This financial alchemy is called securitization, an ungainly term for an increasingly important way that capital gets funneled around the economy. Securitization creates a financial instrument, or security, by pooling the cash flows from a number of similar assets, such as mortgages, or credit card accounts, or the future sales of David Bowie records, and putting them into a separate legal entity, often with insurance or extra collateral in case the cash flows do not materialize as expected. Creating a separate entity isolates the collateral, so that the security is not a general claim against the issuer, just against those assets. Pooling diversifies risk and reduces the need to monitor each underlying payment stream, making it possible to create liquid instruments out of assets that would be too cumbersome or expensive to sell off individually.

Thus, securitization gives large investors, such as pension funds and insurance companies (among the prime purchasers), access to borrowers and others with payment streams to sell across the broad swath of the economy -- and vice versa. It fosters the flow of funds across time and space, connecting spare capital with those who want to invest in productive assets or simply increase their consumption. It links investors to firms via capital markets, and thus provides an increasingly viable alternative to traditional channels of finance, especially lending by local banks. And for companies with balance-sheet assets beyond cash, securitization can transform these assets into financing.

The U.S. economy has seen enormous expansion in asset-backed securities since the first mortgage pass-throughs in the 1970s. Today, the market for securitized mortgages has grown larger than the entire market for corporate bonds. And the more exotic deals, though far smaller in total dollars, continue to expand in importance and in the range of assets that can be transformed.

Securitization is part of a wave of financial innovation that has lowered the cost of moving funds. It has created new securities that are tailored to the needs of investors and issuers, elaborate and mutable financial instruments in which the underlying assets are rendered unrecognizable. It has linked borrowers and investors together, and created thicker and more efficient national capital markets. And financial institutions are being reconfigured as many traditional distinctions among them dissolve.

Liquified Mortgages

Securitization had its roots in the unglamorous mortgage market of the 1970s. At that time, banks, especially thrifts, and other financial institutions tended to operate regionally -- taking in deposits from local residents and making loans to local homeowners and businesses. This meant that in some parts of the country, demand for mortgages and other funding might exceed the supply of deposits, leaving banks without the resources to make otherwise attractive loans. In other areas, the supply of funds might exceed demand. Thus mortgage rates tended to reflect local market conditions and often varied from region to region, with the difference between rates in the highest and lowest regions as much as 1 percent or more. Financial institutions were susceptible to regional downturns because they were unable to hold a geographically diversified loan portfolio that would allow them to offset local losses against profits from regions that were thriving. Savings and loans -- legally required to keep a certain percentage of their holdings in residential mortgages -- were especially vulnerable. Rising interest rates also could be treacherous, since banks, constrained by a legal ceiling on the interest they could pay depositors, were subject to an outflow of funds. Later on, removal of the ceiling only increased the pressure. Although banks could pay high rates, they were receiving payments on mortgages and other loans originated when rates were low.

Securitization helped to change all that. In 1970, the federal agency, Government National Mortgage Association (GNMA), bundled together a pool of single-family mortgages as collateral and sold securities, called "Ginnie Mae" pass-throughs. Principal and interest on these mortgages would be collected by the bank or entity that originated the loan, and then "passed through" to investors, minus a servicing fee. GNMA guaranteed the timely payment of both principal and interest, eliminating risk of default. Over the decade, other government agencies and private institutions, such as Bank of America, developed similar pass-throughs and national standards emerged for qualifying loans for these programs. By 1980, the total amount of mortgage pass-throughs outstanding had climbed to $100 billion. By 1990, it had reached $1 trillion.

Securitization allowed investors in pass-throughs to make use of the diversification inherent in a large number of pooled loans and thus reduced their need to gather information and monitor the payment history of the underlying cash flows. Loans that would have been too costly or cumbersome to sell separately could now be bundled into one large security and more cheaply and easily sold. And, by creating a tradable instrument out of illiquid assets, a lively secondary market commenced. During the 1980s, trading in mortgage bonds went from a backwater job on Wall Street to a very profitable vocation.

