Twenty Trends. Brian P. Sullivan, Risk Information Inc.

Transcription

1 Twenty Trends Brian P. Sullivan, Risk Information Inc.

2 1. Autonomous Cars Are Years Decades Away Autonomous cars will profoundly reduce the size of auto insurance, so planning for an existential change is critical We re sticking to our prediction that we re five years away from a fully autonomous car (For once we re bearish compared to others) If it takes five years from the first truly autonomous cars, then we re 25 years from an autonomous fleet Our discussions of the complexity of mixed fleet are as relevant as ever (see 2016 Mike Nelson; 11/13/17 Auto Insurance Report)

3 2. Machine Learning Could Be The Autonomous Car Wildcard As I mentioned in last year s trends, when trying to predict the future it is essential to identify factors that could make you wrong When it comes to autonomous driving (and many other things), the advances in machine learning present a true wild card Is it possible that machines teaching machines could accelerate autonomous car decision making and bring a viable car to the market sooner than 5 years? Yes, it is possible But remember, it still takes 15 years to turn over the fleet

4 3. Essential to Plan For ADAS Claims Reduction: 10 Years? As mentioned yesterday, I have lots of confidence around projecting the end game, and much less on the ADAS tipping point But projecting is essential given how important this is So here s my idea: today s cars have only a modest impact on reducing claims, so 5 years seems much too short. We ll have plenty impactful cars for sale in 5 years, but not nearly enough in the fleet So 10 years feels like a point where enough impactful cars will produce a measurable decrease in claims counts

5 4. V2V and V2I Further Away Than Ever In honor of retiring Kim Hazelbaker of HLDI, I revisited the vehicle-to-vehicle and vehicle-to-infrastructure question These powerful safety tools continue to see development, but the lack of a profit motive holds back implementation We re going to need more connected autonomous cars before the benefits are sufficient to justify this infrastructure spend

6 5. Going Out On a Limb: Allstate Ready to Grow Allstate, disappointing for years, bounced to #3 by Progressive Recent retirement of President Matt Winter, 60, hurts Industry grew 8.0%; Allstate brand grew 3.5%; Esurance grew 1.0%; and Encompass declined 8.7%. Ugly, right? So here s my hope: Allstate has done a better job than most at getting prices right, and is now strong enough attack Stock is strong, providing some wiggle room

7 6. Not Much Exchange Love; Here s One More Reason Here s a late addition: lots of pushback on my Exchange confidence yesterday, with insurers reluctant to cede control So, rather than backtrack, let me double down: using more sophisticated telematics data for pricing will eventually expose it to FCRA-level regulatory scrutiny. Exchanges better equipped than individual insurers to present regulatory infrastructure In case you re wondering, yes, the exchange developers love me The whole things falls apart if they can t deliver data normalization

8 7. State Politics Could Make Trouble For Insurers Insurer friendly (i.e. Republican) domination of state legislatures has not delivered a legislative windfall, but has prevented trouble But a double whammy is coming: a shift toward plaintiff bar-friendly legislatures (i.e., Democrats) whose agenda will be fueled by rising auto insurance prices We re out on a limb here, but our state reporting is starting to see action, and if the elections go a certain way, watch out

9 8. While Waiting For Millennials, Aging of Boomers Cuts Premium We introduced this point a year ago, but additional work confirms As Baby Boomers age, they will buy less car insurance Millennials will happily take their place, but there will be a lag Math says there will be a dip in amount of household insurance Caveat: societal change doesn t impact insurance, and other factors will have much greater weight to the top line But the math is pretty strong...

10 Source: Courtesy of Peter Francese, Francese LLC, based on data from Bureau of Labor Statistics

11 Source: Courtesy of Peter Francese, Francese LLC, based on data from Bureau of Labor Statistics

12 Seniors v industry growth rate 5% v 2% 13% v 4% 5% v 5% 4% v 5% 21% v 6%

13 Personal Auto Will Be Better Than 2017 As discussed yesterday, miles driven is moderating, claims are stabilizing, and insurers continue to take rate Perhaps a giant catastrophe could slow things down, but even that would not likely overcome the confluence of positive factors Reserve redundancy (not everyone agrees by the way) helps too Still some problems states, but industry trends are positive Severity will continue to grow, gains come from frequency and rate

