The Creation of Value

My charge in speaking to you today is to address the notion of conducting research in risk management and insurance. I suspect the charge may also be to provide some insight into successfully publishing research in leading journals like the Journal of Risk and Insurance and the Geneva Papers. I’ll try to cover both.

In much of our research we’re confronted with problems that require some dynamic programming. That typically means solving the problem backwards or answering the last question first. It makes some sense here as well. Saying something about what is required to publish any piece of academic research is useful not only in its own right but also in the planning and preparation for the conduct of the research. It may help identify some of the issues that must generally be addressed before beginning a research project. It may also help in framing the right set of questions.

Here is a short list of items to keep in mind when beginning your research. It is not exhaustive but it may help you view your ideas from an editor’s perspective. It may also help you write for your audience.

Publishing guidelines

Do

1. Select a topic that really interests you. You may be working on it longer than you think!

2. Select a topic that is important, e.g., does it advance our understanding or extend the boundaries of the discipline.

3. Do a complete literature search using sources such as JSTOR, the Journal of Economic Literature’s EconLit database, and the Web of Science.

4. Select a topic that is appropriate for the journal. Look at what has been published there but don’t use that as your only criteria for submission.

5. Clearly state your contribution to the literature.

6. Review the literature briefly pointing out the most important contributions of others and where your contribution engages with the literature.

7. Make the tone of your article positive.

8. Select the simplest model possible to make your point.

9. Make your technical work as accessible as possible to the audience.

10. For empirical papers, ensure that the hypotheses are based on received theory and cite it.

11. Do cite all relevant literature.

Do not

1. Do not over emphasize your contribution by either minimizing that of others or making unsubstantiated claims about your own.

2. Do not submit a paper with a negative tone. You cannot make a contribution to the literature by doing nothing more than attacking the work of others.

3. Do not submit a paper to multiple journals.

Of all the recommendations, none is more important than making your contribution to the literature clear and concise.

Research topics

The mergers and acquisitions, consolidation and convergence that are occurring in financial markets make this an exciting time to be studying risk management and insurance. Some of the important topics are changing just like the markets. Other topics have a long history and remain unsolved. The first topic that I note below is convergence and it is represented in the markets by a blurring of the lines between banking, finance, and insurance. The same comment needs to be made about boundaries of disciplines like insurance, actuarial science, economics and finance; the boundaries between these disciplines are also blurring. This makes my task as an editor more challenging because I am trying to identify the new boundaries.

Some of the more important topics include market convergence, catastrophes, terrorism, mortality, genetics and cycles.

Convergence

One conception of convergence is that the lines between banking, finance, and insurance are increasingly blurred. For example, what was once the domain of reinsurance is now that of investment banks or reinsurers acting in capital markets. The convergence is occurring as a result of the innovation, added choices, and improved efficiency.

It is possible to claim, as I have, that risks are commodities that may be exchanged and that the corporation, long viewed as a nexus of contracts, may also be viewed as a nexus of risks. The corporation may be described as a composite commodity or bundle of risks that may be separated or re-bundled to suit the objectives of the corporation. “Indeed, the history of the development of risk instruments is a tale of the progressive separation of risks, enabling each to be borne in the least expensive way,” (Kohn 1999). An economy may achieve an efficient allocation of risks as well as resources through separation and trading, i.e., see (Arrow 1963) or (Debreu 1959). The convergence that we see in financial markets may be viewed a reflection of the historical imperative or equivalently as movement toward a more complete market.

While the historical imperative of separation and cost minimization continues to hold, the pace of that separation of risks seems more rapid now or maybe I’m just older now. The first option pricing models, e.g., (Black and Scholes 1973), could rely on no arbitrage conditions and a security that was already valued in the financial market. Some of the new derivative instruments are being written on indices for which there are no underlying assets that are traded; the weather derivatives are but one class of examples. One of the many challenges for a new risk management theory is the development of a model sufficiently robust that it allows the risks to be unbundled and valued so that the cost minimization can be refined. That model will almost surely be a synthesis of insurance and finance. There is no theory that currently exists that provides the means to price some of the new derivative instruments. The convergence of the markets will depend on the development of such a theory.