As the secondary mortgage market grew, competition increased and traditional lending functions were unbundled. Some financial institutions began to specialize in originating loans, others in service and monitoring, and still others in providing the funds, as each assumed tasks based on relative efficiency. Thus, New York Fed economists Paul Bennett, Richard Peach, and Stavros Peristiani find that the secondary market has lowered origination fees and other costs of obtaining a mortgage. Robert Cotterman and James Pearce, at Unicon Research, estimate that conventional-sized mortgage loans, most of which are securitized, have interest rates from one-quarter to one-half of a percentage point lower than "jumbo" loans that exceed government standards in size, and are mostly not securitized. And this may understate the secondary market's impact, since researchers believe that lowered rates on conforming loans have decreased the rates for jumbos also. Thus, the development of the mortgage pass-through and the secondary market are widely credited with making homeownership more affordable for many Americans.

Today, about one-half of all outstanding home mortgage loans are securitized. In this way, banks can move assets off their balance sheets and reduce regulatory capital. But this is not the most important motivation. Securitization makes it possible for banks to adjust their portfolios so the maturity of their assets more closely matches that of deposits. This has greatly diminished bank vulnerability to a rise in interest rates, especially important after 1970 when interest rates became more volatile.

Banks also use securitization to reduce their sensitivity to local economic shocks. Although deregulation has reduced the legal reasons for geographic segmentation, many banks continue to concentrate their loan originations in particular regions or industries where they have superior knowledge of market conditions, property values, and creditworthiness of borrowers. Now those loans can be bundled with others, and bought and sold anywhere in the country. Money can flow from regions with idle deposits to those with excess loan demand, and thus mortgage rates are more similar across the country. And banks can hold more diversified portfolios and are less likely to be upended by a slumping local economy.

Adventures in Financial Engineering

With the success of the mortgage pass-through, members of the financial community grew more confident of their ability to assess and engineer securitized instruments. They also could see that a pass-through had certain features that limited its attractiveness to buyers Its cash flow is unpredictable, and it tends to have a relatively long maturity -- a pass-through is not retired until the last homeowner pays off the last mortgage. Thus there were profit opportunities and substantial incentives to devise new securitization designs with greater predictability, varied maturities, or other innovative wrinkles important to investors.

The main headache in mortgage pass-throughs is the possibility of prepayment. Most mortgages allow homeowners to pay off any part of their principal ahead of schedule without penalty. Pass-throughs deliver all payments to investors, including any early payments of principal. Divorce, death, and sale or destruction of the property are all reasons why a loan might be repaid early. But homeowners are especially prone to prepay when interest rates drop -- the worst time for investors since they are forced to reinvest at lower rates. Although prepayment can be modeled fairly well, exact rates cannot be predicted and pass-through maturities remain indeterminant. This prepayment risk is borne equally by all investors, since each receives a share of all payments.

But many potential buyers of securitized instruments, such as pension funds, insurance companies, and other large institutional investors, would prefer not to assume this risk. They want safe, investment-grade assets with relatively certain maturities and yields. So in 1983, the Federal Home Loan Mortgage Corporation issued the first collateralized mortgage obligation (CMO). Like pass-throughs, CMOs are backed by a pool of guaranteed mortgages, but their payment rules, designed to provide investors with additional choices, are more complicated.

In its simplest incarnation, a CMO has several tiers, known as "tranches." Bonds in each tier receive interest payments, but all principal payments go to bonds in the top tier until they are entirely repaid. Then, the principal payments flow to the next tier until that tier is repaid, and so on, until all tiers are retired sequentially. Thus, a CMO is, in essence, several bonds with some shorter, some intermediate, and some longer maturities. Investors still face prepayment risk, but this risk is divided up differently among the tiers. The upper tiers have shorter and more certain maturities, reducing reinvestment risk; the bottom tiers have longer maturities and assume more concentrated risk in exchange for a higher return. Some CMOs have an extra risky "Z" tranche at the very bottom that pays no principal or interest until all previous tiers are retired. Thus, investors can choose which tier to buy, depending on their taste for risk, and the maturity and yield that they desire.