14 10. Back Out On a Limb: 2018 Will See Commercial Auto Turn May not show up in full year results, but there are signs that enough insurers have taken enough rate to effect a change by YE 2018 I m also partial to comparing paid to incurred loss ratios. Consider:

15 11. Commercial Auto Profit Spread Will Widen Those who focus on commercial, and set proper prices, are finally willing to surrender business to those who charge too little Though everyone agrees that commercial is troubled, there is still plenty of evidence that accommodation business will remain cheap Look for leaders to lose share but gain profit, pushing up the losses for the laggards (this can happen within a single company!) Eventually, maybe later this year, pain will be too great and the market will finally find some stability

16 12. Big States Remain A Big Worry In 2017 California improved to a 70.1% loss ratio. The only higher ratio in more than 20 years was 72.9% in Not good Texas is a similar story, with the 2017 loss ratio of 76.6% showing improvement over 79.8% in Florida is better, but not good: 68.7% in 2017 after 71.2% in 2016 New York: 69.2% in 2017; 69.0% in 2016 US was 68.8%, thanks to smaller states

17 13. Photo Estimating Will Excel At Internal Damage Only the best adjusters can see damage below the surface Today s photo estimating is no match for those nuances, and may never catch the best adjusters But the best adjusters are not the bar that s the average adjuster, and that s a challenge photo estimating will meet Essentially, deep data sets on structure of cars will enable photo estimating to do a better job at detecting initially unseen damage

18 14. Please Prove Me Wrong: Insurers Won t Help Salvage We know exactly how salvage can be cleared, processed and sold more smoothly and profitably for everyone after a catastrophe: Get in front of challenges of towing and titling Consolidation into IAA and Copart make this possible Local towing industry will balk, but otherwise few problems Alas, insurers don t show enough passion to push for change Can we get just a tiny bit of senior executive backing please?

19 15. Agency Mergers Setting Records, But Startups Continue There has been a long trend toward fewer insurance agencies Last year set a record for agency M&A as investors brought money to the table, and older owners looked for an exit strategy Not the death of agents: the consolidating agencies are stronger and capable of more And though data is hard to come by, we re seeing a healthy number of new agencies attacking opportunities

20 16. More Insurers Explore Less-Customized Infrastructure Why is it hard to install core software systems? Because insurers demand too much customization There is a difference between maximizing the fit and just replicating existing processes Watch for a trend toward accepting third party tools Even more certain: shift toward third-party core systems shows no sign of slowing down

21 17. USAA Won t Be Fully Fixed in 2018 Hard to believe USAA s 2017 personal auto loss ratio of 79.6% is a big improvement, but it beats 86.3% in 2016 USAA can run higher loss ratios thanks to low expenses, but the gap isn t nearly that large Geico, with similar expense, was 10 points lower (and unhappy!) USAA taking rate, but problems run deeper. All indications show that the right steps are being taken, but this is a fix that will take more than a few months to take hold

22 18. Shifting From FNOL to INOL Not Yet Ready The ability of vehicles and smartphone software to make an Instant Notice of Loss is developing rapidly But there remain significant technical challenges to handling this on a large scale, and consumers have not yet bought in The technology roadmap is clear, however, and it feels inevitable that a significant amount of claims will move this way The timing is uncertain so we re not putting a prediction to this one, but it feels increasingly ready for prime time

23 19. Comparison Shopping Still in Gradual Growth Mode It would be more fun to predict a big bang, but 2018 seem certain to be a slow building time for online comparison shopping The agency models require real-world growth work, and even the pure comparison sites have significant development challenges when connecting to insurers The big brands will stay on the sidelines because they re not yet pressured enough to force a change

24 20. Despite Rate Increases, Shopping Rates Hold Steady Normally, rising prices would drive an increase in shopping, and certainly that will show up in certain segments But cultural/political noise is drowning out marketing message Stronger economy has eased need for an extra $50 Shift in advertising spend from big splash to more refined and targeted online sales will take a while to work out Insurers doing a better job of reinforcing relationships

25 See You Next Year! May 5-7, 2018 The Breakers Palm Beach, Florida