November 6th, 2008

Defining the Value of Risk Management

Posted at 1:00 AM ET

John Major, Senior Vice President, Instrat
Contact

How do you put a price on risk management? In the early days of finance theory (1950’s), the value of risk management was questioned—unless, of course, it was costless. The nuances of a more complex business environment have rendered this position untenable, but we still struggle to quantify the benefits of risk management, especially in the (re)insurance industry. Thus, the fundamental activity of risk-bearers has not been measurable, leaving a cloud of ambiguity in the middle of every carrier’s operation.

This changes now.

Guy Carpenter’s Firm-Value Risk Model (FVRM) represents a significant step forward in quantifying the value of insurer risk management through reinsurance. Supported by GC-VALOR™ modeling software, the FVRM method helps (re)insurers arrive at a realistic sense of risk management value. Using an approach that centers on shareholder value, the framework delivers a consistent perspective that relates risk and cost to firm value.

The Challenge

Traditionally, risk management alternatives have been evaluated along the two dimensions of cost and benefit. The first, net cost, is straightforward. It consists simply of the out-of-pocket expense associated with transferring risk from a portfolio—minus the expected value of any recoveries. The benefit measure can vary, involving a reduction in a risk metric such as economic capital (i.e., value at risk), ruin probability, or earnings volatility. Sometimes, a more exotic risk measure is used.

Carriers evaluate risk management alternatives using the two metrics (i.e., net cost and the benefit risk metric selected). After filtering the options that are obviously inferior (such as paying more for less risk reduction), the choices that remain—collectively known as the “efficient frontier”—represent a spectrum of increasing cost versus decreasing risk. If one option is deemed superior according to both metrics used, it is selected. Usually, however, there is no clearly superior choice – more risk reduction is usually available if the insurer is willing to pay for it. At this point, the decision-making process becomes murky. Management is asked to apply its own risk preferences, with little guidance as to which trade-offs would add the most to firm value.

Modeling Financial Risk

Financial risk modeling is the first step in ascertaining firm value, which ultimately leads to a determination of the value of risk management. The cornerstone of this method involves distributed earnings (such as dividends). Shareholders’ net present value (NPV)—i.e., firm value—is reached by applying cash flow discounting to the stream of expected payouts to shareholders.

Shareholder Value = NPV(Expected Future Shareholder Payout Stream)[1]

In this model, known as the “discounted dividend model” of equity valuation, a rigid relationship between market value and a firm’s capital does not exist. If market value dips below the amount of capital held, the firm becomes a takeover target, as it could be liquidated at a profit. In this situation, the optimal strategy might be to return all capital to shareholders immediately (if the firm were allowed to do so). On the other hand, a firm with a market value that is greater than the capital it holds is said to have “franchise value.”

Franchise Value = Market Value – Capital Held

Most firms, consequently, seek to create franchise value … and face the challenge of deciding whether to retain or pay out earnings. The latter, of course, directly affects franchise value, while retaining earnings for reinvestment in growth initiatives can create more value—at the expense of both time and risk.

At the core of the decision to retain or pay out is the cost of capital. In theory, a firm should return to shareholders all capital that is not needed immediately and simply secure new capital (through credit or equity markets) when needed to finance future projects. This oversimplified theory, however, does not always correspond to reality. The transaction costs associated with raising new capital may make keeping and later reinvesting retained earnings a more attractive option. Further, market prices for capital reflect risk, and an investor who sees a company in distress will adjust the price charged for financing accordingly, which can hurt the insurer who waits to raise capital until it really needs it.

At present, there is another concern. The notion that capital will be available (even if at a premium price) conflicts with the capital constraints evident in today’s marketplace—especially in credit markets. Retained earnings, in extreme conditions, may be the only viable source of capital. Thus, the decision to pay out profits must consider the risk that external capital will not be available in the future.

Regardless of prevailing financial market conditions, the starting point is always the same: the amount of capital a firm has on hand today. If a firm does not have enough capital to support a long-term stream of dividend payments, the result is a lower NPV, because, in all likelihood, the insurer will not have the earnings to pay dividends for very long. But, adding a little capital could translate to a significant increase in shareholder value – much more than the amount of capital added. On the other hand, increasing the capital supporting an already secure payment stream would have a smaller impact on future discounted payouts. Adding capital in this situation would decrease franchise value, while paying dividends would decrease value by less than the dividends paid. Part of the problem here is that holding capital involves frictional costs, such as taxation of investment earnings, so after some point it is more profitable for the shareholders to hold the excess capital themselves.

The relationship between a firm’s capital held and market value, therefore, is not a straight line. It is instead curvilinear—and is appropriately called the “M-curve” (where “M” is for “market”). Both capital market factors (like the higher costs of raising new capital) and insurance product market factors (such as policyholders’ desires for a secure insurer) affect the shape of the M-curve.

Central to a firm’s M-curve is the optimal payout strategy. At what point, essentially, is capital excessive?

How Much Is Too Much?

