POSTCARDS from the Home Front.

The experience of mortgage securitization shows that the capital markets may not be as hungry for higher-risk insurance-backed securities.

The recent issue of $200 million of catastrophe bonds to protect Tokyo Disneyland  from earthquake risk raises questions about the future of the property/casualty industry. Since no insurance or reinsurance company was involved, does this transaction and other recent risk-related securitizations signal the beginning of the capital markets as the primary source of casualty insurance financing and the end of property/casualty insurance as we know it?

Many of those who believe it is see parallels between recent developments in the insurance industry and the history of another financial-services industry, mortgage banking. Today, mortgage securitization is a $3 trillion industry that dramatically changed the role of the savings and loan industry.

The impetus behind securitization in the insurance industry has been the mismatch between the potential size of single-event losses and the size of total property/casualty industry reserves--for example, the possibility of a hurricane causing losses of $100 billion to $150 billion in Miami given current industry reserves of $330 billion. The fact that the mortgage-security industry has grown so large so fast seems to offer a model of direct capital-market access that solves many of the insurance industry's current problems.

If the property/casualty industry were to go the way of the mortgage-banking industry, important structural changes would be inevitable. Securitization of mortgages changed the typical savings and loan from an integrated institution which originated, held and serviced mortgages to maturity--with all the attendant risks--to one in which a large part of the mortgage originations are immediately passed on to an investment bank or government agency, where they are securitized.

If securitization were to take hold in the insurance industry, primary insurers would, at best, become risk originators, finding risks and passing them on to investment banks, which then pass them through to the capital market. Since investment banks would then become de facto reinsurers, the survival of traditional reinsurers would be based on their ability to compete as conduits with investment banks.

All of this is predicated on the assumption that success in the mortgage-securitization arena easily translates into success in the area of insurance securitization. Although more insurance securitizations are likely to occur, the total appetite for risky insurance securities is substantially smaller than the appetite for low-risk mortgage securities, making it unlikely that capital-markets financing of risk will radically reshape the property/casualty industry.

Almost all residential-mortgage securitization has succeeded on the basis of the ingenuity of the issuers of mortgage securities in finding ways to eliminate risk. Significant demand from institutions such as banks, pension funds, and savings and loan associations has been generated by a constant re-engineering of these mortgage instruments to keep them essentially risk free. Because insurance-backed instruments by their nature cannot be risk free, no amount of ingenuity can eliminate the risk.

Insurance-backed securitization is much closer in nature to the securitization of commercial-mortgage-backed securities (CMBS), for which the risk of default is high. The risk transfers carried out through the commercial mortgage-backed securities market, however, have been very limited. Only a small percentage of all commercial mortgages have been securitized, and among those securitized it is common for the risk-free part to be transferred, while the conduit retains a significant part of the risky tranches. Based on the limited market success for these instruments, the lesson to be drawn is that the market for insurance-based instruments is likely to remain small.

Origins of Mortgage Securitization

Although a number of assets are now securitized in the United States, asset securitization began with residential mortgages.

The first securitization wave began in 1968 with the issue of Government National Mortgage Association (Ginnie Mae) pass-throughs, securities based on a pool of mortgages already guaranteed by the federal government, primarily through the Federal Housing Association. This was soon followed by the issue of pass-throughs of pools of conventional mortgages assembled by the Federal Home Loan Mortgage Corp. (Freddie Mac) and the Federal National Mortgage Association (Fannie Mae). The next wave was the issue of private label pass-throughs on pools of mortgages not qualifying for acceptance by Freddie Mac and Fannie Mae.

The market for mortgage-backed securities grew rapidly, reaching $1 trillion of outstanding debt by 1991. The key to the rapid growth of the market seems to have been the low level of risk of these pass-throughs.

Various risk-reducing devices enabled the majority of mortgage-based securities to obtain a double-A rating or better from national rating agencies. But even with default risk virtually nonexistent, these securities paid a small premium over the risk-free rate, and this combination of no risk/slightly higher return proved irresistible to institutions.

As the stable interest-rate environment of the 1960s and 1970s gave way to the highly volatile 1980s, an unexpected source of risk emerged for these instruments. Since they were backed by residential mortgages that had a prepayment option, movements in interest rates caused changes in the speed of repayment. Therefore, passthrough could not guarantee duration, and this duration risk was a problem for institutions that had to match the duration of their assets and liabilities.

To overcome this risk, a new derivative asset was developed: the collateralized mortgage obligation (CMO). This instrument split up the cash flows from the underlying mortgage pool into a number of tranches guaranteeing duration for some of the tranches at the cost of increased duration risk for others. This made these instruments palatable to institutional investors.

None of this is to suggest that there is no market for risky instruments. But purchasers of risky assets require not just a tolerance for risk, but also the time and sophistication to make sure those risky assets are being priced properly. The key to the success of mortgage securitization was the development of products that required neither time nor sophistication to price because they were designed to be, for all intents and purposes, risk free.

When it was not possible to design a risk-free instrument--as in the case of commercial mortgages--securitization was much less successful. For example, although in 1997 almost all government-guaranteed mortgages were securitized, only about 11% of commercial mortgages were. This clearly has implications for the securitization of insurance-based instruments.

The primary lesson to be drawn from an examination of mortgage securitization is a simple one, but one that runs counter to much of the current thinking: To tap into the deepest pockets of the financial markets, fixed-interest assets must be as close as possible to being risk free. Since insurance-backed assets cannot be risk free, they will remain a fringe instrument. There always will be some buyers of these securities--just as there are buyers of the unrated tranches of mortgage-backed securities--but there will never be enough buyers to alter in any significant way the traditional structure of the business of insurance and reinsurance.

This is not to say that buyers cannot be found for insurance-based assets. Already, deals of around $3 billion have been made, and more are expected ("Beyond Catastrophes," Best's Review, April 1999).

Insurance-Based Securities

The marketers of insurance-based securities will do everything they can to make them appear as low risk as possible. For example, payoffs will be tied to indexes rather than specific company losses to reduce problems of moral hazard and adverse selection. The deal for Tokyo Disneyland owner Oriental Land Inc. was written for an earthquake of a specified magnitude hitting a specified area.

Moreover, investment-grade ratings can be obtained by guaranteeing repayment of principal with a delay in the event of a loss. In this case, the rating agencies rate the securities used to underwrite the repayment of principal. If the securities are investment grade, the insurance-based asset is given an investment grade.

The market for commercial mortgage-backed securities seems to provide an accurate measure of the likely extent of insurance securitization. Of the current market's annual issuance of about $50 billion, about $5 billion consists of tranches that are unrated because of the high risk of default. Perhaps this order of magnitude--$5 billion to $10 billion a year--is a reasonable frame of reference. Obviously, if the market tops out in this range, the property/casualty insurance market faces little threat of demise at the hands of capital-market financing.

Thomas Russell is an associate professor of economics at Santa Clara University's Leavey School of Business, Santa Clara, Calif.