Risk Transfer: Why Insurance Is Not an Option
By Cynthia Edwalds
Expanding Horizons, April 2021
Enterprise risk management (ERM) has emerged and expanded rapidly over the past couple of decades. It provides a unifying framework for considering all of the risks to which a firm is exposed: business risk, operational risk, financial risk, strategic risk, reputation risk and cyber risk. ERM also brings together professionals with expertise in various aspects of risk under one umbrella. One of the challenges of this unification has been the development of a common language about risk. For example, in the textbook Fundamentals of Risk and Insurance, Emmett and Therese Vaughan define risk as:
[A] condition in which there is a possibility of an adverse deviation from a desired outcome that is expected or hoped for.
On the other hand, in the textbook The Essentials of Risk Management, Crouhy, Galai and Mark define risk as:
[V]ariability that can be quantified in terms of probabilities.
The difference in these definitions reflects the perspectives of insurance professionals, who are accustomed to thinking about the unlikely possibility of a severe blow to the financial condition of the entity, versus those of investment professionals, who are accustomed to seeing portfolio values fluctuate continuously and accept some range of adverse fluctuation as inevitable.
This difference in perspective is also evident in the language surrounding risk transfer. In investment parlance, the risk of excessive adverse fluctuation is transferred out of the portfolio by a hedge, which constrains the range of adverse fluctuation of portfolio values. In insurance parlance, the risk of loss is transferred to the insurer by an insurance policy, a contract that specifies how a payment to the insured by the insurer will be determined if and when a specified loss occurs.
Is this difference merely a matter of semantics? Is insurance just a special type of hedge? Or is a hedge an alternative form of insurance? Or are hedges and insurance actually different?
I contend that these are fundamentally different forms of risk transfer. This article describes the similarities and differences between these two forms of risk transfer, then concludes with a discussion of the implications of this difference for enterprise risk management.
Insurance vs. Hedging
From the perspective of an individual firm or person, these two forms of risk transfer seem to be very similar. In each type, the decision maker must identify the relevant risks then decide how much of each risk to retain and how much to transfer, which will be influenced by the price to be paid for transferring the risk. In both cases, the price of transferring the risk decreases as the amount of risk retained increases.
From the perspective of society, both forms of risk transfer result in a net decrease in total risk, where total risk is measured by the average coefficient of variation, even though the total expected amount of potential loss remains the same. However, the processes by which risk is reduced are fundamentally different between these two forms of risk transfer, and this leads to other differences in their mechanics.
Process of Risk Transfer and Reduction
The fundamental process of risk reduction through insurance is risk aggregation, in which risks that are similar but approximately uncorrelated are combined. The resulting block has an expected value of loss equal to the sum of the expected values of loss for each risk included in the block, but the coefficient of variation is considerably less than the coefficient of variation of any of the individual risks in the block before the risk is transferred. Both the insured and the insurer have a smaller coefficient of variation when a new insured is added to the block, as long as the expected value of loss for the new insured is not large relative to the total expected value of loss for the rest of the block.
The fundamental process of risk reduction through hedging is risk dispersion and diversification, in which a relatively large risk to which multiple entities are inherently exposed with high correlation is subdivided into smaller parcels that are perfectly correlated. Some of the smaller risk parcels are transferred to entities that are either already exposed to the same risk in the opposite direction or exposed to uncorrelated or less than perfectly correlated risks. These entities will reduce the total risk in their portfolios by adding a parcel of the dispersed risk. In some cases, accepting the dispersed risk is their own hedge for their opposite exposure to the same risk.
Counterparty Acceptance of Risk
One of the key differences between insurance and hedging as risk transfer mechanisms arises from the difference in their fundamental processes: the counterparty accepting the risk being transferred. In an insurance transaction, the risk is accepted by the entity aggregating the risks into a block, which is the insurer. Although some authors define insurance as a transfer of risk from an individual to a pool in which all participants share equitably in the pooled risk, this is not completely accurate. The insurer accepts the risk in exchange for the premium paid. The insured is no longer exposed to the risk transferred. If the total premiums collected by the insurer are insufficient to cover all the losses of the block, plus the expenses incurred by the insurer in administering the block, the insurer must absorb the excess loss.
