The cost of property and casualty risk of American businesses, non-profits and governments is roughly $200 billion, taking insurance and self-insured losses into the equation. How organizations account for these costs can range from a model of lucidity to a perverse maze. Thankfully, there are some best practices.

Many organizations charge operating units their fair share of risk and insurance costs. Most controllers and risk managers like to do so. About half seem to believe that getting operating units to budget for the cost risk will lower the cost of risk. Even executives who will not conjecture a cost savings still feel the practice is a good one.

In a recent survey of risk managers and risk consultants, about half confidently predicted that a well-run allocation system would generate savings in cost of risk of about 5% to 10%. Thus a company with a million dollars in annual P&C losses could save from $50,000 to $100,000, simply by making managers more financially accountable for the losses in their departments. The most popular candidates for allocation were workers compensation and liability exposures.

But getting the allocation system right is may not be easy. As one risk manager noted, the system “must be perceived as fair and impartial or it will cause those affected to thwart the system or to leave. It must also have a strong enough impact on the bottom line of those in the best position to reduce loss costs.”

Examples of flawed, even perverse allocation practices are easy to find. Some years ago, a division of a major paper products company assiduously reduced its workers compensation clams costs by 90% only to find that home office increased its assessment for the next year. Its allocation policy simply did not take recent experience into account. In another example, a renowned non-profit repeatedly assessed its local chapters for insurance at rates which were far in excess of what were prevailing market quotes. Without telling the chapters, home office was building up a large surplus in its captive.

Risk managers appear to divide into two camps about how to allocate. The primary concern of one camp is year-to-year stability in assessments to operating units. They aim to make these allocations as predictable as possible. The other camp is concerned mainly about the accountability of divisional managers. This group of risk managers tends to favor allocations which reward or punish behavior, fast.

Every method starts with an aggregation of costs called “cost of risk.” Since the mid 1990s, the Risk and Insurance Management Society and Ernst & Young, the accounting firm, using an annual survey, have tracked prevailing cost of risk for corporations. They count insurance premiums, retained losses, internal risk management, outside services, financial guarantees, fees and taxes. For organizations with $100 million of less in annual revenues, cost of risk has been about 3% of revenues. For those with revenues of over $1 billion, cost of risk has been under 1% of revenues. The vast majority of costs relate to workers compensation and liability.

Actuaries like Kevin Bingham, in the Hartford, CT, office of Deloitte & Touche, can come up with estimates of costs if contingencies are large and can only be guessed at by non-actuaries. Per Bingham, “We can use methods with great names like the Bornhuetter Ferguson method weight actual experience with expected experience, if need be.”

Once the base costs are counted up, the next step is to allocate, or charge back, the costs to divisions. A simple way to allocate costs is to spread the total cost of risk, by line, according to the appropriate common denominator for the line. Thus workers compensation costs could be allocated by full time equivalent employee count or payroll. Liability risk costs can be assigned by sales dollars or units of service. This approach emphasizes stability.

But the simple way can hardly be the best way. It fails to create any serious incentive to reduce the cost of risk. Divisional chief executives might have average tenures of, say, three years. Cost control investments, however fast-acting in cutting claims, are slow to trigger a reduction in allocation. From the perspective of the divisional chief, his or her next promotion in sight, spending scarce dollars on lowering cost of risk might even appear a waste of resources.

Bingham believes that charging back the cost of risk to operating units “can really help an organization drive down its costs by motivating line managers to focus on the importance of return to work initiatives.” “As I have observed at clients, nothing motivates managers more than a compensation system that weighs in the cost of risk when determining their final salary.”

Best practice

A search resulted in several finds of what one might call “best practice.” Aimed at self-insured organizations, with some modifications they can work well to for insured entities.

United Technologies deserves a gold medal with actuarial clusters for an ingenious approach. A since retired executive, David Bowen, originally authored it, and George Levin, a consultant, later refined it. What follows is a somewhat simplified version.

United Technologies values its claims shortly after the close of a period, such as a fiscal quarter. The total value of new claims for an operating unit is computed as a percentage of the new claims value of all operating units. The company then applies this percentage to an actuarially set total “premium” cost for the company, and allocates each operating unit its proportional share.

This approach allocates 100% of “premium” to operating units. This is an important goal as United Technologies maximizes indirect cost reimbursement from governmental customers, which have strict guidelines on indirect cost accounting. .

The focus on new claims cost promotes cost reduction. Even stronger incentives to reduce the cost of risk can be bolted in. For instance, to encourage reducing the frequency of claims, a corporation using this approach could base the premium allocation on a blending of an operating unit’s share of total new claims valuation and share of new claim volume. What about large single losses? United Technologies has a provision to cap large losses so that rare events do no wreck havoc on departmental budgets.

The United Technologies approach has ingredients that risk managers say they want in their allocation method.

• The allocations link to new claim trends. Division executives have a clear incentive to prevent and manage new claims.
• The methodology is acceptable under federal standards for charging risk costs to contracts and grants.
• Allocations can be computed frequently based on recent experience.
• 100% of the cost of risk can be allocated to operating units.

This approach has yet another virtue: it is easy for divisional executives to understand. In contrast, one public entity with over $100 million in annual losses uses an allocation method which one person alone is qualified to explain – if one has an hour to listen.

Lost Time Method for workers compensation

Many organizations tend to think of workers compensation differently than they do other risks. A special method for this line of risk may be a good idea. For example, focus attention on reducing lost time. A lost-time calculator can allocate not just workers compensation but also short and long-term disability costs. Total cost of unscheduled absences has been estimated by the Integrated Benefits Institute of San Francisco at 10% of personnel costs.

In one scenario, a corporation at the end of every quarter computes the accumulated lost time of every new lost time claim through the end of the period. It assigns a standard cost per lost day, derived from analysis of prior claims, reflecting indemnity, medical and other costs. For claims that remain out of work at the end of the period, it can add an additional standard cost per claim. (A claim with 5 lost days at the end of the period to date at period end could be assigned an additional cost much lower than a claim with 60 lost days to date.). The corporation can allocate the total cost of workers compensation risk proportionately.

An alternative method recognizes lost time claims from prior periods as well as new claims. All lost days incurred within a period are given a per day cost value. The corporation allocates total workers comp costs in proportion to the division’s share of total lost days for the entire corporation.

Accelerated Experience Mod Method

Want to inject the right incentives but must use an allocation system which does not quickly reflect recent loss experience? Then consider the following method. Massachusetts used it successfully in the 1990s to sharply lower claims costs in the workers compensation assigned risk pool. An Australian insurer is now using it. If it can work for an insurer vis-à-vis hundreds of insureds, it can work for a corporation with multiple operating units.

Step One - invite operating units to demonstrate competence in a portfolio of safety and claims management practices known to have very fast implementation times. Certify them through an audit process. Then, award them a “credit” to regular allocations. A typical formula for workers compensation is a credit for three years for 15%, 10% and 5%. By the third year the normal, slow-moving allocation system will have essentially recognized the improved loss experience.

The key to this method is that divisions often can produce fast, big improvements. In both the Massachusetts and Australian cases, the plan is designed to save twice the amount of the credit and thus create a big financial cushion. The risk management department acts as if it were an insurer. It makes an “underwriting” bet that the improved results match or exceed the discount.

Allocating cost of risk is today built into corporate culture. A huge share of managing risk is borne by corporate departments, not by home office executives. Is there any better way to spell “accountability”?