Monitoring Use

For property and casualty underwriters, there is probably no line of coverage as badly in need of automation as private passenger auto. Tracking how insured vehicles are actually being used – and detecting high-risk behaviour on the part of policyholders – is critical if insurers want to price according to actual risk.

This type of detection is starting to take shape, as carriers and brokers offer telematics to their policyholders. But at CW Consulting’s recent Property & Casualty Crystal Ball, two speakers who are involved in telematics trials spoke of limitations. Alister Campbell, chief executive officer of The Guarantee Company of North America, suggested he would be penalized for exceeding the posted speed limit of 100 kilometres per hour on Highway 401 in southern Ontario, while Randy Carroll, CEO of the Insurance Brokers Association of Ontario, indicated that in the trial in which he is participating, he has the highest frequency and severity of braking.

Campbell did not say exactly how quickly he drives, but if he is driving 112 km/h on the 401 in optimum driving conditions, he is actually driving what was the posted maximum speed in the mid-1970s. Back then, the 401 had a maximum of 70 miles per hour, reflecting the speed for which the highway was designed.

As for Carroll, it is not clear why he is braking so much. He might be overly risk-averse, but in any line of insurance this should not normally warrant a rate increase. Alternatively, Carroll might be watching what’s in front of his vehicle and actually preventing collisions.

Either way, by flagging excessive braking as a collision risk, a telematics program is being used not to detect risky behaviour. Instead, it is being used as a tool to presume – correctly or incorrectly – that risky behaviour was undertaken prior to braking. Would it not be nice if underwriters could use technology to avoid making presumptions in the first place?

Granted, telematics is a fairer (and probably more accurate) method of determining risk than making presumptions based on demographic data, especially gender and marital status. Certainly, if telematics devices are telling carriers how far the insured vehicles are going, how fast – and if they are swerving, braking hard or drifting out of their lanes -those are probably much better indicators of risk than the policyholder’s gender, marital status or credit score (which most Ontario auto carriers would likely use if it were not prohibited by law).

But in order to have true underwriting in auto, there are two critical things the insurer needs to know in order to assess actual risk. One is whether the driver is adjusting to conditions. If insurers could use instrumentation and computer programs to detect the road surface conditions (whether bare and dry, wet, snow-covered or icy), traffic volume and visibility – and then cross-reference this information with the behaviour of the vehicle (including speed and whether cruise-control is used inappropriately) – this would take insurance automation to a new level. The second important factor is distraction. If video cameras were installed in cars – giving the underwriter the ability to record a visual of the driver at all times – there should be no question of whether a driver was looking straight ahead.

Sure, this type of monitoring would have the privacy advocates screaming blue murder, as if driving on a public road is an entitlement rather than a privilege. Nonetheless, it would give underwriters a good indicator of behaviour that is becoming a leading cause of collisions. Knowing how an automobile is being used and operated should be of some use to the company underwriting the risk associated with the vehicle’s use and operation.