Fact and Fiction Behind Too Big to Fail
Not all large banks designated as “systemic” by Dodd-Frank expose taxpayers to investment banking’s trading risks. Only the largest are the ones packaging and trading derivatives and were “securitizing” and selling residential mortgage-backed securities.
Securities trading, historically of bonds and equities, has always been an investment banking strength. However, financial engineering created new, more opaque securities to underwrite and sell. Securitizations and derivatives were the trading activities that expanded into a bubble, leading to its catastrophic collapse. The taxpayer ultimately provided the financial backing that prevented each trading activity’s risks from bankrupting these largest banks, tying the government guarantee to trading risks.
Financial Engineering Exacerbated Trading Risks
In the 1980s and 1990s, the largest commercial banks pushed to get into securities underwriting. They had been limited to lending by Glass-Steagall, which restricted securities underwriting to investment banks. Large European banks could engage in both.
Securities underwriting is a short-term, market-driven activity. An investment bank purchases a large amount of securities (bonds or equities) from an “issuer”, a corporation or government entity. The underwriter locks in the interest rate on the bond issuance or the price of a share for a stock issuance. The underwriter assumes the market risk associated with the subsequent sale of the securities to retail and/or institutional investors. The security’s sale transfers market risk to the investor. It is no longer retained by the underwriting investment bank.
In earlier decades, the two broad categories of securities — bonds and equities — were rather well understood, even by the public. The details of the borrowing or stock issuance were presented in the corporate issuer’s annual report, along with all sorts of other financial information useful to investors.
The introduction and growth of financial engineering changed the transparency and comprehensibility of securities underwritten by investment banks. Financial engineering was born out of profit pressure and self interest. Its invention can be rationalized in that it provides finer tuning to risk management and better matching of investment products to investors’ needs. That was certainly its reasonable intent. While financial engineering expanded risk assessment processes, borrowing mechanisms, and investment products, many devils were in its details. The primary financially-engineered products, securitizations and derivatives, reveal many of these difficulties.
Securitizations widened the intellectual distance between borrowers and lenders. Murky or unavailable information, shady or non-existent documentation, and multiple series of legal entities between borrower and lender created much more complex securities to sell. But when the seller isn’t responsible for the borrower’s ability to repay, it has little incentive to care.
The new legal entities created by investment banks for securitizations weren’t businesses with humans running them. The borrowers weren’t a single company. They were thousands of residential home mortgage borrowers who knew nothing about their mortgage’s sale other than being told where to send the monthly mortgage payment. The government through its Fannie Mae and Freddie Mac entities was a presence within many securitizations, providing a “halo effect” of government support. Government Accountability Office (GAO) reports as early as 1990, typically requested by Congress, warned of the risks of these poorly-supervised Government-Sponsored Enterprises (GSEs) to the taxpayer.
The new legal entity issued its own securities, sold to the public by the investment bank. No financial firm retained any interest in the residential mortgage security on its own balance sheet once it was sold. Investors didn’t really know their borrowers.
When the investment banks had fully exploited the traditional mortgage potential, they turned to Alt A, liar loan and other subprime sectors of the residential mortgage market to keep volumes crossing their trading desks. Then they underwrote insurance against the default of the engineered mortgage securities.
Competitive Pressures Grow
Relying on outmoded formulas and histories, ignoring the big picture, the rating agencies vastly assisted the investment banks. They provided high credit ratings despite increasing leverage across the residential mortgage sector. Folks always make their monthly mortgage payment, don’t they? The high ratings helped the biggest banks reduce the amount of regulatory capital they were required to hold.
The scorecards for securities underwriting are called the “league tables” — the rankings of underwriters by volume in each category of security. Two famously-failed investment banks which packaged and sold residential mortgage-backed securities, Lehman Brothers and Bear Stearns, led the league tables in the final years before the subprime and mortgage market collapse. They are now history. But the securities underwriting arms of the remaining largest, excluding Wells Fargo, were also on the league tables.
Hidden Trading Desk Promises
While most of the risk involved in mortgage securitization was moved off the books of the underwriting bank when the security was sold, the story didn’t end there. These securities were sold with “representations and warranties” that their quality was as described in the sales prospectus. If not, the purchasing investors could demand the repurchase of the securities by the investment bank. The numerous subsequent lawsuits against the largest banks plus the government’s own actions against them resulted in hundred-million-dollar settlements and even billion-dollar settlements.
Outstanding residential mortgage securitizations were $7.4 trillion at the end of 2007 and fell to $3.1 trillion by the end of 2010. That’s a lot of securities selling backed by promises to repurchase if things aren’t what they were said to be.
The unleashing of financial engineering backed by government support and weak oversight pumped up housing prices. It was a steep spike and collapse. Click on the next link to see the picture of the frenzy, courtesy of the Big Picture blog with some “looking for the price floor” by the New York Times.
Trading: Profitable Until Bubbles Collapse
Trading revenue histograms illustrate the significance of systemic risk. Bank holding company annual reports on SEC form 10-K now contain this chart. In the years outside the bubble’s final collapse (2008) these histograms show most trading days resulted in profits. In 2008, it was vastly different.
While all the largest banks with big trading desks present similar pictures, I’ll show Citigroup’s and Goldman Sachs’s, printed from their 10-Ks. Click on the first link to see Citigroup’s 2016 10-K’s daily trading revenue histogram and on the second to see 2008’s. Take a look at the scale labels. Do the same for Goldman’s for three years including 2008. In normal times, they usually make money. When the bubble burst, they experienced significant trading losses.
The Biggest Banks Get Paid While Other Banks Get Stuffed
One behind-the-scenes story little viewed by the public was the sale of $7 billion of GSE Fannie Mae preferred stock on December 6, 2007. By that date the subprime market was collapsing. Credit ratings were plunging on mortgage debt.
Fannie Mae had only two classes of securities — their bonds and this preferred stock. The bonds were rated AAA/Aaa. The rating agencies assumed that the U.S. government would back the debt despite the lack of an explicit government guarantee because Fannie was sponsored by the government for government purposes. Fannie’s preferred stock was also highly rated — AA-/Aa3, according to the offering circular. With the high credit rating, many small banks and even some of the largest purchased the preferred stock. They assumed the government would also support the preferred stock in the event of Fannie’s collapse.
The government didn’t. It took over Fannie, but stopped paying dividends on the preferred stock. The price of the preferred stock plunged.
In a September 22, 2008, letter to government financial officials (click on the next link to see a copy), American Bankers Association President Edward L. Yingling pointed to the severe impact this action had, particularly on small community banks. Nearly 27% of banks held preferred stock of GSEs. The estimated exposure to GSE preferred stock was between $10 and $15 billion. “Since $1 of bank capital supports $7.6 dollars of lending, the actions contemplated mean that lending could decline by between $76 and $114 billion. The credit crunch will be immediate: with capital difficult to raise in the market today, banks will have no choice but to shrink in order to restore their capital-to-assets ratio to previous levels.”
The offering circular’s front page shows which institutions were the underwriters of that preferred stock sale. All five active in investment banking were there. The underwriting group was paid a $70 million fee.
The even more direct government support of derivatives is explained in other posts. Suffice it to say that the taxpayer guarantee inserted in 1933 was not intended to subsidize the trading activities of investment banks. But the huge, formerly-commercial banks sure made money after the removal of Glass-Steagall barriers between trading and the deposit guarantee. Click on this final link to see the profits grow. Until they didn’t.