"When you measure what you are speaking about, and express it in numbers, you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meager and unsatisfactory kind . . . ." William Thomson, Lord Kelvin, Popular Lectures and Addresses (1891--1894).

(i) The Nature of Securitization

Most attempts to define securitization make the same mistake; they focus on the process of securitization instead of on the substance, or meaning, of securitization. Hence, the most common definition of securitization is that it consists of the pooling of assets and the issuance of securities to finance the carrying of the pooled assets. Yet, surely, this reveals no more about securitization than seeing one's image reflected in a mirror reveals about one's inner character. In Lord Kelvin's terms, it is knowledge of "a meager and unsatisfactory kind."

A better definition of securitization is that it consists of the use of superior knowledge about the expected financial behavior of particular assets, as opposed to knowledge about the expected financial behavior of the originator of the chosen assets, with the help of structure to more efficiently finance the assets. This definition is superior because it better explains the need for the most essential aspects of any securitization any where in the world under any legal system, and it better defines the place of securitization within several of the broader financial trends that have occurred at the end of our century.

The first trend has been the break down of individual, segregated and protected, capital markets into one, increasingly world-wide, capital market. The result of this trend has been a drive to find ever more efficient forms of raising capital, particularly in the form of debt financings. The more efficient forms will, by definition, in capital markets that are not segregated or protected from other competing markets, replace the less efficient forms.

Securitization, in the correct circumstances, is one of the very most efficient forms of financing. This is because of two additional trends. The first is the increasing importance of the use of information to create wealth. The second is the increasing sophistication of computers and their uses. Securitization is made possible by the combination of these two trends. Computers enable one to store and retrieve extensive data about the historical behavior of pools of assets. This historical data in turn enables one to predict, under the right circumstances, the behavior of pools of such assets subsequently originated by the applicable originator. Because our knowledge about such behavior may be so precise and reliable, when structured correctly, a securitization may entail less risk than a financing of the entity that originated the securitized assets. Again in Lord Kelvin's terms, our knowledge about the likely behavior of pools of assets is "measurable" and we "express it in numbers." It is a superior sort of knowledge from the perspective of the world of finance. Accordingly, such a securitization may be fairly labeled to be more efficient and indeed may require less over-all capital than competing forms of financing.

The preferred definition of securitization with which this essay began thus reveals why securitization often is preferable to other forms of financing. It also explains most of the structural requirements of securitization. For, to take advantage of superior information of the expected behavior of a pool of assets, the ability of the investor to rely on those assets for payment must not be materially impaired by the financial behavior of the related originator or any of its affiliates. In most legal systems, this is not practicable without the isolation of those assets legally from the financial fortunes of the originator. Isolation, in turn, is almost always accomplished by the legal transfer of the assets to another entity, often a special purpose entity ("SPE") that has no businesses other than holding, servicing, financing and liquidating the assets in order to insure that the only relevant event to the financial success of the investors' investment in the assets is the behavior of such assets. Finally, almost all of the structural complexities that securitization entails are required either to create such isolation or to deal with the indirect effects of the creation of such isolation. For example, the (i) attempt to cause such transfers to be "true sales" in order to eliminate the ability of the originator to call on such assets in its own bankruptcy, (ii) "perfection" of the purchaser's interest in the transferred assets, (iii) protections built into the form of the SPE, its administration and its capital structure all in order to render it "bankruptcy remote", and (iv) limitation on the liabilities that an SPE may otherwise incur are each attributes of the structure of a securitization designed to insure that the isolation of the transferred assets is not only theoretical but also real.

Similarly, attempts to (i) limit taxes on the income of the SPE or the movement across borders of the interest accrued by transferred receivables, (ii) comply with the various securities or investment laws that apply to the securities issued by the various SPEs in order to finance their purchases of the assets, or (iii) comply with the bank regulatory restrictions that arise in connection with such transfers, the creation of SPEs and the other various roles played by banks in connection with sponsoring such transactions each constitute a reaction to indirect problems caused by the structuring of the above described transfer and the SPE to receive the transferred assets.

(ii) Current and Future Trends

A recent trend in the United States in securitization has been attempts by regulators to "catch up" with the market. Two examples are the modifications to the risk-based capital rules of the Basle Accords proposed by certain federal bank regulators, and the proposed replacement of FAS No. 77, the accounting rules for when a transfer of financial assets removes the assets from the transferor's balance sheet, with an entirely new conceptual framework, the financial components approach. 

