I want to talk about insurance reporting. I’m in New Orleans and I’ll be here for the next few days at the WSIA Insurtech conference. I thought, why not? Let’s talk about insurance reporting. I’m going to go over an unordered list. This is a brain dump of the types of reporting that I’ve seen be very useful for our insurance clients. They are in no particular order.
I want to start with exception reporting. This can be anything that can help your business identify out of range or unexpected scenarios. You might have data entry errors, for example, in your policy administration system. A field that might be important to be populated like state, for example, is often important in certain regulatory situations. You can create exception reports that call out missing data elements to make sure that someone is alerted to that situation and takes care of it.
Exception reporting includes out of range scenarios. You might have a type of an exception report that shows you when you have certain rate change situations that are unexpected. If the rate changes from A to B, and the difference between those rates is an unexpectedly large or small amount, you could always create an exception report to call out those outlier situations.
Insured obligation tracking in certain situations where the insureds are responsible for doing certain things. For example, in the trucking industry the insured must report any new vehicles that are added to the fleet that is insured. If they’re not doing this, I think it varies from state to state, the carrier is still responsible for those vehicles, even if they’re not reported. So, it’s important that the insureds are reporting the vehicles.
You might have an exception report that identifies vehicles that were inspected but aren’t on the policy as a list of scheduled vehicles. There are various ways you can create exception reports to help understand outliers, things that are out of range, things that are unexpected, and things that should be happening that aren’t happening.
Next is actuarial reporting. There can be quite a lot of different data elements involved in actuarial reports. One of them that’s common is a loss triangle. For example, let’s start with a Y-axis representing an accident year and an X-axis represents age. If you have an accident year, and in the first year following that accident year, we have $50,000 amount of losses. Then the second year, we have $60,000 total losses, and the third year we have $65,000 in total losses. You can see in a loss triangle how your losses mature over time. That’s very useful for a lot of reasons, probably more than I can get into. One could be making sure you have adequate rates to cover your expected loss ratios. By expected loss ratios, I mean expected loss ratios over the entire maturity of all claims for a given accident year.
Loss triangles are very important to actuarial processes. So are loss ratios, but before you can get to a loss ratio, you need to know your losses. Like we described in loss triangles, but for loss ratios, you’re comparing your losses to your earned premiums.
Earned premium is premium that covers a one-year period. Let’s say you have a $1,200 premium and the policy is at year one. When the first months is complete, you’ve earned 1/12th of the premium. So, you’ve got $100 in earned premium. Then month two, you’ve earned $200 total, $300 the third month, and so on until you earn the entire premium for the policy. If you take all your policies and cumulatively add up your earned premium, that’s money that a carrier has earned. You compare that to how much losses you have on those policies, you then come up with a ratio between your losses and your earned premium. Those ratios vary quite a bit from program to program as well as from line to line. When you have that number and you’re monitoring it, you can really see through time if you’re improving or getting worse in your loss ratios. This is a key metric for pretty much anyone in the insurance industry to be tracking.
If you’re a carrier, you’re responsible for the risk. It’s super important that you understand the loss ratio. Also, if you’re an MGA it’s important that you understand it because you’re working on behalf of the carrier. It’s important for MGUs as well. Frankly, if you’re a producer or a broker, it doesn’t really matter. We’re all trying to keep the losses as low as possible so that we can have a profitable business. Along those same lines, you also have things like exposure change. If the exposure changes from one policy year to the next, what’s the ramifications of that exposure change? Is the rate adequate to cover that change in exposure? That’s another way of providing data to actuarial inputs.
Another subject area we can assist with is in marketing. One of the common metrics I’ve seen used quite a bit is the hit ratio. It is basically out of all the applications you’ve received and quoted, how many of them get bound. So, if you’re receiving 100 applications, you’re quoting all 100 of them, and only 50 of them are getting bound, you have a 50% hit ratio. There are different ways of looking at that ratio, but overall, it’s just saying of all the opportunities to write a policy that we see, how many of those actually turn into a bound policy. And what’s that ratio?
Another area for marketing that might be useful is insured analysis. A marketing arm of the company wants to try to market to the insureds that they want to write policies for. What are those insureds? Which insureds result in the lowest loss ratios? It’s quite important that marketing is focused on the right target, and that will drive all the downstream processes. This will end up with bound policies that hopefully have profitable outcomes.
Another area for marketing is geographical analysis. You might look at your current book of business and look at a state view of things. For example, if you’re looking at Louisiana and find that you have an overweighted portion of your book that is residing in New Orleans, then maybe you don’t want to write as many policies in New Orleans. Especially if you’re talking about property coverage. Maybe you see that in the north side of Louisiana, you have very little exposure up there. Understanding geographically where your exposure is can be important, especially for certain lines like auto and property. So, geographical analysis can be helpful for marketing purposes.
Now you also have what I would call your “meat and potatoes” type reporting. You’re looking at premium trends, for example, what does your premiums look like? Do you have written premiums, gross written premiums, or earned premiums? What does that amount look like when you chart it? Maybe you’re looking at total premium by policy effective date. Maybe you’re looking at premium by another metric. You can look at transaction effective dates. You can take into consideration endorsements. In any case, the “meat and potatoes” of insurance reporting is the actual premium. That’s the revenue side of things. We need to have a good understanding of what our revenue trends look like.
