Guest post: The real cause of the America’s housing bubble was foreign money

In this guest post, Martin Lowy argues that America’s housing bubble couldn’t have inflated to dangerous proportions without massive inflows from Europe.

The crisis was built in just three years: 2004 through 2006. If no new financing sources had been added to the housing market after 2003, nothing extraordinary would have happened even if house prices had declined.

What made those three years extraordinary was the influx of foreign money into securitized mortgage-based products. That inflow enabled mortgage money to be advanced to people who couldn’t repay it.

Foreign money knows little about the domestic market and therefore relies on local banks, governments, or rating agencies. Investment banks seized on that weakness of the foreign money to use weaknesses in the system to promote badly underwritten mortgage products and pass them off as money-good.

That is what drove the final price increases and final over-financings to weak credits that, when prices turned down and credit went back to normal underwriting, caused the Great Recession.

Simply put: no foreign money, no fraud to convince the foreign money to come in, no bubble and no bust. It is sort of like the legal doctrine of proximate cause. These factors are the last “but for”. Thus the fraud committed in the mortgage loan process and in the selling of the mortgage-related securities was the proximate cause of the crisis.

There are three ways to demonstrate this:

The magnitude of the private label securities market and corresponding increases in house prices in the key years.
The magnitude of foreign purchases of securities based on those private label securities.
The pattern of fraud that permitted the packaged mortgages to be sold at prices that supported otherwise defective mortgages.

Start by looking at the types of mortgages originated in the 2000s and how that changed over time. The following table comes from p.154 of my 2009 book Debt Spiral and is based on data from Inside Mortgage Finance:

Note the changes in private label securitisation. The high levels of originations in 2004 through 2006 were due principally to private label (as opposed to Fannie and Freddie guaranteed) securitised product. Private label securities surged as a share of originations in those years — approximately double the highest prior year — at the expense of Fannie’s and Freddie’s market share. Private label securities issuance collapsed by 2008.

The reason the private label market took share away from other origination channels was that the loans were pushed by the brokerage and banking industries in exchange for high fees. They were attractive to borrowers because of low teaser rates, low down payments, low underwriting standards, and opaque fee structures that hid the true cost of the loans.

The next excerpt from Debt Spiral (p. 123) shows the extent of home price appreciation that took place in the four boom Sunbelt states (California, Arizona, Nevada and Florida) in the critical years 2004-2006:

Prices jumped in all four states in 2004-2005 and fell precipitously in 2008. The steep rise started earlier in California.

None of this would have happened without foreign demand for these private-label mortgage bonds.

The Chinese government bought debt issued and guaranteed by Fannie and Freddie while European banks and European-based conduits established by European and American banks went for the toxic stuff.

(I do not believe that the Chinese government purchases of Fannie and Freddie securities were significant to the mortgage market. See my 2017 book Instability under the heading “Foreign lending into the boom” for a further explanation.)

Without the European banks and their affiliates, the market would have been too small to sustain the massive increases in private label mortgage product. Michael Lewis wrote about this phenomenon in The Big Short. And the data support Lewis’s story about the European (and particularly German) banks being the key to the market.

The following graph is based on Bank for International Settlements data. It shows how much the major European banks invested in dollar assets in the key years 2004-2006:

Total assets grew by about two trillion dollars in that period. Not all of it went into the US housing market, but a substantial part did go there—enough to absorb a material part of the three trillion dollars of private-label mortgage bonds issued in the mid-2000s.

European bank involvement was encouraged by bad regulations: the miniscule capital requirements for triple-A corporate securities and American regulators’ response to accounting pronouncement FIN45. That rule allowed US banks — and through adoption by the Bundesbank, German banks as well — to guarantee debts of SIVs and similar conduits without counting the guarantees against capital requirements.

