Does Anyone Really Know What Time It Is?

Observations of inconsistency in excess coverage pricing in the professional liability market.

Nick Economidis is Vice President, eRisk at Crum & Forster. The opinions here are solely his own and do not represent opinions or positions of Crum & Forster or any other organization.

————————-

Although dice games have been around back to the age of the Roman Empire, it wasn’t until as late the 1600s before gamblers fully understood the odds of various number combinations associated with any single roll of two standard six-sided dice. This is not to say that gamblers didn’t have some idea that rolling a combination of 7 was more likely than a 12, but the precise odds, that we seemingly take for granted today, were not commonly known^[1].

Despite a wealth of actuarial time and attention on historical professional liability results and rate-making, it appears that underwriters only have vague and imprecise information of how to price excess professional liability coverage. As an industry, we seem to exist in sort of a Dark Ages of only understanding a vague semblance of the odds without precision.

For the uninitiated, rates in the excess professional liability market are generally expressed as percentage of the underlying premium, and it is also worth noting that underwriters typically offer excess limits in the same size of blocks as the primary policy. For an equal block of coverage, 60% of underlying exists as a standard benchmark. It important to note, however, that different segments of the professional liability market often gravitate toward higher or lower benchmarks for market-standard pricing, and different underwriters will voluntarily elect to use a higher or lower percentage of the underlying premium to price specific accounts based on various underwriting considerations. Even if inconsistent, the professional liability market for excess coverage is both competitive and dynamic.

Nevertheless, the market exhibits a number of inconsistences that suggest that underwriters are not fully understanding the odds of loss associated with the terms being offered. To illustrate the lack of consistency, consider the pricing in the Table below for a hypothetical risk based on a standard excess rate of 60% of underlying.

The hypothetical risk in Table 1, shouldn’t look too usual to any participant in the excess market for professional liability insurance. Table 2 presents another similar example involving multiple layers.

Both of the above charts should look perfectly normal for underwriters, brokers and other participants in professional liability markets. But, there’s a catch. Do you see it?

The catch is that both towers are pricing the same risk for the same amount of coverage but yielding significantly different premium results. A summary of both towers is in the Table below:

• Both towers provide $20,000,000 of coverage.
• The first $10 million of coverage for both towers produces a premium of $100,000.
• Tower #1 generates a total premium of $160,000 versus a premium of $136,000 for tower #2; a difference of 15% for the same hypothetical risk for the same total limits. With no clear explanation, tower #2 produces a premium savings of $24,000 (or about 15%).

It stands to reason that in a market where the market participants have a reasonable understanding of the odds, the premium for a tower of $20 million of coverage should be approximately the same. There’s no real logical explanation for a 15% difference between Tower 1 and Tower 2.

If you doubt that a 15% difference is significant, there’s a second catch worth noting. The 15% difference in the total cost of the tower is achieved by a reduction on the $10M excess $10M portion of the towers, The 3^rd and 4^th excess carriers in Tower 2 have essentially accepted the same risk as 2^nd excess carrier in Tower 1 for a total premium of $36,000 compared to a $60,000; a significant discount of 40% on this layer.

Again, for emphasis, the dynamic displayed in the Tables above is common and normal. At least one broker has even anecdotally expressed that they are aware of this inconsistency and, at times, will manipulate the structure of the tower to attain a more favorable result.

Given these observations, it’s impossible to argue that professional liability underwriters pricing excess coverage have properly calculated or considered the odds of rolling a 12 (or in this case, a loss in excess of $10 million), and this is creating significant inconsistencies in the pricing of excess risk. It’s not that every underwriter should be using the same curve, or actuarial analysis, to price excess. Rather the free and open market requires that responsible underwriters develop their own pricing guidelines and independently decide if and when to deviate from those guidelines.

Involuntary and unintentional discounts of 40% don’t reflect sound underwriting. The obvious conclusion here is that underwriting officers should work with underwriters to fully understand the odds associated with excess loss and to provide underwriters with better tools to enable consistent pricing.

So, the question is raised, does anyone really know what time it is?

—
Postscript: What’s an underwriter to do?

Foremost, underwriters are encouraged to seek input and assistance from their actuarial colleagues. There’s a clear and undeniable inconsistency in many common excess pricing models. The more underwriters can understand the size, magnitude and frequency of large loss events, the better their position to properly select and price risks seeking excess coverage.

Second, underwriters should consider a more holistic approach to excess pricing. In general, it appears that inconsistencies in excess pricing are often related to the underwriting practice of pricing each respective layer solely based on the pricing of the layer directly beneath the proposed layer. Underwriters should consider the pricing of all of the underlying layers in pricing each new layer.

To illustrate this approach, let’s take a fresh look at the hypothetical risk number 2, and give fresh consideration to how the 2^nd (and 3^rd) excess layers might be priced. As previously illustrated, the primary and first excess layers generated a premium of $100,000 for the first $10,000,000 of coverage.

Following a holistic approach, in order to price the 2^nd Excess ($5,000,000 xs $10,000,000), the underwriter should start by considering the pricing of each underlying layer. Presuming that the underwriter wants to charge 60% of the underlying premium, the underwriter can quickly and easily determine that the premium for the $10,000,000 of coverage excess of the underlying $10,000,000 should be $60,000 [= ($62,500 + $37,500) * 60%].

If the underwriter is satisfied that $60,000 is a fair premium for $10,000,000 of coverage excess the underlying $10,000,000 of coverage, then the underwriter simply needs to determine what share of the $60,000 excess premium they feel is reasonable for their proposed 2^nd Excess and what should remain for the 3^rd Excess Layers. For illustrative purposes, if the 2^nd excess underwriter decides that 55% of the $60,000 premium should be allocated to the 2^nd excess, and the underwriter pricing the 3^rd excess takes a similar approach while deciding that 45% of the $60,000 premium is fairly allocated to the 3^rd excess, the pricing of the hypothetical tower would look like this:

If you compare the total pricing the $20,000,000 of coverage for Hypothetical Risk #1 with the revised pricing for Hypothetical Risk #2 in the table above, you’ll note that both towers produce an identical premium of $160,000.

In fairness, it is not correct to say that any of the hypothetical pricing illustrated in any of the examples generates the right premium. Rather, the key point illustrated here is that when two different structures for the same hypothetical risk routinely produce premiums that deviate by as much as 40%, the inconsistency suggests a significant flaw in how underwriters are routinely pricing risk. The holistic approach illustrated above reflects one approach to produce more consistent pricing of risk across various program structures.

^[1]Peter L. Bernstein (1998) Against the Gods: The Remarkable Story of Risk. John Wiley & Sons, Inc.