Making each business line count
Adrian Rands, Chairman, Quantemplate
It has become more important than ever to look at unlocking data opportunities to ensure each business line is carrying its weight to achieve underwriting profitability

News early this month that some lines of business, for some reinsurers (the mighty Munich and Hannover Re, no less), are not meeting their cost of capital serves to underline the importance of positive underwriting performance.

The confluence of soft market pricing, low investment returns, and new competition have made it more important than ever for underwriting operations to ensure each line of business written is carrying its weight.

The challenge is heightened by the approaching exhaustion of surplus reserves. Releases have carried the market for several years, but it seems likely that soft-market underwriting and less-than-prudent reserving (for events such as Tianjin) will reverse the trend, leading to strengthening for some companies in some lines. Paid claims ratios have been moving in the wrong direction for some time. Given this backdrop, in theory everybody must be underwriting for profit, all of the time.

In theory. But we all know what’s continuing to happen to rates. It is impossible to push them up in a softening market, even when you know the results of the previous year or two have been travelling south. That makes underwriting for profit all of the time, all the more difficult. That fact, and the current market situation, combine to make our ability to understand underwriting performance more crucial than ever.

Meanwhile, exposures are increasing in ways that can’t always be seen clearly, using conventional methods of accessing data, when monitoring underwriting performance. One example is the exposure creep arising under steadily increasing line sizes. This tactic has allowed carriers to maintain premium levels in a declining market, relative to the number of risks on their books – a key performance indicator (KPI) used by many. However, the premium/risk KPI masks a rise in assumed risk: the net effect of premium rate reductions and bigger lines is to crank up internal exposure.

The practice has an impact on reserving too. The traditional approach, based of triangulating loss ratio development, totally ignores increases in exposed limits or adequacy of premium rates. In the early years, claims development is likely to look similar, so conventional reserving methodology will forecast the same ultimate figures, even though actual underlying exposures may be much higher. Such granular drivers of loss development are easy to disregard, but may have a serious impact on a company’s underwriting performance.

Changing policy conditions are another force behind changing development. The hours clause is the most obvious example. Increasing the hours under the event definition can have the effect of transforming multiple independent meteorological events into single re/insurance events.

In the UK, for example, this year’s alphabet of recent named winter storms (particularly Eva-Frank and Gertrude-Henry-Imogen) have occurred in such a concentration that the extended hours clause may have a costly impact for reinsurers bereft of reinstatements. It is difficult to determine if such coverage extensions have been adequately priced.

The same can be said for the soft-market extension of geographic scope under many treaties. But since rates are going down, it is perhaps safe to assume it is not being priced.

Sometimes underwriters, armed with technological wizardry, which to a great extent is under their control, are able to justify broadened cover by manipulating modelled outputs. It’s easy to choose to ignore a loss on the edge of the distribution by declaring it an unrepresentative anomaly. As prices fall, model-tweaking becomes an ever-more tempting tactic to reduce loss projections.
Selecting the right loss ratio

The solution is to embrace a multidimensional approach to risk, one which captures, at the organisational level, the correct performance indicators, properly calculated, upon which to base forecast results. In other words, you have to pick the right loss ratios. Further, though, this information needs to be available all of the time, so that unexpected developments can lead to decisive remedial responses.

That demands a harmonious connection between underwriting and the actuarial team creating and analysing key performance indicators. Underwriters have to be focused on the KPIs which directly affect their day-to-day underwriting profitability, and those KPIs have to be designed with particular risk types in mind. Selecting them well demands that underwriters build in their product and class expertise. Generic KPIs won’t do.
Individual classes, lines of business and even specific geographies need to have their own KPIs, because the underlying risks are different. Political risk, for example, issues 10-year policies as standard. However, by the time such risks are grouped with other classes and examined from a high level, factors critical to exposures are camouflaged. Changes to risk due to hours clauses or discovery periods under liability policies are very difficult to spot at that distance, through the noise of a whole portfolio.

Once in place, a robust management information system is required, one that sends out signals when a certain type or geography of business is off-plan and throwing up anomalies. A handful of claims may occur outside expectations, but they have to be spotted to prompt action. When such anomalies are to be investigated, the process has to take place with easy access to the relevant contextual data, which helps to separate trends from genuine anomalies.

Achieving such a performance-monitoring regime is also in part dependent upon corporate culture. It is best achieved by companies that are agile. This agility must be demonstrated not just in the reaction to anomalies, but in the creation (and re-creation) of key performance indicators which relate directly to the value-added areas of a company’s business by ensuring each KPI has a direct link to underwriting profitability.

Agility must also be sufficient to allow underwriters to convert findings into actionable insights. Companies must be able and empowered to change their underwriting policy quickly enough to halt trends before they become serious losses.

To achieve that responsive level of underwriting performance monitoring of course requires that all the relevant risk, claims, and reserving data is collected and compiled in a consistent format, so it can be compared across datasets.

For many businesses, this technical challenge simply compounds and exacerbates the others, but on a fundamental level, adequate data analysis can be performed, and deviations identified, only when the data standards are at least interchangeable.

For many companies, the data required to unlock superior underwriting performance is present, but exists simply as a grey mass. Making sense of it may seem impossible, but doing so is critical to achieving profitable underwriting performance.

Originally published in Insurance Day on 18 February 2016

The Federated
Data Model
Taming the complexity of
regulatory reporting
Insurance models
for the technology age
Understanding changing
distribution channels
Making each business
line count
Brokers as
risk consultants
Four steps to
data-driven reinsurance
Portfolio steering
in the soft cycle
What is
responsive reserving?
The price is tight,
but is it right?
Improving data capture
through collaboration