Willis and Towers Watson have merged. Visit

Today's analytics – changing our understanding of the past and the future of risk

John Merkovsky | 20th April 2015


The future is not what it used to be. Neither is the past.

As we build our stores of data and begin to analyse this data we are discovering correlations and trends that had not been readily apparent in the past. These correlations are challenging, if not upending, the way we look at the future. Still, as the proverbial black swans become increasingly grey, we know that data cannot entirely predict future events.

So how do we get the most from this “new past?” That is, how can the study of data and the building of analytical models powered by insightful algorithms change our understanding of the past in a way that gives us a view into the future?

With more to learn, this future is coming into focus for all of us in the risk world. At its most basic, predictive analytics allows us to align current risk management priorities with spend. It empowers us to embed risk management decisions into C-suite strategies. At the edges, it is beginning to allow us to think about risk finance in new ways, supporting the development of new and often customized risk transfer solutions for risks historically considered “uninsurable.”

Better weather and natural catastrophe forecasting

Take weather for example. Weather forecasting and natural catastrophes are two areas where analytics has already made a real difference. Just since Hurricane Katrina ten years ago, many improvements have been made to the use and application of catastrophe models – including:

  • the collection of more extensive claims and engineering data
  • the introduction of climate change features
  • the use of more accurate exposure data

After Hurricane Katrina flooded more than 80% of New Orleans, catastrophe models were improved.

This results in a much better view of prospective catastrophic events, such as hurricanes, tsunamis, windstorms and earthquakes.

Similarly in this area, new and alternative sources of data –including amateur weather observations via smart phones and even social media – are increasingly being used to help predict and manage flooding, heat waves, and other extreme weather events.

Faster restoration after catastrophe

Most exciting are the societal ramifications. Better analysis of better data is helping emerging economies mitigate the impact of drought and other natural catastrophes, boosting productivity and reducing the threat of food scarcity.

Index-based risk finance capabilities, for example, analyse rainfall data in certain African countries and pay out as soon as severe weather events occur—minimizing losses and speeding the restoration process. The African Risk Capacity (ARC) catastrophe insurance pool is one such vehicle, supported by US $55 million of index-based reinsurance capacity from the international weather-risk markets (secured by Willis). The underlying insurance policies issued by ARC Limited are cutting-edge, index-based coverages, with parametric triggers tailored to reflect each country's specific rainfall requirements for growing staple crops

The calculation of claims to the program is based upon satellite rainfall data, which is used to objectively determine whether a drought has occurred. This allows claims to be calculated quickly so funds can be deployed far more quickly and efficiently. This is one of the first times in Africa that the insurance process has become such a key instrument in achieving humanitarian and development goals—but it’s just a taste of analytics’ role in the not-too-distant future.

Better analytics using satellite rainfall data can lead to more stable food supplies in Africa.

Better risk-transfer options for business

Equipped with improved data and analytics, companies are able to tailor their risk-transfer options more keenly, attract new kinds of capital, and better customize their risk-transfer and risk-financing strategies. From captives to financial-markets-backed catastrophe bonds for property exposures, analytics are empowering solutions that enable businesses to transfer traditionally uninsurable risks to a third party.

Precise data—both historical and exposure-related—is enabling a high level of risk transparency and opening new vehicles for mitigating and financing specific risks or portfolios of risks, including property exposures, business interruption consequences, project risks and in some cases liability risks

Beyond traditional insurance market options, the alternative risk transfer market offers options like the optimal use of captive insurers, or capital market-related solutions such as catastrophe bonds or contingent capital solutions.

The combination of diverse forms of data, from an individual entities’ own experience to broader industry data classes, provides a much more comprehensive view of risk. This allows for a look around the proverbial corner with company-specific, forward-looking loss scenarios. That gives entitles the chance to constantly adapt their risk management approach.

Healthier workforce

Today, health and claims data can help employers to predict future costs and possible adverse scenarios, enabling organisations to direct risk-control activities. Sophisticated analytical insights are providing companies greater visibility into their employees’ health, giving them a deeper understanding of the organization’s healthcare and workers comp costs. This knowledge enables companies to develop more tailored cost-reduction solutions—ergonomic or wellness programs specifically suited to their employees for example.

With so much of a company’s budget going to employee wellness, cost savings are a significant benefit for improved analytics, but they’re not the only one. The more sophisticated a company’s approach is to employee wellbeing, the healthier and more productive its workforce will be—and the more benevolent the company’s reputation will be. Which doesn’t hurt in retaining and attracting talent.

But what does this mean for the future?

The emergence of applied predictive analytics has inadvertently led to a split, in some areas, between people who think like a physicist, relying on the power of technical, algorithmic approaches, and those who rely on a more heuristic approach, trusting their experience and judgment. In the future, successful risk management professionals will need to synthesize these two approaches and mindsets into a form of thinking adept at combining human judgment and analytics to reach the end goal: greater understanding and better decision making.

Companies should adopt a healthy skepticism when first measuring data and looking for patterns, and here industry knowledge will be vital. Ultimately, getting the most from data will require the effective use of human capital. That’s where the future challenge lies – that’s where risk managers and Willis can work together.

About the author

John Merkovsky

Risk & Analytics

John Merkovsky is Group Director of Risk & Analytics.

  • Copyright © 2019 Willis Towers Watson