Wednesday, August 18, 2010

Causes of the Crisis: Financing of the Shadow Banks

And so we return to the old question, how does the shadow banking system finance its operations? One way, obv iously, is by selling its financial products to investors. But since ultimately the finance system is a system of making loans, it has to raise the funds to make loans somehow. There appear to be two main answers. One way is by "selling commercial paper," that is by taking out a series of short-term loans (known as commercial paper). Securitizers, in particular, used this method, using their securitizations as collateral.

These short-term loans came mostly from money market mutual funds, the most user-frienly portion of the shadow banking system. Money market funds accept"deposits" by selling shares for exactly $1, no more and no less. Shares are kept at $1 so that investors can cash out at any time for the same amount they paid in, like closing out a bank account. They pay interest and dividends to investors, usually at a higher rate than a certificate of deposit at a conventional bank. Some even offer checks like conventional bank accounts. But they are not insured. To avoid losing value, money market mutual funds limit their "commercial paper" loans to municipal governments, blue chip companies, and the like. Unlike in conventional banking, ordinary customers do not make direct deposits, but have intermediaries make their investments. Direct deposits generally come from local governments, corporations, investment institutions, and high net worth individuals. However, the Federal Reserve comments that even the most sophisticated investors in money market mutual funds have no tolerance for risk and are prone to "herd-like" behavior.

The other main source of funding are "repurchase agreements" or "repo" for short. As everyone knows, the FDIC insured conventional bank deposits for up to $100,000 before the 2008 crisis and up to $250,000 since. For ordinary depositors this is not a problem. Anyone with more money than that will probably not want to keep in in cash anyhow, but to invest it in something that offers better returns. But large corporations, banks, and so forth, may have large blocks of cash between investments, much greater than anything the FDIC would insure. (Nor are conventional banks equipped to handle such large deposits and withdrawals). But they still want somewhere short-term and risk-free to park it. Typically, the owner makes a one-day loan to an investment bank or other shadow bank and takes some safe financial instrument, such as a senior tranch security, as collateral. The short-term nature of these loans keeps them safe and liquid. Each day, the lender has the option of rolling over the loan or withdrawing it (allowing the bank to "repurchase" the collateral). Like ordinary depositors, repo lenders often close out all or part of their loan. So long as new lenders are opening new accounts and an equal rate, this is not a problem for the bank. Unlike ordinary depositors, repo lenders are powerful enough to change the terms of their loans as a condition of rollover.

All these funding sources -- investment in money market mutual funds, the short-terms loans known as commercial paper, and repo agreements -- have in common with bank deposits that they are risk-adverse and can be closed at will. A leading argument, therefore, is that when securities started looking shaky, these lenders did just that, setting off a run on the shadow banking system. Most ordinary people did not recognize the run because their own bank deposits were insured, and because it had none of the traditional features of a bank run -- long lines outside the bank, panicked mobs in the street and so forth.

The run began (so far as I can tell) in mid-2007 on banks' structured investment vehicles (SIV's). These were subsidiaries of banks, established off their balance sheets so as to evade capital requirements, that invested in securitized loans, purchases with "commercial paper" loans from money market mutual funds. As these investments came more and more into doubt, money markets stopped making loans. SIV's were experiencing a run. The Treasury Department tried to pool all the SIV's into a single SIV, but banks were not interested. Instead, they took the SIV's back onto their own balance sheet. A portion of the shadow banking system had vanished.

It was about this time that repo lenders started getting nervous. Up until then, repo lenders accepted the bank's collateral at face value. That is to say, they would deposit one dollar of repo for every dollar in collateral. Around August, 2007, repo lenders stopped taking collateral at face value. Instead, they would loan only 90 cents for every dollar of collateral. This is called a 10% "haircut." As the crisis escalated, haircuts kept rising, to over 45% when Lehman Brothers failed.

Nor was this haircut limited to sub-prime mortgages. All housing securities were suspect because they were based on grossly inflated prices. But the problems were not limited to real estate. All securitized loans became suspect. Lenders demanding more and more collateral effectively means that the system is getting less and less funding. If not a full-scale run, it is suffering a severe pinch.

