flows are instead generated from active trading in the securities being insured against. Such programs must make the dubious assumptions of liquid and continuous markets, assumptions that have proved quite problematic for huge speculators like LTCM in distressed markets and for derivative traders generally with the WTC attack. We have discussed these issues previously, and will undoubtedly return to them again. But the focus today will be on insuring against credit losses. This is an especially ill-conceived notion, fraught today with extreme systemic risk (financial and economic).
If one considers the traditional example of derivatives so often espoused by proponents - the farmer selling his crops in the forward market to the food processor - it is not difficult to recognize the mutually beneficial relationship between the two parties to such a contract. The farmer desires the security of locking in a price for his production as he embarks on the significant investment of capital and labor, while the food processor also has good reason to fix his raw material prices through forward purchase agreements. If farm prices collapse, the processor may lose on some contracts but his business position is at the same time strengthened by the decline in raw material prices. If prices rise, the farmer may have missed an opportunity this year but is now ready to profit by selling next year's crops forward. And while important, there is another aspect of this type of commercial arrangement that seems to be unappreciated. I would argue that this mutually beneficial commercial relationship tends to maintain the integrity of the market pricing mechanism. If farmers begin to speculate by selling forward large amounts of grain, for example, the excess supply of future grain available in the market above expected business demands of the processor would immediately cause a decline in prices and reduce the incentive for farmers to speculate.
Now let's consider a market in financial derivatives. Instead of actual commodities, we're dealing with contracts based on security prices and future cash flows. And unlike the farmer and food processor (or oil extractor and petroleum refiner), it is no walk in the park to match counter-parties with mutually beneficial commercial interests. Yes, there are a limited number of businesses that benefit from, say, lower equity prices or higher interest rates, but these types of enterprises would be in the distinct minority and would certainly not constitute a sufficient market for accepting the risk of significant market declines. Capital markets by definition create large potential demand for insurance to protect wealth (insure against declining security prices), with the sellers of such insurance necessarily either speculators or sophisticated derivative traders/dynamic hedgers (actively selling securities as necessary to generate cash flows to pay on the insurance). This is a key point. Financial insurance markets differ in some key respects to the farmer/processor forward sale. For one, any significant move in financial markets where there is significant insurance outstanding would be self-feeding and destabilizing, as the dynamic hedgers (and speculators) are forced into trend following trading. This increases volatility on the upside, with obvious potential to foster liquidity problems and create dislocation when hedging programs require aggressive selling into declining markets (to generate the cash flows necessary to pay on escalating derivative losses).
There is another critical aspect of this marketplace that does not get the analytical attention it deserves. Unlike the grain market example where increased speculation tends to be reflected in corrective market price adjustments (market pricing mechanism integrity) with minimal distorting impact on the real economy, demand for financial derivatives are both open-ended and unstable. I would argue that such markets have a proclivity for wild speculative excess, subverting the market pricing mechanism and over time imparting great distortions onto the financial system and real economy. Since the supplier of such insurance is by the very nature of the marketplace a financial player ("dynamic trader" or speculator) and prices are determined by computer-based historical pricing models, as long as increased demand for this type of insurance is being met by increased speculative supply, there is absolutely no natural limit to the size of this market. And, importantly, it is the nature of these types of markets that, in the context of a speculative environment and Credit Bubble dynamics, rising demand for market protection (hastened, itself, by credit and speculative excess) only heightens speculative interest in "writing" market insurance. These are acutely unstable market processes.
Moreover, the expanding markets (and speculative interest) in financial derivatives do have profound and distorting impact on the real economy. Consider the example of heightened speculative interest in, let's say, profiting from the difference in the returns from agency securities and short-term interest rates. As speculators increase exposure to this trade, the "writer" of this derivative acquires mortgage securities in the marketplace. This exacerbates mortgage finance credit availability (and financial system liquidity generally), while further stoking speculative interest in mortgage securities. Credit excess in the real economy fuels financial derivative trading, while increased speculative interest in derivative trading stokes real economy credit excess. And the greater the credit boom the greater the demand for protection against eventual credit problems. Such a situation becomes particularly problematic recognizing the inevitable bursting of speculative Bubbles. I will admit that this is a too convoluted explanation of how the traditional farmer/processor derivative relationship is a stabilizing force for the greater system, while financial derivatives are covertly highly destabilizing and eventually dysfunctional.
I had previously believed the probabilities were high that a dislocation would develop in the interest rate derivative area. With the exponential growth of this market paralleling credit expansion (and financial system leverage), a view that the 1994 experience was but a prelude for a major systemic interest rate derivative dislocation was reasonable (if so far incorrect). But it is important to appreciate that it was the very nature of the amazing contemporary U.S. credit system - with the capability of supplying unlimited credit (and a completely elastic system of money creation) to meet what was extreme borrowing demands - that really subverted the market pricing mechanism. The traditional function of interest rates in facilitating market and economic adjustments became largely inapplicable. And although unfettered credit creation worked wonders in prolonging the boom, the resulting continuation of Credit and speculative excess has been devastating to the U.S. system. These cost are becoming more conspicuous by the week.
Especially at the late stage of the boom, not only did the amount of new credit go to extremes, the quality of this additional debt deteriorated significantly (enormous issuance of telecom and subprime debt, for example). At the same time, it became an increasing challenge for Wall Street to structure these risky credits into readily marketable securities and instruments. The CDO (collateralized debt obligation) market was key, where sophisticated structures could isolate risk to suit the tastes of the aggressive speculator. We've also seen an explosion of asset-backed commercial paper that incorporates liquidity and credit protection to garner pristine credit ratings (and marketability to the money fund complex). For these myriad instruments and structures - the marketability of risk - the proliferation of various markets for credit risk was invaluable. Credit derivatives and insurance, in reality, became the key to sustaining the entire Credit Bubble. And while we see serious problems with insuring market values, we view insuring against credit losses as a sure recipe for disaster.
