Inside Job (32 page)

These short-term borrowing markets finally collapsed in the late summer and early autumn of 2008, as managers of money market funds and other short-term capital
sources became increasingly nervous about the collapse of the bubble. They stopped lending to the banks, thereby creating a massive heart attack throughout  global financial markets. It
started with Lehman, which was nearly totally dependent on short-term financing. When its lenders stopped accepting even its AAA mortgage paper as collateral, Lehman ran out of money almost
instantly. Its bankruptcy filing trapped tens of billions of dollars that had come from money market funds. Unable to get access to their cash, money market fund operators had to dip into their own
reserves to maintain their sacred $1-per-share price; the Reserve Fund, one of the largest and oldest, “broke the buck”. This was one of the events that Paulson and Bernanke should have
predicted but didn’t. It caused a huge run on money market funds, as depositors withdrew their funds in panic. The money market funds therefore stopped lending money to banks and industrial
corporations. This immediately threatened the solvency of every major investment bank, until the Federal Reserve stepped in and pumped oceans of funding into the banking system.
⁵
The Treasury Department also announced that for the first time, the federal government would guarantee assets in money market funds.

In other words, excessive dependence upon short-term funding is dangerous. In crisis conditions or even moments of stress, it creates volatility that can rapidly endanger the system. This danger
was multiplied exponentially by excessive leverage. The enormous growth in assets during the bubble was permitted only by enormous borrowing.

Since deregulation began, total leverage in financial services has increased astronomically, and frequently in ways that are invisible to regulators, customers, and
lenders.

Once the Glass-Steagall wall between commercial lending and
broker-dealers was torn down, financial conglomerates such as JPMorgan Chase, Citigroup,
and Bank of America started to take advantage. The investment banking and trading side of these institutions now had a vastly larger capital base to support its buying and trading. Then, in 2004,
rigid numerical leverage constraints on investment banks were eliminated, replaced by the banks’ own estimates. That change alone allowed the five largest investment banks to increase their
leverage by about 50 percent, which was deplored even at the time. At the SEC meeting that approved the change unanimously, one commissioner is heard saying, well, we all know that these are the
big guys, so if it goes wrong, it will be a really big mess—a comment followed by nervous laughter.
⁶

But the leverage that was shown on bank balance sheets was only a portion of the real leverage in the financial system. There were several reasons for this. First, perhaps taking a cue from
Enron, which used off-balance-sheet entities to conceal risk, the banks created many new special purpose entities (SPEs), structured investment vehicles (SIVs), and the like, which were used to
transfer risky assets off their books. They lobbied heavily for these entities to receive favourable accounting treatment by the Financial Accounting Standards Board (FASB), by auditors, and by
regulators, and they won.

Such obfuscations were being used more and more frequently by the mid-2000s. Citigroup wins the honours for the sheer scale of the numbers involved. Citigroup’s closing 2007 balance sheet
showed total assets of $2.2 trillion, of which $1.5 trillion—in loans, trading assets, and financial investments—were “risky” assets. Deep in the financial footnotes,
however, we discover that Citi had
another
$1.1 trillion in risky assets in addition to those on the balance sheet. Its true balance sheet was therefore 50 percent larger than the one shown
on its books, and its exposure to risky assets 73 percent larger. Of the $1.1 trillion in undisclosed assets, moreover, many were among the worst produced during the bubble.

The logic of the off-balance-sheet treatment was that Citigroup and the other banks had “sold” the assets to new entities that bore the risks of any defaults. But like so much bank
accounting during the bubble,
that was misguided. First, in order to reassure investors, Citi had frequently guaranteed these entities against losses. Buried in the footnotes
to its financial statements one would see allusions to commitments implemented by “writing a liquidity put or other liquidity facility . . . [and] guarantees, letters of credit, credit
default swaps or total return swaps,” which shifted risk from these entities back to Citigroup. In other cases, there was no legal guarantee, but there was a complete expectation by
investors, encouraged by Citigroup, that losses would be covered in a way similar to the US government’s implicit guarantee of Fannie Mae and Freddie Mac.

Merrill Lynch had its own issues. According to the company’s accounting,  2005 and  2006, the peak years of the financial bubble, were the best in
Merrill’s history, with revenues of over $26 billion in 2005 and over $34 billion in 2006. Since Merrill paid out roughly half of all revenues to employees, those were also years of
record cash compensation—$12.4 billion and $17 billion respectively.
⁷

But over the next two years, Merrill
lost
more money than it had made in all the years since 2003 combined. Merrill held toxic assets whose value collapsed in the crisis. In the end, for
the entire period of 2003 through 2008, Merrill Lynch had a total of $100.4 billion in revenues, and paid $80.4 billion in compensation, but lost a cumulative $14.6 billion. Merrill paid 80 percent
of its real revenues in compensation because many of its earlier supposed revenues were only on paper, so they evaporated, while the bonuses based on them were converted into very real
payments.

In October 2007, both Merrill and Citi took very large losses, and both firms “accepted the resignations” of their CEOs. In both cases, losses were primarily related to the forced
repatriation of their SPEs and SIVs to on-balance-sheet status. In most cases, they had no choice, for the guarantees they had written required them to do so.