"Once market participants got the idea of splitting up the cash flows from a pool of mortgages, there was no stopping them," writes economist Joseph Haubrich of the Federal Reserve Bank of Cleveland. They were able to tweak each feature to create a seemingly endless number of variations. For example, investors who want even more certainty can now buy a PAC, an elaborate tranche that renders payments according to a prespecified schedule. Those who want greater return (and can accept more risk) might buy PAC's companion tranche, which is created from payments left after the PAC schedule is met. There are now bonds that pay interest only (IOs) so long as the underlying tranche gets principal payments and others that pay principal only (POs). There are floaters, which are linked to a particular interest-rate index, and superfloaters (which multiply the effect of a change in the index), and jump Zs (a Z tranche that, under certain circumstances, jumps to the head of the tranche line), and more. Similarly elaborate structures have also been created for assets other than mortgages, called collateralized bond obligations.

The impact of CMOs, especially the exotic varieties, is far more controversial than simple pass-throughs. After their introduction, the market grew steadily, from $20 billion issued in 1985 to $320 billion in 1993, attesting to their appeal. They offer investors an expansive range of choices as to risk, maturity, and yield. Thus, the value of a CMO exceeds that of its underlying pool, say Julie Fernald, Frank Keane, and Martin Mair in a study for the Federal Reserve Bank of New York.

But CMOs cannot make prepayment risk disappear; the best they do is shift the risk from one tranche to another. The more exotic configurations sometimes behave in quirky and complicated ways, with amplified effects, which makes them difficult to understand and predict. They are not widely traded in secondary markets, so investors must rely heavily on mathematical models to determine how to price the risk. And lower tranches, sometimes dubbed "toxic waste" or "kitchen sink bonds," can be especially volatile and hard to evaluate. While intended for sophisticated buyers who can judge the relationship of risk to return -- mutual funds, hedge funds, and equity investors -- these exotic CMOs have sometimes been marketed to those far less experienced. In the chase for higher yield, even the sophisticated have been burned. During the early 1990s, CMOs were implicated in losses at Kidder Peabody, Goldman Sachs, Bear Stearns, and others. The market contracted, as outstanding issues fell by $200 billion between 1993 and 1994. And according to economists Fernald, Keane, and Mair, exactly how much homeowners have gained through lower mortgage rates remains an open question.

"A Bond is Just a Promise"

The second innovation frontier was the securitization of new asset classes. As the mortgage market grew, the financial community recognized similarities between mortgages and other assets. Take car loans, for example. The collateral might be different Houses tend to stay put, don't rust, and maintain their value better. But both generate payment streams. Each stream has a history that can be analyzed, for rates of default, early and late payment, and response to other economic variables. Thus, a cash stream that appeared reasonably predictable was a candidate to be securitized. In 1985, Salomon Brothers developed the first bonds backed by auto loans and, within a year, had also launched credit card securitizations. By 1996, firms were issuing $50 billion in credit card securitizations, $40 billion in bonds backed by home-equity loans, and $33 billion backed by auto loans and leases, according to the newsletter, Asset-Backed Alert.

After 1986, the pace of innovation quickened, with the creation of bonds backed by an expanding range of assets, from computer leases to municipal property tax liens to health club fees. In many instances, the investment banks that helped package and market the deals simply identified the logical extensions of previous securitizations, then approached and courted owners of assets that were attractive candidates. And for many of these owners, securitization meant access to capital at rates that would not have been possible with traditional financial relationships.

Thus, it is unlikely that David Bowie and his business manager, Bill Zysblat, would have been able to approach a bank and walk out with $55 million against the royalties from such albums as "The Rise and Fall of Ziggy Stardust and the Spiders from Mars," and other early efforts. Bowie, who reportedly still sells more than two million albums a year, is somewhat unusual in that he owns the copyrights and master tapes to his early recordings; generally the record company owns them. His two existing ten-year distribution deals were about to end. Negotiations for a new deal were under way when David Pullman, of Fahnestock and Company, a small New York City investment firm, proposed bonds as a way to get money up front. While Pullman tested the feasibility of a bond sale, Zysblat looked to see how big an advance record companies would offer. Said Zysblat to the press, "His numbers were bigger than my numbers."

