Rising loan defaults are a normal feature of any credit cycle. What makes the past few years different is the degree to which derivatives and aggressive loan securitizations have spread the risk around, beyond the financial institutions which historically have specialized in creating and managing such illiquid risks. According to the FDIC, in 2005 almost 68 percent of home mortgage originations were securitized. The US bank insurance agency reported last year:
"A significant development in the mortgage securities market is the recent and dramatic expansion of “private-label” MBS, which are securitized by entities other than the GSEs and do not carry an explicit or implicit guarantee. Total outstanding private-label MBS represented 29 percent of total outstanding MBS in 2005, more than double the share in 2003. Of total private-label MBS issuance, two-thirds comprised nonprime loans in 2005, up from 46 percent in 2003. With the increased exposure to private-label MBS and a large share of higher-risk nontraditional mortgages being securitized in this sector, investors appear willing to assume greater risk in their search for yield."
Just as bad trades in the natural gas market took down a hedge fund formerly known as Amaranth, equally bad asset quality inside sub-prime mortgage and other consumer loans, which have been packaged and sold in vast quantities into the fund community, potentially represent time bombs that could damage or even kill dozens more hedge funds.
First consider the restatement by HBC and how the bank's credit profile illustrates the difference between sub-prime and mainstream consumer credit. We've noted in past missives that HBC and sub-prime lender peers such as Citigroup (C) are unlike other money center peers in terms of default rates and other aspects of their portfolios. Even excluding the sub-prime Household Finance business, HBC's $170 billion in US banking assets were 1.7 standard deviations (SDs) above-peer in terms of loan defaults, as shown by the Basel II summary public data benchmarks below.
In the bank-only profile shown above, actual metrics for defaults, Loss Given Default, Exposure at Default and Weighted Average Maturity or "M," are shown for HBC, with SDs shown for the peer group comparison. HBC comes in well-above peer in most respects with the notable exception of M, which like C (1.6 years) is very short compared to other money centers. Note that EAD at 272% also is high, some 2 SDs above peer, but this is not unusual for a consumer lender.
Now when you look at HBC's US business on a consolidated basis, however, using the data from the Fed's Form Y-9, the credit risk disparity between the sub-prime business of HBC and other large bank holding companies becomes more apparent. When Household Finance and other non-bank businesses are added in to the calculation, HBC's consolidated assets rise to $456 billion as of the end of the third quarter of 2006, making HBC the sixth largest banking group in the US.
HBC's net loan and leases losses in the US for the first nine months of 2006 were 183 basis points (1.83%) vs 105bp for C and just 21bp for the large bank peer group. Provisions for loan and lease losses originally reported by HBC were 120bp vs 38bp for C and just 14bp for the peer group, according to Q3 2006 data from the Fed. But we know now that the Q3 disclosure by HBC understated the actual loan portfolio credit risk -- not to mention contingent risks from securitized loans.
One reason investors were spooked by last week's HBC restatement is what it implies for future valuations across the financial services sector, both for on-balance sheet risk and for risk which many banks previously sold. Wish we could be a fly on the wall for the discussion between HBC management and the bank's external auditors, KPMG, as to the appropriate level of provisions for Q4 2006 and beyond. But HBC is not the only bank with this problem.
Over the rest of 2007, we fully expect to see most of the major money centers announce higher loan losses and provisions for retail mortgage portfolios, and some banks may even be forced to restate previous periods. But the real threat to all of the major US banks involved in significant asset securitization lies in the probability that many of the collateralized loan structures employed to shift risk off bank balance sheets will unwind.
As Jody Shenn of Bloomberg wrote this week:
"Subprime loan buyers typically can force lenders to buy back the mortgages they sell if borrowers miss their first few payments, any type of fraud is discovered, or the loans otherwise fail to meet the guidelines laid out in a sales contract."
This is true even if the loans were packaged into a collateralized debt structure or CDO, anointed with a credit derivative enhancement from a hedge fund, and blessed with an explicit credit opinion from a rating agency, before being sold to yet another hedge fund.
Jamie Dimon, chief executive of JPMorgan Chase (JPM), disclosed last week that JPM held only $5bn of higher-risk sub-prime loans, just two per cent of its total retail portfolio. He then bragged that the bank had sold much of its mortgage exposure -- but to whom? Fact is that JPM, the largest derivative dealer on earth, likely sold much of its loan exposure to its hedge fund clients, highly leveraged entities that have significant clearing and credit exposure to JPM.
As the wheels start to come off of the mortgage collateral wagon in 2007, a number of money center banks and broker dealers, particularly the ones with large prime brokerage operations, may be forced to repurchase CDOs from hedge funds, mutual funds, banks and other clients who discover to their dismay that there is no bid for this paper, credit agency rating or no. This situation will be particularly poignant for JPM, which seemingly was the proximate cause of the Amaranth hedge fund failure and even profited from the fund's demise, as we wrote in a previous issue.
Question: Why do you suppose that neither the asset managers at Goldman Sachs (GS) nor Morgan Stanley (MS) have filed a legal claim against JPM for the Amaranth collapse? According to our colleagues at Institutional Investor, quoting Gas Daily, MS Alternative Investment Partners Absolute Return Fund lost two-thirds of its $4.3 million allocation due to the Amaranth collapse, while its Institutional Fund of Hedge Funds lost 55% of its value. II also reports that GS, in its Nov. 14 filing with the SEC said only that its Global Relative Value Fund, with $95.6 million, "experienced negative performance for the quarter due primarily to significant losses in energy-related investments by a single GRV advisor, Amaranth Advisors, following a dramatic move in natural gas prices."
While Amaranth's failure arose from losses in the commodities markets, the example seems relevant to a discussion of the relationship between a prime broker and a hedge fund regarding other types of investments, say sub-prime loans and CDOs. Indeed, in the wake of the Amaranth fiasco, we hear that some of the larger hedge funds have demanded and won new credit facility provisions which essentially turn conventional credit lines into term loans, with no reset or call provisions to adjust for changes in collateral value. This means that a hedge fund holding a portfolio of rancid, sub-prime CDOs could effectively force a dealer bank or BD to finance same indefinitely, even though there is no cash bid for this entirely unique, opaque collateral.
As and when a dealer is forced to bite the bullet and repurchase the CDO from a recalcitrant customer, doubtless at or around par value, the dealer then will be forced to immediately mark-down the collateral and take a loss -- assuming the auditors are paying attention. The alternative for the dealer would be to pull the fund's credit line and potentially force yet another fund collapse. Either way, the effective moratorium on litigation which seems to have prevailed in the wake of the Amaranth collapse will end and the full ugliness of the CDO marketplace will become increasing transparent. Stay tuned.