But the most devastating part of the crisis, the part truly threatening a run that would threaten the entire system, occurred in the aftermath of the Lehman Brothers failure. When Lehman Brothers failed, Reserve Primary Fund, a money market mutual fund, was forced to write off Lehman debts that "broke the buck," i.e., caused its shares to fall below $1 (for only the second time in the history of money market mutual funds). This set off a panic by depositors, who had always assumed a money market was as safe as a bank account, and the beginning of a run. And, as everyone knows, the greatest danger of banks runs is that they are highly contagious. A run on money market mutual funds threatened to put an end to major corporations' normal system of financing day-to-day operations, and also threatened day-to-day funding of municipal governments, to say nothing of funding shadow banks. It drove interest rates on commercial paper from 2% to 8%. Desparate, the Treasure Department moved to insure all existing money markets. So far as I can tell, it was the prospect of a run on money markets more than anything else that convinced Chairman Ben Bernanke and Treasury Secretary Hank Paulson to request, and Congress to pass, the TARP.

So just how bad was the withdrawal of funding from the shadow banking industry?

This graph gives a good estimate. Most striking is the explosive growth of the finance industry, particularly since the mid-90's. (Unfortunately, the graph does not give these figures as a percent of GDP). Equally striking is the shadow banking industry's growth, overtaking conventional banking, and then its rapid decline. Its funding has fallen from $20 trillion to about $16 trillion, a 20% drop. Conventional banking, by contrast, has suffered a minimal dip, but definitely an interruption of growth. It should be noted that at their peak conventional banking had total funding of $13 trillion and shadow banking $20 trillion for a total of $33 trillion. Total GDP, by contrast, was approximately $14 trillion, or about half the total.

Around the time of the Savings and Loan Crisis, uncomfortable questions began to be raised about the health of the banking system in general. There was a rise in gallows humor on the subject. "Safe as money in the bank" suddenly didn't seem so reassuring. Cartoons showed depositors opening a bank account telling the bank, "But first I need two forms of ID and a major credit card." Such has not been the case in this financial crisis. "Safe as money in the bank" has, indeed, been a comforting mantra, compared to other things we thought were safe as money in the bank, but weren't.


Thursday, August 05, 2010

Causes of the Crisis: Shadow Banking

When Alan Greenspan kept interest rates low, housing and housing prices took off. It is, as Zandi points out, a well-established rule that asset prices and interest rates move in opposite directions because an asset that can be purchased as a lower rate of interest has a higher present value. Unfortunately, unlike earlier booms, the housing boom was not based on any technological advances in how houses were built, but only on easy credit. Mortgages, in turn, were fed into an insatiable pit informally known as the shadow banking system.

Despite its colorful name, the shadow banking system simply refers to any financial institution that is not a traditional commercial bank. Shadow banks are nothing new. In fact, as this paper points out, all banks could once be considered shadow banks, in the sense that they were loosely regulated and financially fragile. The FDIC brought commercial banks out of the shadows, insured deposits, and tightly regulated lending, but left a small "shadow" financial industry in the form of investment banks playing the stock market. Since then a wide range of non-traditional institutions have grown up, from money market mutual funds that are almost like traditional banks, to wild west hedge funds.

One of the things the shadow banking industry does (or did) is securitize. Someone, either a commercial bank or a finance company backed by an investment bank, would make a mortgage loan. Instead of keeping the loan, the originator would sell it to a "loan warehouse," either a "conduit" belonging to the bank holding company but off the balance sheet of the commercial bank, or a federally backed conduit. Once there, someone, Fannie, Freddie, or the Special Purpose Vehicle (SPV) for a bank holding company or investment bank, would pool loans and slice them into tranches (slices) for sale. Senior tranches got priority in payments, but the lowest interest rates, with risk and payment escalating up the ladder. Securitized loans were bought by banks, insurance companies, pension funds, asset managers, and hedge funds.