In the case of the risk-based capital rules, the bank regulators believed that the rules contained a logical anomaly in that a bank that credit enhanced, on behalf of an SPE, a pool of assets that the bank had not originated was required to maintain less capital against its enhancement obligation than a bank that provided the same amount of credit enhancement to such an SPE with regard to assets that the bank had originated and subsequently transferred to the SPE. The regulators believed that the risk was the same and merited the same amount of capital. The industry objected and demonstrated to the regulators that in the case of third party provided credit enhancement the bank nearly always was taking the second or third level of risk and almost never the first level of risk (which was usually born by the transferor/originator). On the other hand, in the case of the bank that transferred the assets after originating them, the bank, in its combined role of originator and credit enhancer, usually was taking the first or second level of risk. Further, an even greater logical anomaly would be created by the rules as originally proposed as, if adopted in such form, unsecured loans made by commercial banks to holding companies would require many times less capital to be held by such banks than secured contingent obligations in the form of credit enhancement given for the benefit of companies actually holding the related assets. The regulators then reformed their proposal and eventually proposed that the amount of capital required to be held for credit enhancement should be a function of a rating agency assessment of the level of the risk involved. The proposal is now under consideration by the Federal Financial Institution Examination Council ("FFIEC").

In the case of accounting rules, members of the Financial Accounting Standards Board ("FASB") had for some time believed that FAS No. 77 had been rendered obsolete by the rapidly developing complexities of the securitization market place. Hence, in 1994 the FASB proposed to adopt an entirely new conceptual framework in which the material question would no longer be whether the risks or benefits of the assets to be transferred had been retained or transferred by the transferor but instead who controlled the various benefits of the transferred assets. Those that were controlled by the transferee would be derecognized by the transferor, and those that remained with the transferor would be reflected on its balance sheet. While the new approach is a brilliant conceptual breakthrough, the FASB made the mistake of not defining "control" consistently with various accounting interpretations but instead relying on the legal concept of whether the transfer constituted a "true sale at law".

This aspect of the proposal provoked an uproar from the entire securitization community as it failed to recognize the practical structural means by which the market accomplished the isolation of the assets from the transferor's bankruptcy estate and instead relied on a legal doctrine on which there was no agreement or accepted standard in the legal community. Accordingly, the FASB is now attempting to fashion a definition of control that takes account of the practical means by which market participants in fact transfer control over the transferred assets.

What do these processes reveal? In each case, the applicable regulators believed that the markets had evolved so quickly and diversely that existing rules that had not been adopted with securitization in mind were hopelessly inadequate to deal with the intricacies of the securitization market. However, when regulators who were unfamiliar with how the market functioned in fact attempted to craft new rules, the new rules were even more impractical than the existing ones. As a result of a dialogue by regulators with market participants, the amended proposed rules become much more practical and consistent with market practice. In the U.S., because of the extensive breadth of the market, and the rapid pace of change, no regulatory body can hope to adopt practical rules without an extensive dialogue with active participants in the market place. This experience also demonstrates that it is extremely difficult to craft rules that will work in any event because of the immense and ever changing diversity of market practice.

The situation in Europe is, of course, some what different. The United Kingdom, with its tradition of common law, is the most similar to the United States. Different forms and structures may evolve freely for the most part so long as not prohibited. Structures would function more smoothly with some law and accounting changes, but it is not necessary to change the laws for securitization to thrive so long as this form of financing makes economic sense. However, countries with civil law traditions are in a different predicament. In large part, they must pass new laws in order for the securitization market to develop. This is a partial (though by no means complete) explanation for the volume of securitization in France or Spain (where such laws have been passed) compared to Italy or Germany (where less extensive laws have been passed). But because it is so difficult to build legal and accounting rules that can accommodate the tremendous creative energies of the market, markets in countries where rules must be created to permit securitization to flourish will inevitably develop more slowly than they could otherwise develop. If the market is dependent on regulators in order to thrive, the market is constantly catching up. Conversely, if the market will thrive so long as the regulators do not discourage such activity, it is more likely that the regulators will be the ones who are catching up.

Consistently with the aforesaid discussion, because of the widespread belief in the United States that securitization is socially valuable, legislation has continued to be proposed that promotes securitization. Hence, there are proposals to change the tax laws to make securitization of non-real estate assets easier and laws have recently been passed to facilitate the securitization of small business loans and leases. It is reasonable to expect that, so long as there is no highly visible disaster with regard to a prominent securitization transaction, laws of this nature will continue to be proposed and adopted in the U.S. However, there has been no explosion of transactions securitizing small business loans. This is undoubtedly because such transactions, given the mix of the high returns of such loans and the present abundance of bank capital, do not today make great economic sense. Legislation can enable transactions to occur but it will not make them occur if they do not make economic sense.

While this principle applies to Europe as well, for the reasons stated in the second preceding paragraph, the situation is more complicated. While enabling legislation usually won't change basic economics, such legislation may be necessary in order to enable basic economics to take their natural course.