In a hedging transaction, the risk is initially transferred to but does not stay with a market maker. The market maker expects to transfer the risk to another counterparty quickly and ensures this by continuously adjusting the prices at which it will accept a risk or transfer it away. The risk may get transferred many times, and the counterparty at any given time is usually unknown to the entity initially transferring the risk using the hedge. The counterparty may or may not be reducing its existing risk by diversification or hedging; the counterparty may be increasing its risk by speculation. In fact, the activity of speculators is essential to the risk transfer process in financial markets.
It is important to note that the process of risk aggregation in insurance is not the same as risk diversification. It is essential to the process of risk aggregation that the individual risks being combined into a block be similar to each other. Insurers underwrite each risk that they will accept into their block and reject risks for which the propensity for loss is outside the acceptable range for the block. The lack of correlation between these similar risks arises from the nature of the risk itself. For example, every house is exposed to the risk of damage by fire. However, with modern firefighting and prevention technology, it is rare for the burning of one house to increase the chance that another house will burn too.
Risk reduction through hedging involves two key elements. The process of risk dispersion reduces the impact of the risk being transferred by spreading it broadly to those willing to bear it for the right price. Speculators are essential for this process, which does not actually reduce the risk. The risk is reduced through diversification, which is achieved by combining inherently dissimilar risks, where the lack of correlation arises from the dissimilarity.
The need for underwriting in insurance points to another difference between risk aggregation and risk dispersion. In the risk aggregation process, there is a lack of symmetry in the information about the risk being considered for inclusion in the block. The entities inherently exposed to the risk and seeking to transfer it know more about their own exposure to the risk than the insurer does. The underwriting process reduces the imbalance of information, but there is the possibility of antiselection—which is the inclusion in the block of a risk that has a higher expected cost and/or higher variability than would normally be included in the block—due to inadequate information provided to the insurer.
In a hedging transaction, it is reasonable to assume that both parties have equal information about the risk being transferred. An efficient market is presumed to operate with transparency and liquidity. All relevant information about the risk should be available to the public in order to provide transparency for the market. There must also be a large enough number of investors who trust and understand the public information to provide liquidity to the market.
With either type of risk transfer mechanism, it is usually not practical (and often not possible) to transfer the entire risk. The entity inherently exposed to the risk must usually retain some portion of the risk. In an insurance transaction, this risk sharing is specified in the insurance contract through exclusions, deductibles, coinsurance and policy limits.
In a hedging transaction, there are two distinct types of risk sharing by the entity inherently exposed to the risk. First, the risk manager for the entity may select the extent of acceptable adverse deviation before the hedge takes effect—for example, by buying put options that are more versus less out of the money. The more out of the money the put option is, the less it costs, and the more the risk of adverse deviation is being retained. Second, the risk manager retains basis risk to the extent that the hedge instrument is not perfectly negatively correlated with the risk being hedged after the acceptable amount of adverse deviation has been breached.
Duration of Transfer
The time horizon over which the risk is transferred is different for insurance policies versus hedging instruments. Most insurance policies transfer the risk for at least one year. Some, such as whole life insurance, transfer the risk for many decades. On the other hand, most hedging instruments involve financial derivative securities, which usually have a maturity date that is less than a year from the transaction date.
Assurance of Payment
In either type of risk transfer, it is important for the party initially transferring the risk to be confident that the counterparty accepting the risk will deliver payment as promised upon the occurrence of the adverse outcome contemplated in the risk transfer. In an insurance transaction, this is accomplished through government supervision and regulation. Insurers must be licensed by the government regulator with jurisdiction. The insurer must demonstrate a certain level of available capital in order to obtain a license and must report on its evolving financial condition at least annually. The regulator may restrict the insurer’s ability to sell more policies or may intervene in other ways, including taking control of the insurer if the capital held by the insurer is found to be inadequate to ensure payment of all losses on all blocks insured.
In a hedging transaction, the exchange or market maker establishes rules about accepting risk, usually requiring assets to be held in a margin account or as collateral from any party interested in accepting risk and collecting a premium for it. If the assessment of the risk increases, the risk taker is usually required to deposit more assets. This mechanism ensures that the risk taker will be able to make the promised payment if the contemplated adverse outcome is realized.
