Warranty Risk Management 101: A Four-Part Series
Introduction
Developing a successful Warranty Program (whether limited or extended) is no easy task. There are a number of decision factors to consider, including but not limited to: extended warranty vs. service contract, type of coverage (break-fix, replacement, accidental damage, etc.), limits of liability (what to cover), deductible amounts, add-on vs. Day One, where to offer the plans (states and countries), how to offer (POS integration, website, alternative channels), regulations / filing requirements, claims administration (in-house vs. third party), pricing and risk management.
Risk management is the least understood, yet most impactful aspect of Program Design. Accurate Cost Per Unit (CPU) forecasts are the very foundation of stable prices, right-sized reserves, higher profits, and higher customer satisfaction/loyalty. Yet, the issue with these calculations is that – nine times out of ten – we find significant errors. With virtually no exceptions, these calculations are generated using antiquated quantitative methods that do not work well in sectors where the historical perspective is rendered less useful due to changes in technology, product mix, consumer behavior, etc. Hint: if your analysts are relying on loss triangles, chain ladder, augmented chain ladder, Weibull, Cox regression, and similar methods – you almost certainly have room for improved CPU forecasts.
After, Inc. has developed this four-part series on “Warranty Risk Management 101” to give you a basic understanding of this complex discipline. In Part 1 we begin by defining the key components of any Risk Management exercise. In Part 2, we will introduce the types of risk structures and the pros/cons of each. In Part 3, we will show you how to accurately estimate losses and illustrate how errors can impact program profits. And in Part 4, we will conclude with a case study on Polaris, a long-time After, Inc. client, who we helped with a risk roadmap that led from a completely outsourced risk management structure to creating its own warranty company over the course of nine years in iterative steps – and the lessons we all learned along the way.
Now, let’s start with the definitions.
Warranty Risk Management Definitions
Below we have listed thirteen key components of Risk Management and provided definitions of each. You can use this table as a cheat-sheet to refer back to throughout the Series, as well as download for your own files: Risk Management Definitions – Feb 2021.
| Component | Definition |
|---|---|
| Premium | A premium is the amount of money charged by an insurance company to keep warranty coverage active. It is typically paid monthly but can be billed a number of ways. The insurance company calculates the premium using this equation: Premium = CPU + SM + RF + PT. Each of these four components are described below. |
| Cost per Unit (CPU) | CPU is the core calculation in risk management and is defined as “the average dollar amount needed to cover claim costs, including parts and labor, and in some cases, claims administration.” As an example, let’s assume the CPU = $100. |
| Safety Margin (SM) | Safety margin is the amount charged by the insurance company to cover “unforeseen changes” and is added to CPU to come up with the premium. If the CPU is $100, and the safety margin is 15%, the amount of safety margin would be $17.65. The equation is not “$100 X15%”, but instead “($100 divided by .85) – $100 = $17.65.” |
| Risk Fee (RF) | Risk fee is the “base wage” for the insurance carrier. It is used to cover carrier resources and deployment of capital/surplus and is added to CPU and SM to come up with the premium. At a $100 CPU, a 10% risk fee would add $13.07 or “($100 divided by .9) – $100 = $13.07.” |
| Premium Tax | The premium tax is the tax imposed by each state on the carrier in relation to the gross premium allocated to risks located in that state. The insurance carrier passes this on to the consumer or manufacturer. For example, a 2% premium tax would add $2.67 to the CPU or “($100 divided by .9733) – $100.” Now that we have the four component costs, we would just add them to get the premium. $100 + $17.65 +$13.07 + $2.67 = $133.39. |
| Investment Income | Insurance companies have two sources of income: underwriting profits and investment income. The latter is generated through investing premiums into interest-bearing assets like stocks and bonds. Dividends and interest are then used to offset underwriting operations which can sometimes be unprofitable. The longer the term of the contracts sold, the more investment income there is. |
| Underwriting Profits | Underwriting profits are calculated as “revenues that come from cash collected on policy premiums minus money paid out on claims and for operating the business.” In forecasting claims costs, insurance companies will always be conservative to protect their margins. Manufacturers should recognize this and do their own analysis to make sure that they are not paying higher premiums than they should. |
| Loss Fund | The amount left over to pay claims after risk fee and taxes are taken out. It is typically equal to CPU + Safety Margin (and sometimes includes investment income). |
| Profit Share | Percent of the underwriting profits that a manufacturer receives, based on terms of the agreement with the insurer. It can include all or a portion of the underwriting profits (usually 50-100%, depending on the size of the program, the known risk, etc.). |
| Earned Premium and Premium Earning Patterns | The term earned premium refers to the premium collected by an insurance company for the portion of a policy that has expired. Earnings curves provide a company with the percentage of premium that should be recognized as revenue at each point throughout the life of the service contract in order to appropriately match revenue with expected claim costs. Approaches commonly used are Pro Rata, Rule of 78s and Reverse Rule of 78s. The Pro Rata approach earns premium evenly over the life of a contract, which assumes that losses are expected to occur evenly throughout the term of the contract which rarely happens. The Rule of 78s method is commonly used when losses are expected to be weighted toward the beginning of a contract such as with Used Vehicles. The Reverse Rule of 78s is commonly used on New Vehicles where the exposure is weighted toward the end of the contract after the manufacturer’s warranty has expired. In general, these benchmark curves provide the benefit of being easy to calculate and explain but rarely reflect an accurate expectation of claim costs throughout the term. (Source: https://www.providers-administrators.com/348218/earnings-curves-matching-premium-with-losses-and-refunds) |
| Claim Emergence Patterns | The claim cost curve shows the percentage of claims that occur each month. |
| IBNR | The acronym IBNR stands for claims “incurred but not reported”. In risk management, it refers to claims that have happened, but the insurer is still unaware of them. As these claims must still be paid out, the insurer must set aside money (“reserves”) to cover its soon-to-be-discovered costs. Since the insurer knows neither how many of these losses have occurred, nor the severity of each loss, IBNR calculations are estimates that can be subject to a range of errors. |
| Loss Ratio | The loss ratio is defined as “incurred claims divided by earned premium.” The issue is that major errors occur as a result of earned premium miscalculations. A better measurement would be “total forecasted claims / total premium” as this would take into consideration the volatility that can occur throughout the contract term. |
To learn more about After, Inc.’s Warranty Analytics & Risk Management Consulting Services, please visit http://afterinc.com/warranty-analytics-solutions/