Like any business, companies need to sustain higher revenue than expenses in order to stay viable. Auto insurance is no different; companies make money off of the premiums customers pay but lose money when they fulfill their obligation to pay for their customer's damages. They also have a host of operating expenses to pay, including agent compensation and advertising. The proportion of expenses to revenue is called the "combined loss ratio," and whenever it is above 100%, the company loses more money than it is earning. In 2016, only two of the top 10 auto insurance companies in the country had combined ratios below 100%--and just barely.
So even if you have never been in an accident, your rates are still going up because the auto insurers are trying to bring their ratios back below 100%. Imagine a major drought destroying part of a farm's crops, and then the farm charging more for the crops that survived in order to make up for the ones they lost--it is the same principle with your auto insurer. In 2010, the situation was the exact opposite. Only two of the top 10 companies were operating with combined loss ratios over 100%. The average combined loss ratio was 99.7% in 2010 compared to the whopping 107.1% average in 2016. The average combined ratio has climbed year after year, and as a result, auto insurance rates have gone up an average 20% across the country.
Unfortunately, the rates hikes have not been effective at closing the gap between profit and loss. All the insurers in the table increased their written premium revenues in 2016 (mostly due to rate hikes), yet they still (with the exception of Allstate) ran higher combined loss ratios. The current trend indicates that the companies are getting further from turning underwriting profits again. Even after three solid years of increases, the companies are not just still losing money - they're losing an even greater amount.
In their financial statements, GEICO, Progressive, and Allstate are not hesitant to blame bad weather as a significant source for their losses. In its financial statement, Progressive said catastrophe losses as of the end of the third quarter were "$121.0 million greater than in the same period last year". They attribute 85 of that 121 million to Hurricane Matthew alone. Comprehensive claims, the ones that would result from catastrophic weather, can average upwards of $1,700 per claim according to the Insurance Information Institute, so those figures make sense given the number of people affected by a hurricane. The floods in Louisiana, which were also explicitly cited in financial statements, also ended up costing insurers millions. It's not just the weather though.
The National Safety Council found that fatal motor accidents went up 6% from 2015 to 2016, for a total of 40,200 fatalities--the most since 2007. The National Highway Traffic and Safety Administration blames distracted driving due to texting as a large source for the increase in fatalities.
More disasters and more accidents are leading to more claims, thus more payouts from the insurers. In 2017, the number of households with at least one auto insurance claim in the past three years increased by 3,869,969 compared to the same statistic in 2014. In 2017, 22.2% of households had at least one auto insurance claim, while 20.5% had one in 2014. Nielsen, which compiled the data, projects that by 2022, 22.5% of households will have at least one auto claim.
It is difficult to pinpoint future auto insurance pricing with certainty. What you will pay for auto insurance in the near future will most likely be more than what you are paying now--even if you have a great driving record. The trends that are causing more accidents--lower gas prices leading to more drivers, those drivers being distracted with texting--are not likely to abate. Other important factors like severe weather are difficult to predict. The Colorado State University Tropical Meteorology Project has forecasted this hurricane season to be below average in terms of the number of 'named storms' that will form in the ocean. They cannot predict however how many of those storms will hit the U.S. So even if the hurricane season is below average in terms of the number of storms if the number of storms that hit the U.S. is above average, the property damage costs would be enormous and an even heavier burden on the insurers.
On the other hand, if the weather turns out to be favorable, or something discourages people from driving, the insures may start to see better margins again, and feel no need to raise rates further. Again, it is all speculative and only time will tell.