There are other risks involved in the convergence phenomenon that firms may choose to bear, transfer, or hedge. The risks include property risk, casualty risk, interest rate risk, foreign exchange risk, commodity risk, credit risk, and weather risk. Some risk management tools, e.g., insurance and leverage, can have either a direct or indirect impact on the consequences of several of these risks and hence may not be the lowest cost method for managing the risks.

The weather derivatives provide one recent example of an instrument designed to manage a risk that does not have an immediate and direct impact on other risks. These derivatives are written on an index of the weather conditions such as cooling degree-days or heating degree-days and can take forms such as that of calls, puts, swaps, caps, collars, or floors. The development of these instruments provides an interesting example of the convergence in the insurance and finance markets. It is also an example of the historical trend of the development of instruments that separate risk that will lower the total risk bearing cost.

Finally, there is an instrument that plays a central role in much of the new development of risk management. Most of the instruments noted here can only succeed if the contracts are honoured. Hence, credit derivatives must play a central role in the new risk management. The notional volumes have grown from $5-$10 billion in 1995 to between $50 and $100 billion in 1996 to over $250 billion in 1997. Like other derivatives, the credit derivative can take a variety of forms. The simplest is the credit swap. Suppose firms C and J enter a CAT swap and, as above, suppose that L_c and L_jare the random net losses on the books of business for C and J, respectively. Each firm incurs some credit risk. Firm C may default when a California quake makes the net loss positive. To protect itself against this credit exposure, J enters a deal with firm A in which J pays A a fixed amount and in return A agrees to assume C’s obligations to J in the event that C defaults. This reduces J’s credit exposure and J only suffers a loss on the California quake exposure if both A and C default. Such a credit derivative can be viewed as an assurance of the separation of risks in the CAT swap.

Catastrophes

Natural catastrophes include storms, floods, earthquakes, etc. 2001 was an average year prior to September 11th. The catastrophe (CAT) risks can be managed with a variety of instruments like insurance, reinsurance, CAT bond and options

The Law of Large Numbers implies that insuring risks is possible by constructing a large pool or portfolio of independent risks. This, however, is not always possible. Some risks have large positive correlation coefficients. This makes insuring difficult but not impossible.

In recent years, the magnitude of catastrophic property-casualty disaster risks has become a major topic of debate. The insurance industry now regularly discusses potential U.S. earthquake or hurricane losses of $50-$100 billion, a magnitude of loss that was unthinkable ten years ago. The disasters of Hurricane Andrew and the Northridge Earthquake alone totaled over $45 billion in 1997 dollars, with the insured component running to almost $30 billion. This compares with cumulative insured losses from natural catastrophes in the decade prior to those events (roughly 1980-92) of only about $25 billion (according to data from Property Claims Services).^{^[1]}

Catastrophic risks will continue to grow since, for example, the population of hazard prone states like California and Florida have grown at rates two or three times that of the national average. An event loss will exceed $50 billion at some point. Estimates from A. M. Best place the capital and surplus of U. S. insurers at about $239 billion but this capital and surplus is for all risks, not just catastrophes. Reinsurance provides little more help since its capital and surplus is also small relative to the potential losses, i.e., $26.7 billion for U.S. reinsurers, $6.5 billion for Bermudan reinsurers, $7 billion for German reinsurers and $16.8 billion for others.^{^[2]} The estimates are all in 1997 dollars.

In 1992 the United Services Automobile Association (USAA) experienced a $600 million loss when Hurricane Andrew hit south Florida; the industry loss was $16.5 billion but a storm the size of Andrew about forty miles north would have resulted in insured losses of over $50 billion. A variety of options are available to handle the catastrophe (CAT) losses. The insurers can expand reinsurance, issue catastrophe bonds, purchase catastrophe options, engage in catastrophe or basis swaps, etc.. Reinsurance has some advantages because of the relationship that exists between insurer and reinsurer; this type of contracting can reduce the moral hazard and adverse selection problems but the capacity of the reinsurers raises questions about credit risk.

The CAT bond represents a securitization of the risk. Securitization is the process of aggregating similar instruments into a negotiable security; this process is well known in real estate where the instruments are loans. In 1998 USAA issued the first catastrophe (CAT) bond. This type of bond is designed to provide a return to the issuer in the event of catastrophic losses due, for example, to a hurricane or an earthquake. The money used to purchase catastrophe bonds is invested in securities and the bondholders receive a return from the portfolio of securities in the event that a catastrophe does not occur. Some CAT bonds provide principal protection and others, with higher interest rates to compensate for the risk, do not. As in the USAA case the instrument may be designed to cover a particular layer of CAT losses in its book of business and that will affect the probability that loss event coverage is triggered. The CAT bonds provide a good instrument for diversification because the losses they cover are not highly correlated with other financial instruments. One of the biggest advantages, however, is the securitization. The catastrophic losses may represent no more than a normal day’s fluctuation in U.S. equity values and the CAT bonds tap the much greater capacity of the capital markets. Since the U. S. financial markets represent about $12 trillion, a $50-$100 billion event would only represent about 40-80 basis points of wealth.

A CAT swap represents a diversification of the risk. The diversification benefits are easy to see. Suppose L_c and L_j represent the random net losses for a book of business on California quakes and Japan quakes, respectively. Suppose a fraction q of the California book is swapped for that proportion of the Japan book. Then the new book is the portfolio (1 - q) L_c + q L_j. If the net random losses are independent then the diversification advantages are clear.^{^[3]} The swap represents a form of geographic diversification but it does introduce some credit risk that will have to be borne or transferred.

A CAT option is a derivative. An industry index of losses I is constructed and the options are written on that index. A call option would have the form max{0, I – i} and so would be in the money for all indexed losses exceeding the strike price, equivalently, the loss i. This instrument taps the capital markets for coverage and avoids some of the credit risk problems of the CAT swap. This type of instrument, however, introduces the potential for basis risk because the loss experience of the firm may be different from that of the index. If L represents the random loss of the firm then the basis risk would be the difference I – L. One of the advantages of the CAT option is that the use of an index eliminates the moral hazard and adverse selection problems that hamper the risk markets.

It is possible to view the decision for or against the CAT option as a trade-off between the agency costs and the basis risk costs. It is also possible for the firm to select a CAT option and then eliminate the basis risk with a basis swap. With a basis swap the firm trades the call option for loss coverage at a price. Such a swap would reintroduce the agency costs that the CAT option eliminate but those costs would be reduced because they would be on the value of the basis risk rather than the risk on the whole book of business, i.e., any deterioration in underwriting standards due to the moral hazard problem would have to be reflected in the swap price but to a lesser extent.

Terrorism

Is terrorism insurable? Is private insurance feasible? Is private or public insurance cover the best response? Pool Re provides terrorism cover in the UK CCS (Consorcio de Compensación de Seguros) in Spain; SASRIA (South African Special Risks Insurance Association) in South Africa; PTCF (Property Tax and Compensation Fund) in Israel.

Once a national concern, terrorism has become an international one of unprecedented proportions. The September 11^th losses unlike natural catastrophe losses were entirely unexpected and, more importantly, unfunded. The total insured losses are estimated at 30-58 billion USD and the reinsurance and some insurance markets have hardened; the markets had hardened prior to September 11^th. What is more, insuring some property became quite a problem after September 11^th. There has been an increased demand for terrorism and other covers but with few exceptions^{^[4]} terrorism insurance was not readily available.

Mortality (Longevity)

The age distribution has continued to change as the baby booms ages in many countries. The changes in the age distribution in conjunction with the problems financing social security and similar state pension plans have increased the demand for private pension plans, e.g., annuities, around the world. One of the key problems facing annuity providers is mortality risk. Like individuals, annuity providers face the prospect of outliving their resources and so not being able to honor their annuity contracts if the mortality prediction is wrong. Hence, mortality risk is a problem and it can and probably has inhibited the growth of annuity markets.

Mortality risk can be hedged in a variety of different ways, including the recently proposed survivor bonds, i.e., see (Blake and Burrows 2001). The survivor bond provides a coupon payoff commensurate with the proportion of the population that survives to a particular age and so the provider can use this bond to hedge the mortality risk. Given such an instrument, it would also be possible to construct mortality options and, given the existence of the bond, the options could be priced based on the bonds. A natural hedge may exist if the insurer balances the life book of business with the annuity book of business but the age distribution may make it difficult to maintain such a hedge while survivor bonds and options would allow much greater flexibility.