All of this “dumb money” flowing into the US housing market encouraged shoddy underwriting standards. Many mortgages could not have been created and sold without fraud being committed by the mortgage brokers, mortgage bankers, aggregators, and underwriters of the eventual securities, aided and abetted by the ratings companies. Some of that fraud was accomplished merely by a wink, other parts of it were more active. But the system flourished for a brief period of three years based on dividing up the tsunami of fees that the mortgages generated.

A chart from CoreLogic published in their current “Housing Credit Insights” illustrates how inferior the 2004-2006 loans were in historical terms:

Creating and selling such poorly underwritten loans required cooperation among a chain of service providers.

The mortgage brokers were lowest on the food chain, but they were a crucial element because they talked the borrowers into the high-fee mortgages and helped the borrowers to falsify the financial data in their applications, in many cases by simply filling out the applications and presenting them to applicants to sign.

The mortgage brokers knew what the applications needed to say because the aggregators nearer the top of the food chain gave them periodic descriptions of the mortgages that would be acquired for sale in the securitized market. Mortgage bankers also played along with the aggregators’ postings, and the result was billions of dollars of mortgages that met the requirements.

The hardest part of the scheme, however, was selling the securities that were backed by mortgages that did not meet traditional underwriting criteria. That job was critical because without an end market, the mortgage banker funding for the mortgages would dry up.

Substantially all of the major underwriters—and some of the lesser ones—participated in this lucrative market. The basic game was to package the mortgages so that large parts of the packages would earn high ratings from the semi-official oligopoly. The ratings enabled the underwriters to sell the highly-rated pieces to European banks, structured investment vehicles established nominally in Ireland by American and German banks, and other institutions. To accomplish these sales, the underwriters had to mislead the rating agencies and the buyers of the securities concerning the nature of the underlying mortgages. I tend to believe that the rating agencies were complicit as well.

How do we know that the major underwriters misled investors during that period? We do not know that directly—at least not yet. The major underwriters have settled the cases for tens of billions of dollars rather than going to trial and exposing themselves to public scrutiny.

Two late-to-the-party copycat non-U.S. underwriters, however, have not had such wisdom. Nomura and Royal Bank of Scotland (RBS) went to trial against the supervisor of Fannie and Freddie. And Nomura and RBS lost at trial. The U.S. District Court opinion is long and detailed. It convinces me that Nomura and RBS did intentionally mislead investors and that they were following a pattern of conduct that had been laid out by the major underwriters before them.

RBS and Nomura have appealed but the appellate briefs are damning in themselves. The fundamental premise of the appeal is not that the mortgages were properly represented in the offering documents, but that Fannie and Freddie were sophisticated purchasers who knew that the mortgages did not meet the stated criteria. The appellants may win on that ground, but they still would be guilty in the court of public opinion.

I am still working on how the sales worked. But there is enough in Michael Lewis’s The Big Short, the District Court opinion and elsewhere in the public domain that I am confident that securities fraud was a key component of the salesmanship.

Without the debt-fuelled house price increases in the key Sunbelt states in 2004-2006, the stage would not have been set for a crisis. The market would have calmed down, there would have been some defaults and foreclosures, but the securities markets would not have been involved, Fannie and Freddie would not have been taken into conservatorship, the Fed would not have had to open vast swap lines with European central banks or back the value of many other securities, and the bank bailouts would have been unnecessary.

That said, without a bubble there would have been a cost: 2004 through 2006 would not have seemed like boom years in America.

Martin Lowy has seen banking from many sides: as a lawyer for banks (partner at Hughes Hubbard & Reed and at Gibson, Dunn & Crutcher), as a senior officer of a bank, as an outside director of a bank, as a staff member of advisory committees to regulators, and as lawyer for the FDIC and for the NY Superintendent of Banks. He is the author of High Rollers: Inside the S&L Debacle, Practical Handbook for Bank Directors, Debt Spiral: How Credit Failed Capitalism, and InStAbILItY: Booms, Busts, the Fragility of Banks, And What To Do about It.