There is absolutely no way that credit losses are an insurable event. Credit conditions are indeed hyper-cyclical, with minimal and declining incidence of loss during financial and economic expansion. Credit conditions are as well extremely interdependent, with individual credit risks highly correlated and interactive. There is also strong correlation between business credit losses and losses in the consumer and mortgage sectors. And particularly problematic, historical data is keenly deceptive. A protracted period of credit expansion, with corresponding minimal losses, is actually a harbinger of severe credit problems to follow. I have often used a flood insurance analogy to underscore how the availability of insurance against loss incites risk-augmenting behavior. But even this analogy does not capture one dangerous aspect of credit insurance. While flood insurance does lead to more building along the river, at least it does not impact the amount of rainfall. Credit insurance can dramatically alter "weather patterns."
After all, the availability of credit insurance fosters credit creation that specifically fuels economic growth and asset inflation, thereby minimizing credit losses during the expansion. Indeed, powerful Credit Bubble dynamics are at work over the life of the cycle, with the interplay of the increasing profitability of credit insurance, heightened speculative interest in risky securities and instruments, and the natural tendency to finance increasingly risky ventures at the latter stage of economic expansion. Over the course of the boom, then, the amount of credit insurance written directly influences the extent of future credit losses. As we have witnessed over the past few years, the availability of credit insurance impacts both the quantity and quality of debt issued. And in what continues to be an extremely dangerous feedback mechanism, credit insurance has become a key aspect of "structured finance" generally. This then becomes a critical systemic issue with "structured finance" having become key to the U.S. credit system's capacity to intermediate risky lending.
Now that we are firmly in a post-Bubble environment, and especially after the huge financial losses and deteriorating business conditions following September 11th, it is worth pondering the changing circumstance of the credit insurance market. There are clearly huge losses looming out there somewhere, and we would expect the speculative interest in this market has been severely shaken. And returning to our comparison of credit derivatives with traditional insurance, traditional insurance maintains a pool of collected premiums from a large number of policies to pay what are expected to be a relatively predictable small number of random and independent claims. With their reliance on dynamic trading strategies, credit derivatives are similar to equity and interest rate derivatives. As the theory goes, when credit problems manifest, short positions can be placed in the underlying credits, thus providing the cash flow to pay future loss claims. Any sudden and unexpected event such as the WTC attack can be fatal to such strategies, highlighting the dangers lurking with the flawed assumptions of liquid and continuous markets.
So today, with the accelerating downswing of an historic credit cycle (one that included the proliferation of credit insurance), we would expect the derivative industry to now be an anxious seller of credit risk (hedging and offloading their exposure) INSTEAD of the key buyers. If this proves to be the case, then this development has profound ramifications for the entire "structured finance" complex. The key to what has been to this point insatiable demand for risky commercial, consumer, and mortgage loans has been the ability to create structures and readily sell the riskiest components in the securitization and derivative marketplace. If, as we suspect, the market for credit risk has experienced a critical sea change (the speculators and credit insurers transformed from accumulator to liquidator of credit risk) it is not an exaggeration to view this as a momentous financial and economic development. Only time will tell, but there is no doubt to the market's waning enthusiasm (revulsion?) for risky securitizers, as well as all the players along the securitization food chain (lender, insurer, & broker).
Yet we do know today with 100% confidence that acute systemic credit problems have led the Federal Reserve and U.S. financial sector to pursue extraordinary reflationary measures. At some point we would expect this process to have a significant impact for at least one of the two monster derivative markets, interest rate and currency. If this reflation does take hold, we don't even want to contemplate the consequences for the extremely leveraged and maladjusted U.S. financial sector in the event the Fed is forced to "slam on the brakes." Today, however, the greater risk appears to be in the currency derivative area. We do suspect that "structured finance" and currency derivatives have played a key role in sustaining speculative flows into U.S. financial assets. Why these flows would continue considering the alarming developments in the U.S. credit system and economy is difficult to fathom. It makes more sense to us that the faltering market in U.S. credit risk and the ramifications for the securitization marketplace is in the process of changing the whole game. This dramatically alters the risk profile of a large portion of the U.S. debt market and increases the potential for a significant reduction of general credit availability for the maladjusted U.S. economy. There is, as well, the looming issue of the viability of those who have been writing all these derivative contracts.
The huge and growing losses in the credit area come on top of what must be enormous losses in equity derivatives. There are surely also major problems in the important market for syndicated bank loans. Losses in currency derivatives could prove the "back breaker," and I am not sure why the more sophisticated international players would not take such risks seriously today. It is not unreasonable to view recent developments as setting in motion a potential serious dislocation for the entire U.S. credit system. Yet we'll be the first to admit that there is virtually no transparency in the derivative area. During bull markets when confidence is thick, the lack of transparency works as an advantage. The bear is a much different animal, with nerves not calmed by not knowing.
I will conclude with one of my favorite quotes, by novelist and essayist G.K. Chesterton, 1909 (from Peter L. Bernstein's Against the Gods - The Remarkable Story of Risk):
"The real trouble with this world of ours is not that it is an unreasonable world, nor even that it is a reasonable one. The commonest kind of trouble is that it is nearly reasonable, but not quite. Life is not an illogicality; yet it is a trap for logicians. It looks just a little more mathematical and regular than it is; its exactitude is obvious, but its inexactitude is hidden; its wildness lies in wait."