Personalised risk data profiling is building stronger resilience

[pg:TP-Personalised-risk-data]

In-depth and accurate data from onsite risk surveys deepens insight and builds resilience

Accurate data collected onsite by professional risk engineers drives pinpoint underwriting. It ensures loss prevention programmes are effective, allows companies to engineer risk out of their operations and empowers them to be more resilient.

At the top end of the commercial property sector, where site values can run to billions of pounds, data-led underwriting and risk management, based on real information and in-person observations, deliver operational and financial benefits that a statistical modelling approach cannot match.

Limitations of statistical modelling

Pure statistical modelling, which does not incorporate real risk data, is prevalent in many aspects of the low and mid-tier commercial property markets and the technology is increasingly sophisticated. But for the largest and most complex commercial property risks its effectiveness is limited. There are just not enough comparable risks at this end of the market to make pure statistical modelling as reliable or accurate as it needs to be.

For example, there are no end of small, high street retail properties and their volume and similarity lend themselves well to statistical modelling. But there are relatively few complex chemical plants or power stations dotted around the world. The age, design, location and construction of these plants can differ significantly and effective statistical comparisons for underwriting and risk management purposes can often only be made at a macro level.

Onsite risk evaluation surveys allow professional risk engineers to assess and understand each site in detail. Working to the COPE methodology – construction, occupancy, protection, exposure – enables them to analyse and quantify risks on a consistent basis.

Physical surveys catch up to 1,000 data points, some of which indicate likelihood factors while others quantify potential severity of exposures, which adds depth to the risk assessment. This detailed and flexible approach creates a rich, accurate and reliable data set for each location.

FM Global operates to this standard and visits around 60,000 sites annually. Its 1,800 risk engineers attend some sites more than once, completing a total of 100,000 surveys and collecting about 75 million data points each year.

The road to resilience

Collecting detailed data that is specific to an individual client’s risk allows insurers to price cover more accurately, benefitting both the carrier and the policyholder. It enables both parties to negotiate, design and implement insurance programmes that reflect the exact nature of underlying exposures instead of relying on standard wordings and schedules.

The data also informs the design of hazard management and loss prevention strategies that guide insureds how to engineer risk out of their businesses and improve enterprise resilience. Based on up-to-date and professionally captured data, these strategies will identify and mitigate the biggest exposures borne by companies and deliver steady improvements in existing risk profiles.

Where losses occur, risk engineers also have an important role to play in capturing data that will help refine and evolve risk management and prevention strategies.

Conducting onsite, post-loss surveys enables them to identify the root causes of the event. The analysis lets insurers build up a more detailed understanding of individual risk factors and the frequency and severity of correlating loss outcomes. These factors can then be addressed directly in the evolving risk management programme, ensuring it is constantly refined and improved.

Predictive analytics that rely on real-life data rather than statistics

The richness of real-life data gathered on individual risks also allows insurers to improve their risk prediction and modelling capabilities.

For example, FM Global ranks the 60,000 risk locations it insures for their perceived likelihood of having a loss valued at over $3m in the coming year. This ranking is based largely on the data collected during onsite surveys. The insurer has found that around 30% of these large losses come from the top 1,000 (1.67%) locations on that list.

Prioritising the design and implementation of detailed risk management and mitigation programmes at these locations benefits both policyholders and the insurer. It also demonstrates the value of loss predictions based on real data.

FM Global is also using its wealth of data to develop its ability to predict machinery and equipment failures. When breakdowns happen suddenly and/or unexpectedly, they can create significant losses for businesses due to the immediate impact on output and the long lead times that generally exist to source parts or find replacements.

Where insurers can help inform the risk management process of the effectiveness of maintenance programmes and schedules for essential systems and equipment, this can serve to reduce the likelihood of unexpected failures and the business interruption losses they generate.

Understanding the root causes of machine failures also informs training programmes for staff, improving operating procedures, safeguarding users and extending equipment lifespans. Developing effective predictive analytics for specialist machinery used for very specific purposes is only possible by inspecting the equipment physically and assessing how it is operated and the workload it is under.

Data is currency in today’s digital age, but it is the quality rather than the quantity of data that generates value. No matter the computing power available to sift through and analyse data mountains, information that is inaccurate, irrelevant or out of date will not generate valuable insights.

How does your insurer go about understanding your complex commercial risk? What data does it collect and how does it use it to improve your company’s resilience?