By this time, any kind of loan receiving regular payments could be similarly sliced and diced. Here is one thing that was never made clear to me. Different types of loans carry different interest rates. A home mortgage, for instance, is traditionally considered very safe. Defaults are (traditionally) rare because defaulting means losing one's home, and even when they do happen, the loan is backed by a stable asset, easily located, and (until recently) reliably appreciating. A car loan, by contrast, has a higher default rate and is backed by a very portable asset that is easier to hide and depreciates rapidly, so it carries a higher rate. And a credit card is an unsecured loan with no downpayment, so interest rates are very high indeed. So were these different rates of risk and return reflected in the security slices? Or were all senior tranches treated as equivalent, regardless of the loans behind them, with safer loans simply creating more senior tranches and riskier loans more high-risk tranches?

As housing prices continued to escalate and the securitization machine kept wanting more, several things happened, some in the world of housing, and some in the world of high finance. In the world of housing, the building boom started reaching its natural limit, as booms always do. This one came faster than most because it was not, after all, based on any technological advance that made housing cheaper to make, but only on easy credit. People had been buying houses at an unsustainable rate. And so, as often happens when a boom reaches its natural limits, people tried to artificially sustain it, creating a bubble. In the world of housing, that meant moving more and more into sub-prime territory. Sub-prime lending, as a share of the total, had fairly stable from 1997 to 2003. From 2004 to 2006, it took off.
Making so many sub-prime loans, notoriously, also led to the abandonment of common sense underwriting standards. Lenders all too often stopped doing assets checks or asking for proof of income. As housing prices soared higher and higher, they increasingly went out of reach to many buyers. Lenders started asking less and less for down payments, sometimes waiving them altogether. When even that failed to make a house affordable, lenders offered "teaser" rates, with artificially low interests rates for two years, to increase later. Loan originators who tried to maintain reasonable underwriting standards found themselves undercut by less scrupulous rivals and had little choice but to lower their standards as well. A similar process went on with regulation. Financial regulation was hopelessly fragmented. If any regulatory agency attempted to maintain discipline, lenders could simply choose a laxer agency. So regulators, too, saw little choice but to lower standards or lose out to rival agencies.

So how did high finance deal with all these bad loans? By disguising them through ever more complex systems of securitization. The favored form was a collateralized debt obligation or CDO. These, too, were well established forms of securitization before the 2000's, usually a custom-made pool of corporate bonds or loans that were securitized. If an asset-backed security or sub-slice of such a security started looking shaky, financial companies would pool together the shaky securities and securitize them. The product was known as an ABS CDO, short for asset-backed security collateralized debt obligation. A more accurate name might have been "chef's special" because an ABS CDO essentially took the leftovers no one wanted, threw them in the same pot, and offered up slices (tranches) of the resulting product. Toward the end, securitizers started making CDO's of CDO's, known as CDO's squared, or even CDO's cubed.

Credit ratings agencies regularly certified senior tranches of the chef's special as triple A. Credit ratings agencies had the same problems as everyone else -- if the didn't tell securitizers what they wanted to hear about their securities, securitizers would simply hire some other agency to do their ratings.

And about the time sub-prime lending the starting to take off, the SEC relaxed standards limiting how much debt the five major investment banks could run, relative to their equity.

All of which raises an obvious question. Couldn't anyone see the system was headed for trouble. The answer appears to be yes, some people did indeed see that the system was headed for trouble -- and set out to make a buck off it.


Monday, August 02, 2010

Causes of the Crisis: Too Much Money Chasing Too Few Investments

The prosperity of the Roaring Nineties had two main bases, a solid foundation of productivity growth stemming from information technology, and a more dubious foundation of easy credit and financial innovation. The good times came to an end the old-fashioned way -- with the bursting of a bubble in the stock market. Zandi describes the bursting mostly of a tech bubble while Krugman describes a broader bubble in the stock market as a whole. Both agree that something went seriously wrong with the economy once the bubble burst, but neither seriously explores what.

The decline following the bursting of the bubble was shallow, with unemployment only rising to about 6.5 percent.* But recovery was painfully slow, especially on the employment front, even though Alan Greenspan cut the interest rate the Fed charged banks all the way to 1%, well below the inflation rate of 2.5%, and the Bush Administration gave a Keynesian stimulus in the form of a huge tax cut. Greenspan feared a Japanese style lost decade and, as the recession pushed down already low inflation rates, began to worry about deflation.