Types of Risk Transferred
The type of risk that can be transferred by insurance differs from the type that can be transferred by hedging, although there is some overlap. For a risk to be insurable, there must be enough similar risks that are approximately uncorrelated so that a block can be aggregated, resulting in a reduction of risk. The risk should also be one for which the outcomes are either neutral or adverse. There should be no possibility of gain relative to expectations. For a risk to be hedgeable, there must be sufficient information about the risk publicly available so that any interested party can understand the risk, and there must be several independent potential counterparties willing to accept the risk at some price. Hedgeable risks often include the possibility of gain as well as the possibility of loss. Risks that are systemic, such as exposure to interest rate fluctuations, are hedgeable but not insurable. Risks that are specific, such as the risk that a particular person will suffer a heart attack, are insurable but not hedgeable.
Potential to Increase Total Losses
Although it was stated earlier that the total expected losses for society as a whole do not change with either an insurance transaction or a hedging transaction, there is actually potential under either type of risk transfer for total losses to increase, at least marginally.
After an insured purchases an insurance policy, the insured might be less diligent about reducing their exposure to the risk than they were before purchasing insurance. This potential for lax risk control is called “morale hazard.” Some insureds also intentionally defraud their insurer to collect indemnification for a loss that was not fortuitous. This is called “moral hazard.” Both hazards increase the total expected losses for society as a whole.
In hedging transactions, particularly those using financial derivative instruments, there is the potential for “speculative bubbles,” when a large number of speculators choose to increase their risk exposure in the same way at the same time. This disrupts the market equilibrium for the security or securities involved, eventually leading to a crash when the speculative bubble bursts. As the bubble expands, the total losses for society as a whole increase.
Effect of Accumulation
A key difference between insurable risks and hedgeable risks is the effect of accumulation. Since the fundamental process of risk reduction in insurance is risk aggregation, accumulating more insurable risks of the same type reduces the total risk for the insurer. If Allstate insured every automobile in the United States, it would have very little risk as measured by the coefficient of variation. On the other hand, accumulating more hedgeable risks of the same type increases the total risk for the investor. Long Term Capital Management learned the hard way that becoming the dominant writer of DOOM (deep out-of-money) put options on the Standard and Poor (S&P) 500 Index meant being exposed to a very large risk.
The difference in the effect of accumulation has important implications for risk management. For any type of risk, the effects of accumulation should be considered carefully when designing a risk transfer mechanism. Many risks that are insurable have residual correlations that limit the risk reduction that can be achieved through aggregation.
For example, an individual retiree is exposed to longevity risk because she does not know how long she will live, so her accumulated assets may or may not be adequate to meet her needs for the rest of her life. If this retiree participates in a defined benefit pension plan, the pension plan sponsor aggregates the longevity risk of the plan participants and reduces the total longevity risk of the group. However, the fact that mortality rates have decreased relentlessly but irregularly for more than a century creates a residual correlation between the longevity risk of the individuals in the plan. The plan sponsor can be reasonably sure that estimating future liability cash flows using only unadjusted recent experience will underestimate the funds needed to meet these obligations but does not know by how much. As more pension plans are aggregated together in a block, the risk of mortality improving more rapidly than anticipated becomes the dominant risk, and the benefits of further accumulation cease. At this point, hedging becomes the more appropriate risk transfer mechanism.
Implications for Enterprise Risk Management
In enterprise risk management, all risks to which the enterprise is exposed must be addressed. Some risks, such as reputation risk, are not transferrable and can only be addressed through risk reduction techniques. For those risks that are transferrable, the risk manager should be aware of the inherent nature of the risk to determine whether insurance or hedging (or some combination of both) is the more appropriate approach.
Statements of fact and opinions expressed herein are those of the individual authors and are not necessarily those of the Society of Actuaries, the editors, or the respective authors’ employers.
Cynthia Edwalds, FSA, ACAS, MAAA, is a Clinical Professor in the finance department and Executive Director of The Fred Arditti Center for Risk Management at DePaul University. Cynthia can be reached at firstname.lastname@example.org.