For the capital-rich (re)insurance industry, “excess” capital has been a topic of conversation through most of 2008. Several consecutive benign loss years have left carriers flush with cash. Even with the acceleration of the credit crisis into a financial catastrophe, balance sheet damage has still left insurers with a premium-to-surplus ratio of 0.88; reinsurers remain slightly above 1.0. The discussion of capital repatriation is as relevant as ever.

In order to maximize the value of a (re)insurer as defined by the shareholder NPV model—which includes optimizing its risk management practices—Guy Carpenter has developed the FVRM. The first step of the FVRM is the capital model:

Change in Capital = Predictable Earnings – Random Losses + Capital Inflows – Capital Payouts

Change in capital, quite simply, involves subtracting random losses and capital payouts from a firm’s predictable earnings (like premium and dividends) and adding any inflows of capital. In this definition, normal fluctuations are combined with catastrophic jumps.

Interestingly, the FVRM assumes that predictable revenue can be a function of a firm’s current level of capital. A phenomenon called “customer risk aversion” is at play. Policyholders want their insurers to be able to pay claims when they arise and consequently want the carrier to be financially sound. If a carrier is not on solid financial footing, insureds will refuse to pay premiums that are as high as those they would tolerate for a stronger insurer. This “penalty” could exceed the actuarial “fair value” of the risk of claims non-payment by a large factor. Concerns as to the risk of insolvency therefore could constrain earnings potential. This issue must be addressed before a firm seeks to maximize the second part of the FVRM—shareholder value:

Shareholder Value = Maximum {Expected [ NPV (future payouts – (1+k)*(future inflows))]}

This is just the original shareholder value equation (above) with the payout stream split into outflows and inflows and with a factor k included for the cost of raising external capital. Under normal circumstances, a large firm can expect to issue new equity with underwriting and administrative fees totaling around 5 percent. A factor of k over 30 percent is hard to imagine. However, the financial distress of a firm after a catastrophic loss might not be considered “normal circumstances.” In conditions of extreme uncertainty, there is a chance that investors in the capital markets could require expected capital gains amounting to several multiples (i.e., hundreds of percent) as a cost of injecting new capital. This could show up, for instance, in the dilutive effect of issuing new shares on the value of existing shares.

The Value of Risk

The benefits of effective risk management practices are extensive, from perception of the firm to the reduction of earnings volatility. Yet, it has been difficult to quantify the value added by such practices. Ultimately, risk management drives franchise value, that is, the value that a (re)insurer adds to the capital it holds. Using the FVRM framework (re)insurers can quantify the financial advantages of their core activity to their customers and the value of their own risk management as well.

In economic terms, the market value of a risk management program consists of the difference between the market value of the firm with the risk management program and that of the firm without the program. The result of this calculation is the amount that risk management adds to franchise value.

Risk management affects the amount of capital a firm holds in two ways:

  1. The cost of risk management can decrease predictable earnings (and consequently shareholder payouts)
  2. The benefit of risk management is to make the probability distribution of random losses more favorable

In exchange for the cost of risk management, a firm reduces the likelihood that a catastrophic loss will push it into bankruptcy—which would stop shareholder payouts completely. Also, it protects the firm from the probability of having to operate in financial distress, which generates costs of its own, such as distractions of management time, dealing with and paying for outside oversight, suffering employee turnover, and raising capital at unfavorable terms.

Taking Action

The operational steps needed for FVRM start with modeling the firm’s risks in probabilistic terms, including how those probabilities would change under different reinsurance alternatives (with prices also modeled). Another input required is some quantification of the customer risk aversion effect. This could be a ratings cliff – a drop in capital below some level will cause the insurer to lose business as well as pricing power for the business it keeps. Having a measure for the cost of raising external capital and how that relates to impairment level is also necessary.

Solving the FVRM with GC-VALOR™ modeling software entails finding the shape of the M-curve and determining the reinsurance program and dividend-paying strategy that will optimize shareholder value. Often, this means retaining earnings until capital reaches its optimal level on the M-curve. Then, the carrier pays out all earnings in excess of that level. Further, this approach typically includes a reinsurance strategy that varies with a company’s position on the calculated M-curve. By comparing the shareholder value under the optimal strategy to those from alternative strategies, the FVRM quantifies the value of different risk management approaches.

By focusing on the sources of franchise value, the FVRM approach gives (re)insurers a tangible way to assess their effectiveness. The decisions made in the process of identifying and hedging risks can be evaluated by their impacts on the value of the firm to shareholders. The result is a consistent yardstick for measuring the trade-off between risk and reward.

Additional Contributors:

  • Donald Mango, Chief Actuary
  • Gary Venter, Managing Director

Footnotes:

  1. Note that this is not the same as market capitalization, which is the value of all outstanding shares priced at the level that those shareholders who want to sell right now would take for their shares. Shareholder value as defined here is usually higher than market capitalization, as can be seen in premium prices for acquisitions, and is also more stable.
AddThis Feed Button
Bookmark and Share


Related Posts