A further trend is the evolution of the structure of SPEs that permit maximum flexibility for the issuer. Hence the development in the U.S. of owner trusts, master trusts and multi- seller conduits. In the case of owner or master trusts, SPEs that take these forms may issue multiple tranches of fixed or floating rate securities with maturities that vary from 30 days to 12 years or more. Such trusts may enter into private or public offerings and access the capital markets only once or repeatedly over an unlimited period of years. Similarly, conduits that issue asset-backed commercial paper or medium term notes may purchase assets of all types (including operating as well as financial assets) from multiple sellers, of investment or non-investment grade ratings, and in multiple countries. The U.S. Securities and Exchange Commission allows issuers of asset- backed or mortgage-backed securities to utilize so-called "shelfs" to access the public markets without prior review, in many circumstances, after the initial filing, allowing issuers to take advantage of favorable market conditions in a matter of days. The efficiency of execution in the asset-backed and mortgage-backed world has become, even to the most experienced, startling. Another similar development has been the creation of automobile titling trusts. To overcome the difficulties caused by state automobile titling statutes, trusts have now been developed that hold title to the automobiles and related leases and the structures of which permit securitization of their assets by the transfer to investors of beneficial interests in the trusts themselves.

This trend must certainly be reflected in time in Europe, but on the whole it will be more difficult to duplicate in countries where legislation is necessary to permit the creation and operation of the form of SPEs now being developed in the U.S. than in countries where market forces may operate naturally without such legislation.

Yet another trend in the U.S. has been the evolving notion of what constitutes a "true sale." This has been a particularly vexing problem. For legal bankruptcy purposes, when a transfer is a "true sale", the transfer is so complete that it effectively removes the assets from the bankruptcy estate of the transferor. In the U.S., courts have focused on the substance of the transaction, rather than on its form, and thus, as the risk and rewards of the transferred assets are distributed between both the transferor and transferee, for legal purposes there is often a not trivial amount of ambiguity. There is the added problem that there is no agreement in the legal community on what constitutes a "true sale" for bankruptcy purposes. The market has developed practical means of dealing with this problems with the so-called "two-tier" structure, but the recent FASB proposal described above has highlighted the problems caused by the uncertainty in the legal arena.

The FASB proposal also illustrates the difficulty of determining conceptually what a true sale should be for accounting proposes. While the legal precedents focus, in some part, on the transfer of risks, accounting standards have traditionally focused on rewards. The new financial components approach is an attempt to break out of these conceptual traps as all securitizations will divide risks and rewards among both the transferor and transferee leaving such transactions difficult to characterize with certainty by those who believe that risks or rewards should be the determinative factors. The idea behind the financial components approach is that a pool of assets and their attendant liabilities can be divided into an unending succession of conceptual risks and benefits. Rather than agonizing over how much of each must be transferred and to whom in order to find a sale (or a financing) of all of the assets and liabilities at issue, why not recognize that the transferor and transferee may each record on their own balance sheets those aspects of the assets that they control, and those liabilities for which they are liable. This will create balance sheets that are less misleading than balance sheets constructed on the basis of an all-or-nothing formulation that by definition must be at least partially misleading.

It is still too early to determine if the intellectual ferment occasioned by the FASB proposal will cause a more rational true sale standard to evolve for legal or accounting purposes in the U.S.

Again, the situation in Europe is more complex. Both accounting and legal rules vary widely form country to country. On the legal side, form seems to be much more important than in the U.S. Ironically, this means that isolation of assets, at least insofar as a true sale is required, may be easier to accomplish in many European countries than in the U.S. Often this advantage is negated, in practice, however, by the necessity of giving notice to obligors on the receivables if the investors in the securitization wish to avoid certain risks. Hence the need for legislation in countries such as France or Belgium. These is also tremendous diversity among different nations in their accounting rules. The United Kingdom ("U.K.") has gone through a very painful process in reforming its accounting rules in leading to the adoption of Financial Reporting Standard 5. By focusing on risks and rewards, the U.K. Accounting Standards Board faced the all too typical dilemma of how to satisfy market participants when the use of such standards inevitably lead either to ambiguity for many transactions or the outright elimination as sales of many market structures. The only solution is a compromise that causes some damage to the integrity of such standards as the basis for making decisions. The "linked presentation" is clearly such a compromise.

The lesson to be learned from these difficulties in defining "true sales" is that notions of what constitutes - or should constitute - a "true sale" will continue to evolve until a more satisfying intellectual framework is found that can accommodate market realties.

Finally, we may end where we began. Securitization is no more than a tool, but what a wonderful tool it is. It is ever changing, ever growing, ever striving to become even more efficient. While, of course, I cannot know, I believe that Lord Kelvin would have found securitization to be the most satisfactory of all forms of finance.