So what was the problem? The 9-11 attacks certainly didn't help, but there must have been something more than that. As a mere Enlightened Layperson, I can only offer a few guesses. One is that the '90's productivity growth resulting from adopting new information technologies had reached its natural limits and was not going to keep on giving. Another (much discussed these days) is that recovery following a financial crisis always tends to be slow and painful. As in the 1920's, stock market speculators had been buying on the margins and running up a lot of debt. The 90% margins of the 1920's were no longer allowed. Margins were 50%, but borrowing even half of what you invest can lead to ruinous debt when the underlying stocks end up as so much designer toilet paper. Dot-coms and other tech companies had run up large debts of their own, which ended up as worthless after they failed. Another possible explanation is that the stock market crash had scared investors away from financing anything new and dynamic.

When the economy did move into expansion, an abundance of indicators suggested that this was a far less healthy growth than what had prevailed in the '90's, most likely because it was based on housing. So what's wrong with housing? Economists classify housing as a form of investment, but it is more accurately described as somewhere between consumption and investment. Housing is investment in the sense that it involves long term loans several times larger than the initial down payment. But it is consumption in the sense that a house is a consumer product that does nothing to increase the economy's productive capacity or productivity. Or, put differently, housing is investment in the sense of the financial resources it ties up, but not in the sense of promoting future growth. Given the choice of what to base an economic expansion on, housing would be about last on the list. But central bankers don't get to choose what part of the economy takes off, so Greenspan kept interest rates low in hopes of getting something going, and that something turned out to be housing.

Greenspan is often blamed for the current economic disaster for keeping interest rates too low for too long. Still, one would normally expect too-low interest rates to set off either inflation or an ordinary bubble of the sort we have experienced many times in the past without a comparable disaster. Something else must have gone wrong as well. So far as I can tell, two things went wrong. One was a classic case of investment-level inflation; too much money chasing too few productive investments. The other was changes in how the finance system handled all that money.

Too much money had several origins. Remember those disturbing trade deficits that plagued Mexico and Thailand before their financial crises? Remember how they were financed by foreign investment, investment reaching 6%, 7%, even 8% of GDP? Investment that outran productive outlets, set off a speculative boom, and ultimately proved unsustainable and stampeded out as fast as it stampeded in? Well, the same thing started happening to the US as well. **
Trade deficits, running at a very manageable 1.5% of GDP in 1996 and only slightly higher in 1997, began rapidly escalating afterward, peaking at 6% of GDP in 2006. These deficits appear to have begun escalating about the time of the 1997 Asian financial crisis. These may have been either trade-driven (the Asian downturn hurt ourt exports) or investment-driven (capital flight from Asia to the safety of the US). Some people believe the influx of foreign investment contributed to the stock bubble that burst in 2000. It certainly contributed to the housing bubble.

But at the same time, the housing bubble was international. Many countries besides the US experienced a similar surge in housing prices, followed by a downturn, often even worse than ours.

Several things appear to have been happening at once. For one, Alan Greenspan was not the only central banker leaving interest rates low; the tendency was world-wide. For another, international savings had become more mobile than ever before. And critically, there was what Ben Bernanke called a "global savings glut;" a "vast pool of money" that grew from from $36 trillion to $70 between 2000 and 2008. The source of this appears to be partly recycles petrodollars, but mostly an immense growth in wealth in China, which also has a tremendously high savings rate.

China's extraordinary savings rate led to the same paradox that played so large a role in causing Japan's "lost decade" -- too much savings can lead to the buildup of too much debt. This is because the influx of too much investment tends both to bid up prices of assets and to bid down returns. This leaves the finance system in general in the same place the hedge funds were in at the time of the LTCM crisis, desparately taking more and more risks with more and more borrowed money to feed the insatiable appetites of investors. Rising asset prices and falling returns also strongly encourage speculation over productive, long-term investment.

Next: Internal problems in the finance system.
*It should go without saying that if unemployment gave any sign of falling even close to 6.5% today, there would be dancing in the streets.

**Krugman discusses the dangers of such large trade deficits, but does not address what size of deficit would be safe. He does comment that Australia has been running trade deficits of 4% of GDP "for decades," all funded by foreign capital. So perhaps the danger is not in the specific size of the deficit, but in rapidly rising deficits, which we certainly had.