Lehman and several other banks also used garden-variety accounting tricks to conceal the extent of their leverage. Using the Repo 105 trick mentioned  earlier, Lehman artificially
“sold” and then “repurchased” stuff just before and after each quarter ended, making the firm
seem less leveraged than it really was. They found a
British law firm, Linklaters, that signed off on this, since it was performed in Europe, where it was within the rules—advice the firm later said they intended to apply within Europe only. At
the end of the bubble, Lehman admitted to being leveraged over thirty to one; due to its Repo 105 trick, as well as various off-balance-sheet entities, its real leverage was even higher.
That’s where American investment banking was in 2008.

As deregulation progressed, the US financial sector became far more concentrated, both within markets and across them. By the time the bubble started, American financial
services were dominated by five independent investment banks, four huge financial conglomerates, three insurers, and three ratings agencies. Several firms—certainly Goldman Sachs, AIG,
JPMorgan Chase, Bank of America, and Citigroup, even perhaps a number of smaller firms such as Lehman Brothers, Wells Fargo, and Morgan Stanley—were sufficiently large that their collapse
would endanger the entire financial system. In other words, they were too big to fail.

This condition was exacerbated by the fact that many individual markets were and remain even more concentrated than the whole industry. Five institutions control over 95 percent of all
derivatives trading worldwide, and two—Goldman Sachs and JPMorgan Chase—control nearly half. Three US rating agencies dominate that market globally. A group of about a dozen banks
controls the LIBOR—the rate used to set nearly all short-term interest rates—and has been accused of colluding to manipulate this rate. The top five investment banks dominate the market
for initial public offerings, frequently share portions of such offerings with each other, and charge exactly the same fees. Joint ventures established collectively by all the major investment
banks dominate several markets related to the bubble; for example, the indexes of
mortgage securities used as references in constructing many synthetic CDOs. And so on. As a
result, many important markets are subject to cartelistic behaviour, which can also destabilize them.

Furthermore, some of the biggest banks are also, one fears, too big
to succeed
.
⁸
  They are so large, so complex, with so many disparate
activities, that it became impossible even for well-intentioned CEOs to monitor all of them. Citigroup is perhaps the most obvious case; Chuck Prince was hired in large part to clean up
Citigroup’s legal and regulatory problems dating from the Internet-Enron era. He did that. But he was a lawyer, and he never looked hard enough at what the investment bankers were doing
in the mortgage market. He may have trusted Robert Rubin to watch that. And measured by revenues rather than potential risk, Citigroup’s mortgage business was only a small portion of the
bank, easy to overlook. Thus, while the failure of Bear Stearns and Merrill Lynch could be attributed to grotesque failures of management and corporate governance, the failure of Citigroup was
partially due to sheer size. One wonders how JPMorgan Chase will fare once Jamie Dimon retires, and whether Bank of America can ever be managed effectively now that it has swallowed Countrywide,
U.S. Trust, and Merrill Lynch.

Information technology and globalization, while enormously beneficial in many ways, have a dark side. They have given rise to a level of cooperation, and therefore
interdependence, between many disparate institutions never before seen in economic behaviour. And the cost of hyperefficiency is system fragility. We see it in how a tsunami in Japan disrupts the
operations of electronics and car companies half the world away, in how twenty terrorists can shut down the US for a week, and we also see it in finance. In the pre-Internet, pre-deregulation
financial system, the failure of any one institution or market did not generally endanger everything else. That is no longer true.

Astonishingly, it was not until the Lehman failure in September 2008 that senior government officials—Treasury secretary Hank Paulson, Fed chairman Ben Bernanke, New
York Federal Reserve president Tim Geithner—apparently realized how interconnected the global financial system had become. Perhaps they had believed their own propaganda. The banks insisted
for years that they had off-loaded much of their risk to wealthy investors, hedge funds, pension funds, and the like. Derivatives such as credit default swaps were said to be particularly wonderful
for reducing risk. But the risk shedding was a mirage. If you followed the money and the risks all the way through the system, in the end they came right back to the banks. And once a crisis
started, it could move through the system much faster and more violently than any human being could understand or manage.

Richard Bookstaber, an engineer and economist with deep experience in finance and risk management, termed this phenomenon “tight coupling”. A tightly coupled system is one in which
problems propagate faster than managers can react, and where systemic interdependence means that disruptions can rapidly become catastrophic as they progress. In his brilliant book
A Demon of
Our Own Design
, written
before
the crisis and published in 2007, Bookstaber warned that modern electronic finance was dangerous in this way. He examined a series of both nonfinancial and
financial disasters, seeking their common features. He found that increased velocity and efficiency created a danger that human regulators would be overtaken by events. He also found that measures
that supposedly increased the safety of one institution or part of the system could have the opposite effect on the
whole
system, particularly if widely adopted. One example he considered
was the role played by “portfolio insurance” in causing the 1987 market crash. More recently, think credit default swaps and AIG; selling mortgage assets when the panic started; or
withdrawing money from a money market fund in September 2008.

It is probably pointless to second-guess most of the behaviour of the people at the US Treasury and the Federal Reserve in September 2008, for they were making decisions under great pressure and
with poor
information. What is to be condemned is the way they behaved before and after the catastrophe. Messrs. Greenspan, Bernanke, Paulson, Geithner et al. could have
retained their blissful, oblivious happiness in 2006–2008 only by willfully ignoring masses of contrary evidence. The dangerous financial crises that America had experienced since the early
1980s were both recent and highly visible. And what have these gentlemen advised since? Nothing very impressive.

The chart immediately below, vastly oversimplified as it is, captures some of the relationships that lead to tight coupling within the modern financial system.