Genetic Testing

Should insurers have access to genetic test results? If not then how costly will the adverse selection problem be? If so then will it create groups that are uninsurable? Should the state be an insurer of last resort?

Currently there is a five-year moratorium on insurers’ use of genetic testing information in the United Kingdom. Other countries in Europe ban its use with either a moratorium or legislation. The story is much the same in the United States where some states have gone as far as banning the use of medical histories. It is still too early to measure^{^[5]} the extent of the adverse selection problem that will occur if the bans become permanent. At its worst, the adverse selection problem can cause market failure and so more work needs to be done to assess the problem.

Market Cycles

What is a market cycle? Why do cycles exist? What causes them? Can risk management dampen or eliminate them? Periods of falling insurance premiums (soft markets) followed by rising insurance premiums (hard markets) still characterize insurance markets. While there are a number of theories there is no lasting consensus. Of course, when the industry experiences a large loss it is difficult for publicly held and traded corporations to issue debt or equity because those issues reduce current shareholder value and the corporate management has a responsibility to act in the interests of its current shareholders; if management is paid in stock or stock options then it is also in their interests on personal account not to issue new debt or equity to replenish the capital reserves. Hence, in order to maintain sufficient capital for its existing book of business and any new policies the policy prices, i.e., premiums must increase to replenish and augment the capital. Of course, as the prices increase it becomes more profitable for new firms to write business and so may prompt entry to the industry. The entry may become a problem when competition for existing business necessitates price reductions and despite the moral hazard problem, the limited liability of publicly held insurers may motivate them to increase the corporate risk by reducing prices and so maintain or increase stock value. This does not, however, explain the existence of market cycles or why risk management has not eliminated them.

Concluding Remarks

The research suggestions and topics just scratch the surface of what can and ought to be done to extend our knowledge of risk management and insurance. We are all in the business of value creation; our value comes with the creation of knowledge.

References:

Arrow, K. J. (1963). "The Role of Securities in the Optimal Allocation of Risk Bearing." Review of Economic Studies.

Black, F. and M. Scholes (1973). "The Pricing of Options and Corporate Liabilities." Journal of Political Economy 81: 637-59.

Blake, D. and W. Burrows (2001). "Survivor Bonds: Helping to Hedge Mortality Risk." Journal of Risk and Insurance 68(2): 339-48.

Debreu, G. (1959). Theory of Value: An Axiomatic Analysis of Economic Equilibrium, John Wiley & Sons.

Froot, K. A., Ed. (1999). The Financing of Catastrophe Risk. Chicago, The University of Chicago Press.

Kohn, M. (1999). "Risk Instruments in the Medieval and Early Modern Economy." Dartmouth College Working Paper 99-07.

MacMinn, R. D. (1984). "A General Diversification Theorem: A Note." Journal of Finance 39(2): 541-50.

Samuelson, P. A. (1967). "General Proof That Diversification Pays." Journal of Financial and Quantitative Analysis 2(2): 1-13.

Subramanian, K., J. Lemaire, et al. (1999). "Estimating adverse selection costs from genetic testing for breast and ovarian cancer: The case of life insurance." Journal of Risk and Insurance 66(4): 531-550.

^{^[1]} Froot, K. A., Ed. (1999). The Financing of Catastrophe Risk. Chicago, The University of Chicago Press., p. 1.

^{^[3]} See Samuelson, P. A. (1967). "General Proof That Diversification Pays." Journal of Financial and Quantitative Analysis 2(2): 1-13. and MacMinn, R. D. (1984). "A General Diversification Theorem: A Note." Journal of Finance 39(2): 541-50..

^{^[4]} AIG did start providing terrorism insurance shortly after September 11^th. Other companies such as Swiss Re, Munich Re, etc., have also since joined in offering at least limited coverage.

^{^[5]} Some work has been done on the problem, e.g., see Subramanian, K., J. Lemaire, et al. (1999). "Estimating adverse selection costs from genetic testing for breast and ovarian cancer: The case of life insurance." Journal of Risk and Insurance 66(4): 531-550.