While some insurance products are highly specialised and almost impossible to compare, most common products have increasingly become commodities. Consumers can now easily compare them online.
When comparing insurance policy prices and details becomes as effortless as getting quotes for airline tickets or hotel accommodation on price comparison sites, more insurance companies will eventually enter the market. And thus price competition will increase.
Preparing for a price war
Once the price war starts, there is no way to avoid it. And insurers need to meet their competitors head-on.
To win a price war, insurers need to be meticulous when they set the premium levels. They might also need to rethink the definition of “profit” when they are making pricing strategies for the future. In a market where premium levels are volatile and the competitive situation may change rapidly, insurers also need the capability to evaluate potential future scenarios in a short period of time.
Setting the premiums right
In the fast-paced digital era, customers expect insurance prices to be easily available online. They will make inquiries for insurance covers for their cars or homes on price comparison websites and expect the prices to be available immediately. From an insurer’s point of view, the premium customers will see on their screens when comparing insurance policy prices is the sum of the insurer’s technical premium and the commercial loading.
The technical premium represents the break-even price that the insurance company would charge for the policy if it had no costs and no desire to make a profit. Commercial loading represents the sum of the insurance company’s costs and the profit it expects to make on the policy. Technical pricing is the subject of many actuarial textbooks. But as machine learning algorithms make their way into actuarial departments, we will need to rewrite those books. Modern pricing techniques that include machine learning algorithms are a notable improvement compared to traditional models. If applied properly, ML models will result in more accurate technical pricing given the same data.
But what about commercial loading? How much profit should the insurer aim for?
Every one of us has a different tolerance for how much we would pay for, e.g., a car insurance policy. Some customers don’t consider price to that important. Others will try to search for a better deal elsewhere, regardless of how much time the process would take. Most customers are somewhere in between.
Being able to price the insurance products analytically based on the “willingness to pay” is, for many actuaries, seen as the holy grail of insurance pricing.
Most insurers already do personal pricing to some extent today. For example, they give different discounts to policyholders with equal risk. However, there is often a great potential to do segmentation and price calculations in a more analytical manner. Ideally, insurers would like to set the premiums as high as possible, but not so high that customers move their policies to another insurer.
On the other side, insurers would like to move customers away from their competitors by offering low premiums – but not too low. The insurer must first determine the price sensitivity of insurance customers and then price each insurance policy so that it maximises the profit for the insurer. At SAS, we refer to this as portfolio optimisation.
Insurers that can quickly reoptimise changing prices in the online market will also quickly identify customers that are at risk for churn. They can then perform the appropriate actions to prevent this from happening.
When insurers think “profit,” they usually mean the income statement for next year. This is about to change. The concept of Customer Lifetime Value (CLV) is becoming more and more common in the insurance industry. And many insurers are now refining their pricing strategy based on a maximisation of the CLV of all its customers, thus not focusing solely on the profit definition in the income statement. The CLV of an insurance customer is the net present value of this customer for the insurer, where behavioural effects like renewal, cancellation and cross-selling of other insurance products are considered for the entire lifetime of the customer.
To accurately compute CLV for a customer, the insurer will need data that describes the behavioural patterns of the customer. Most insurance companies have quite a lot of such data available – the problem is usually that it is not adequately structured. In practice, to quantitively identify the customer lifetime value, insurers need to integrate both actuarial and customer behaviour models. Once a system for this is in place, insurance companies will have a strong quantitative foundation to compute the customer lifetime value of their policyholders.
SAS and insurance pricing
Price competition is changing the insurance market right now. When a customer determines where to buy insurance, the price is the most important factor. Thus, to stay competitive and still run a profitable business, insurers need to set their premium levels just right. The evolution of price comparison websites – which provide real-time quotes on competitor prices and increased access to data that contains information about the customer’s insurance risk – has made the actuary’s job of calculating the premium more complicated.
Over the years, SAS has worked together with insurers to ensure that strong system support is in place to compute premium levels down to an individual policy level. These pricing systems have been put through the test in some of the most competitive insurance markets in Europe. They have turned out to be a valuable strategic tool for insurers to balance the desire for profit against the desire for market share. And maybe most important of all, they have enabled these insurance companies to effectively join the price war, fight it and still make a profit.
Source: Global Banking and Finance
Share this article: