Getting under the skin of exposure – a case study in pharma

July 25, 2018 Chris Sandilands

We recently spoke to Libbe Englander, the Founder and CEO of Pharm3r. We found the proposition interesting as an example of how commercial / specialty insurers can improve their underwriting decision-making from the better use of data – right now.

What is Pharm3r?

Pharm3r is a healthcare analytics company. Their flagship product is a dashboard with risk insights on drug and medical device manufacturers. For example, the business can show you the number of adverse events reported for a drug and can compare product problems associated with thousands of medical devices. Underwriters using the product can, for example, assess the relative riskiness of two pharma manufacturers’ portfolios of drugs or gain deep insight into a medical device manufacturer’s exposure profile:

• What specific devices does the company make and which of them exhibit high-risk features?
• What adverse events and product problems are being reported? Which risk signals are
  unexpected based on documentation or class peers?
• What is the trend of signals associated with the company? How does the risk profile compare to
  other similar devices on the market?
• In what countries are events/problems being reported? How many of these are associated with
  litigation activity?

Portfolio analysis is also possible, for example what is the insurer’s aggregate exposure to high risk
companies, drugs or active ingredients. Given that data is uploaded regularly (up to hourly) from various sources, trends can be identified quickly.

Experience to exposure

Commercial and specialty insurers sometimes argue that their segment of the market is relatively immune to change. For example, high layer D&O underwriters might argue that there are too few claims for it to be possible to model risk and continue to price largely on gut instinct supported by some analysis of historic claims.

But industry databases and analytics companies are a game-changer for underwriters. In short, they allow underwriters to move away from using experience metrics (i.e. claims and near misses) as the primary source of rating to an informed and quantified view of exposure.

Let’s consider an underwriter considering a $100m xs $200m layer of a medical device manufacturer’s liability policy. It is likely that there are too few historic claims for the underwriter to perform any reliable statistical experience analysis of that layer’s exposure. The broker may be pushing for below-average ROL pricing due to manufacturer’s good claims history.

Data could offer underwriters huge insight into the exposure of the layer. For example, there might be a sharp increase in the number of lawyers who have recently filed claims related to a specific medical device in the US. For a drug manufacturer, Pharm3r could identify that a particular active ingredient appeared to be linked to an uptick in side-effects.

A clear route to value – unlike IoT

We said in the introduction that underwriters could use insight from businesses like pharm3r “right now”, and we mean it. We see a clear and unimpeded route to value that requires little more than underwriters’ willingness to explore new data sources.

This is in contrast to a more discussed industry theme, (commercial) Internet of Things. IoT involves insurers collecting data from devices in or on an insured’s property and using this for a variety of purposes, for example loss prediction / prevention, more accurate pricing thanks to a better view of exposure, or more tailored products.

But nobody has cracked IoT yet. There are complications: what is the ‘trade’ that means the insured is happy to share their data with the insurer; economic questions about the cost of devices; difficulty linking sensor data to a risk ‘score’ or insurance premium.

In terms of a framework we recently discussed on our blog, IoT needs the innovation team to find the ‘product-market fit’, whilst industry databases are ready for the rocket fuel on the front line.

What this means for underwriters

The existence of such databases changes the role of the underwriter and capabilities required. First of all, we believe that all underwriters should be involved in hunting for data to help them assess risk more accurately. (Indeed, Munich Re now has a team called ‘data hunting’.) In some cases this data will provide insurers with underwriting information to quantify a premium; in other cases the data will provide only a relative risk score. Both are superior to pure underwriter judgement.

This hunt will need creative thinking. Whilst some mappings might be obvious – for example Pharm3r and medical liability – others might not be (e.g. what is a lead indicator for general liability exposure?).

Second, underwriters will need to be more structured and quantitative in their analysis of exposures. In some ways property cat is a good proxy. Here, underwriters already use sophisticated modelling software (e.g. RMS, Oasis LMF) to quantify exposure. Direct insurers will require similar skills – a technical ability to calibrate and run sophisticated analysis looking at both historic performance and projections, an ability to interpret results, not to mention the usual relationship and negotiating skills.

The bottom line: underwriters should be having lunch with data providers both inside and outside the insurance industry as much as with brokers.