Meanwhile, Pullman approached Prudential, the nation's largest insurance carrier, because it was known as one of the few companies willing and able to take on such an unusual deal, notes Prudential Vice President Andrea Kutscher. Bowie is not exactly a perfectly diversified product, but he did have a number of different assets, including royalties from 25 albums and 250 songs, publishing rights, sheet music, and movie rights, all across the globe. He and his record company could also provide up to twenty years of sales data, which Prudential was able to slice and dice, looking at first-year sales, sales thereafter, and other trends, year by year. Kutscher also conferred with several record companies for additional perspective. As a result of its analysis, Prudential became convinced that there was a predictable stream of income there. "A bond is just a promise, anyway," one observer commented.

The rating agency also had to be reassured. Moody's reviewed the transaction for asset quality; attorneys opined on whether the legal structure was truly independent. In this instance, the deal had additional insurance. EMI Records, the company with whom Bowie also signed a deal for his future output, contributed "substantial support" through a combination of "guaranteed payments and other structural enhancements," as Kutscher delicately put it. Thus, EMI probably absorbed a substantial amount of the risk in this deal. When the bonds are retired, the underlying assets return to Bowie's estate. The bonds received an A3, below AAA, but a solid rating, nonetheless.

With the rating in place, Prudential bought the entire issue in a private placement for its general account -- where it invests the money of life insurance policyholders. The bonds have an average life of about 9 years and were priced to pay 7.9 percent interest, which compared favorably to the 10-year U.S. treasury rate of 6.4 percent and was 50 to 100 basis points above similarly rated corporate bonds sold at the time. Prudential considers this premium a return to its time and expertise in putting the deal together. And what did Bowie, reportedly worth $100 million, plan to do with the cash? The singer has been mum on the subject.

In other instances, securitization of new asset classes has become a way to obtain financing where it would otherwise be next to impossible. Troubled companies have been able to obtain capital at favorable terms, by isolating and securitizing high-quality assets and using the proceeds to restructure their balance sheets, including paying down high-cost bank debt. During the peso crisis, José Cuervos and other Mexican companies were able to raise capital by securitizing their earnings from U.S. exports. State lottery winners can get their money up front. Even people with AIDS and advanced HIV patients have been able to get cash up front via securitizations backed by their life insurance policies. Some Wall Street analysts have suggested, half seriously, that the next twist is for Shaq to raise $10 to $20 million by issuing bonds backed by his expected future endorsement revenues.

All That's Solid. . .

Securitization can be a powerful tool for individual firms and a boon to economic growth. Some on Wall Street are now looking to create bonds backed by high-tech royalties so that startups can get money up front for research and development. And, in a recent review of the research evidence, University of Virginia Professor Ross Levine suggests that developed and functioning financial systems are "vitally linked to economic growth."

Still, there may be limits to how far securitization can go. These are expensive procedures that require minimum issue size and, often, the prospect of a repeat deal to make them profitable. And, David Bowie aside, securitization is not useful for pooling and diversifying our biggest individual economic risks, such as a dive in future income from an unexpected depreciation in our human capital or some large macroeconomic shock.

Nonetheless, securitization represents a substantial change from traditional financial relationships, where a local bank makes loans to its neighbors based on community ties and knowledge of the individual business. Some observers, such as Harvard Business School Professor William Poorvu, while aware of securitization's benefits, are concerned that something will be lost as commercial financing moves from a business based on relationships and individual judgments to one built on fees from servicing loans that are standardized, pooled, and sold. He cautions that putting commercial projects through a cookie-cutter loan application process may result in worthwhile projects going unfunded, especially in urban areas, even as we get a surfeit of suburban projects that fit the model.

And while securitization may work well enough during a prosperous economy, will these new arrangements be dependable in the next downturn? According to Poorvu, "The jury is still out." Models and pricing procedures for the more exotic securities have yet to prove they can withstand the test of a severe recession. And even for the less exotic versions, when cash flows slow and borrowers look for relief, Poorvu fears that banks will be hemmed in by the dictates of those who bought the loan. Bankers may be forced to "consult the manual," rather than exercise their best judgment in each case.

But the exercise of judgment takes valuable resources, and also carries risk. And saving the economy from the cost and risk of making these individual judgments is precisely what securitization is supposed to do.