Political risk underwriters do not attempt to get inside a terrorist’s mind. Solid underwriting principles apply as much to this sector as any other. David James describes the evolution of a market with confused origins and an uncertain future.

Political violence underwriting has come along way since the mid-1990s when Lloyd’s took the first tentative steps towards a standalone alternative to Pool Re. At the time, terrorism insurance fell between the spheres of the property and political risk market. Many of the main players in today’s market forged their techniques, modelling and underwriting philosophy during this period. Today, the terrorism arena is discussed as if it were just like any other discipline despite its confused origins and questionable future as a standalone class.

Terrorism is inevitably an emotional subject. For this reason, terrorism underwriters are never short of interview requests and are often asked for an insight into the process. There is a fear that responses could be sensationalised in print and therefore, underwriters are not always willing to provide answers.

Common myths

“Do you try and get into the mind of the terrorist?” is a common question posed to terrorism underwriters. Trying to second guess a terrorist is not the underwriter’s job. Fundamentally, underwriting terrorism risk is much like any other class of insurance. The principles of good underwriting apply here as they do in all other areas. Aggregation control, efficient management reporting, timely credit control, focused marketing and distribution channels are all important.

There is, however, an additional layer to the underwriting process that is particular to terrorism – that of subjective risk assessment.

“Do you think that a terrorism event is going to happen at a given location during a particular period?” It is possible to become weighed down by such theorising but if the objective underwriting factors are properly controlled, in many cases the subjective becomes almost irrelevant. The first responsibility of an underwriter is to control his exposure.

One of the reasons that terrorism cover is expensive in Manhattan zip code 10017 is the accumulation of exposure, rather than the threat level in New York or the target probability of Lexington Avenue. That is not to say we are unconcerned with occupancy or attachment point – these are very relevant factors – they are just not the main drivers of appetite or price.

Moving away from the high accumulation areas, subjective factors become increasingly important. This is particularly true with large industrial exposures such as utilities, semiconductor and desalination plants. Here the scale and spread of the risk is such that even in the event of a large bomb or attack it is unlikely that more than one “risk” will be physically impacted.

This is where the analytical skills of the political risk underwriter allow him to differentiate the acceptable from the unacceptable risk. One of the interesting aspects of this market is that the more challenging the risk, the more varied the underwriting response. The risk ceases to have a “market” price and the broker has to arrive at a consensus between disparate positions.

Having managed the aggregate position the risk is then evaluated in terms of its likely impact post loss. Is the event significant enough to “turn” the market – or even make it pause for thought? Finally, there is the “I told you so” test. Does a particular loss leave us open to the wisdom of hindsight? Should we have seen it coming? Terrorism isn’t random. Necessarily the protagonist seeks a target where he believes he can realise the maximum impact, and the highest likelihood of success. This is unlike any other peril and highlights the importance of effective risk management.

Use the right tools

“It seems ironic that the legislation currently being considered may indeed threaten the viability of the very thing it was designed to encourage

Aggregation control is at the heart of what terrorism underwriters do. In 2002 the accepted view was that underwriters needed a predictive tool capable of identifying likely targets. Understanding the risk is crucial but an aggregation tool needs to be a determinative instrument first.

The determinative approach was not initially accepted by all but it is now becoming the market standard. Unfortunately for those seeking off-the-shelf products, developing the systems takes time and populating the tool with the right quality of location data is crucial. Fortunately, most brokers have recognised that it is in everyone’s interest to provide underwriters with good quality address data including zip codes and street numbers. The more accurate the data the better underwriters are able to monitor their accumulations and maximise the available capacity in any given blast zone.

To underwrite effectively in this class it is crucial to have an effective pre-risking tool that is able to deliver results quickly. It is equally important to monitor this dataset regularly to anticipate accumulations before they become critical, rather than simply reacting to having reached a maximum accumulation.

“What is the impact of short-term events and incidents?” It is inevitable that after every terrorist attack underwriters will be asked to assess the impact on the market. Post-2001, most attacks have not led to large insured losses. Therefore the impact has been to affect risk perception rather than just price.

The nature of the loss has focused underwriters’ attention on one or more of a combination of price, appetite or aggregation control. For the most part, the impact on price over the last six years has been negligible in the immediate aftermath of an attack. However, this is not the case where a shift in pattern or trend is evident. In the wake of ex-prime minister Rafik Hariri’s assassination in Lebanon, for example, there was a long-term deterioration in country risk and this informed appetite for business in that region.

In future, if an event showed that new attack devices or methods were being used, it could have a more long-lasting impact on price and appetite in the market. Underwriters would need to develop new aggregation methodology. At present, the standard industry event that we model our “blast zones” to is a two-tonne truck bomb and this was used in the attacks in Bishopsgate (London) and Oklahoma (US). If it becomes evident that the “improvised explosive device” has been replaced by an alternative method of delivery then it may be prudent to reconsider our methodology. Fortunately successful attacks are limited and the opportunities for testing theories likewise.

The fundamentals of terrorism underwriting are no different from any other class even if the loss trigger is perceived as particularly challenging. The subjective risk factors may evolve but they are principally relevant to challenging our assumptions on aggregation in the long term rather than being the main driver of price or underwriting appetite in the short term.

The way forward

The perceived wisdom is that the Terrorism Risk Insurance Act (TRIA) and its extensions should be a permanent part of the insurance landscape. Yet it seems ironic that the legislation currently being considered may indeed threaten the viability of the very thing it was designed to encourage, namely a long-term viable standalone terrorism market. If TRIA is cancelled it would prove challenging for certain classes – most notably workers’ compensation – but would inevitably bring new investment and capacity to the class. Insurers would find the technology to provide cover for all but the most extreme risks.

Lloyd’s can be proud of the way it has risen to the challenge of a blanket exclusion on terrorism post 9/11. The market is now entering a crucial period where its shape and size is likely to be significantly affected by the path the legislature within the US seeks to take on TRIA. Whether we like it or not, the last months of 2007 will have a significant impact on where this new market goes. As underwriters, all we can seek to do is be ready to meet these challenges head on.

Terrorism: Something's gotta give

There is much talk at the moment of an industry-wide softening. It is inevitable that a class with excellent loss ratios attracts new capacity. Nevertheless, there is a two-speed market at the moment. In certain high accumulation areas the new capacity has had little impact on pricing.

New underwriters in the terrorism market in certain cases lack the necessary skill, controls or commercial acumen to achieve consistent pricing. They will generally find that they have quickly utilised their aggregate on very competitively-priced business. Non-clash risks can almost be considered separately as there are fewer constraints on aggregate and so capacity and competition is plentiful.

The market is delicately poised. Many participants are poorly balanced with insufficient net income to pay for their likely net retained loss. A major loss, especially within the US, could therefore have a decisive impact on the market. The market seems to be poised on a fulcrum – which could easily tilt back post a major loss. This makes for an uncertain period as we look forward to 2008.

2008 will be an increasingly difficult period for buyers, brokers and regulators as the market “relaxation” continues. It is likely we will see an underwriting market that remains soft on price but focused on underwriting controls and coverage. Whether, in the absence of any major events, the mathematics will continue to make sense beyond 2008 is another question. For many it is likely that there will be insufficient net income to ride out a loss. We are almost “waiting for something to give”.