Another area reporting is used for in insurance is claims. You need to have a good understanding of your claims. Claims can create interesting complications, because you have all sorts of ways of measuring. When a claim comes in, it’s initially not going to have anything paid out on it. Something is going to be reserved in anticipation of an obligation to pay out in the future. You have categories of claims: Some are reserves, some are actual paid amounts, some are recoveries, and so on. All add up to get to a number that we call our total insured. Different components of the claims measure across the maturity of a claim which we often call the valuation date. Claims are very similar to loss triangles. You will have a certain valuation date on month one and then a certain valuation amount on month two and so on. Those same exact claims will have different amounts as time moves forward. Claims and the understanding of the incurred losses for claims is quite important to understand. It contributes to the loss ratio, but claims are the main expense of an insurance business in most cases.
We talked about loss ratios. We talked about loss triangles. Those again, I think you could pretty much consider “meat and potatoes” of insurance reporting. Loss triangles don’t really have a lot of information when it comes to a loss triangle if you’re a new business. We’ve seen that happen at least a couple of times. You may need to have a loss triangle that’s based on industry benchmark type data. So you actually can use that as your reference until you have enough history in your business to create a valuable loss triangle.
Another area of reporting that we get into for our insurance clients is feeds for partners. This could be a lot of different things. It could be bordereau that an MGA might report back to their carriers. It could be coverage verification that a carrier is providing to a claims third party administrator or an MGA reporting to a TPA. It could be pretty much anything where you have a partner that needs a data feed that you’re going to provide to them. It could be either in the form of an interactive report, some type of standard schema, like a CSV or a fixed width file, or any type of standardized input that the partner can then consume and use that information for whatever purpose it was intended. Partner feeds can be a huge value add, especially if you consider doing them manually. You’re trying to pull data out of your various data sources, cobble it together and try to create a feed. You’re having to do this, if you’re lucky, every quarter, but often it’s every week if not every day. It’s important that carrier feeds are automated. That can be a really big value add.
While we’re talking about feeds, let’s talk about regulatory feeds. Certain lines have serious requirements to report to the regulatory agencies. Workers comp I know is one of them. Auto type policies also have certain regulatory filing requirements. With filings it is critical that those filings are accurate and timely. Otherwise obviously there’s going to be some pretty severe fines. Automating those processes and making sure that they are accurate and timely is something that we help our insurance clients with.
Another area of reporting that’s important is goal tracking. This depends on your perspective of things or what your role is in the insurance industry. If you think about a carrier, a carrier is probably interested in understanding the size of the book of business. They want to know how much premium is out there in a particular book. They want to track against goals. If they have a goal to grow their book by 20% by the end of the year, they are going to do that through whatever organic policy sales, as well as acquisitions of other companies. They want to see if they are making it to that goal. There are all kinds of ways to slice and dice goals.
Another example would be program specific goals. If you have a program that maybe is underperforming, and you’re trying to correct that problem. I don’t know exactly what the problem would be, but let’s just say it’s high losses. It might be important that we track the loss ratios for that program very closely and compare them against the goal so that we can see if it is improving or is it getting worse? Is this simply a line to get out of because the losses are too high? Tracking against goals can be important, especially for the executive and the managerial team to keep their fingers on the pulse of what’s going on.
Another area that’s a bit technical, but also important is reconciliation automation. I’m working on this for one of our clients right now. There are often cases where we’re using multiple data sources that have the same information in them. You might have one policy administration system where policies are being entered originally, and you might have another system that’s doing the billing.
There’s all kinds of scenarios like this, but in these cases where you have duplicate entry for whatever reason. It’s important that these systems are in sync. Problems can happen. Humans make mistakes when typing. Create a reconciliation process that is automated to the point where a human can then do the final steps. This means we can align data elements and figure out what’s already synchronized properly, where we have matches, and then call out those things that don’t match. Then we can look at those items and figure out if there is a data anomaly. Is there a failure of the process? What’s the issue? Why are we not reconciling? Then we can address the issue. Reconciliation automation can be a big time-saver.
Another area that’s kind of interesting is program manager scorecards. If you have a one or several programs, you’re going to want to have a scorecard to help the manager understand how things are going. Is it being successful or is it not being successful? Are we on track? That scorecard might have a lot of the same things we already talked about, such as hit ratios, rate change, premium trends, and loss ratios. The idea is you compile a lot of information, focus on a particular person’s role or a particular group’s role, in this case a program manager, and give them a high-level view of things. In addition to that high level view, you should be able to drill into the various aspects of that scorecard and understand the components that make up key performance indicators.
For example, if my hit ratio is 50%, let me drill into that so I can see the applications that we bound. Show me also the ones that we did not bind so that we understand, are we being too stringent on the selection of our applications that we bind or vice versa. You can make those decisions with that information.
Another area that’s big is simply the elimination of manual data manipulation. You know, whenever your business gets to a certain size and you’ve got various data feeds going on, whatever it might be, you are quickly going to get to an unsustainable situation. If you’re an MGA, you’re reporting back to a carrier, you’re reporting to regulatory, and you’re reporting to partners. Once you have this stuff going on, you are quickly going to get to an unsustainable situation if you’re doing these things manually every single time. Not only will it be very expensive to have your costs grow linearly as your additional data feeds and reporting requirements increase, but the process is going to be very slow. It is going to be error prone. It is not going to be sustainable. There’s not a lot of people in the world that want to do this work that is repetitive day after day, cobbling data together, and trying to make it accurate. You could be in Excel and whatever other tools you can cobble together so that you get this data in the shape that it needs to be.Eliminating manual processes, isn’t real exciting, but it’s often a very big part of what we do. It’s not trivial either. A lot of times these manual processes require a considerable amount of human input. Some things are not able to be automated, but it’s very rare that nothing can be automated. Often there’s a big chunk that can be automated before it gets to the need for a human to evaluate it. The reconciliation automation and a lot of the upfront work can be done right away. Let someone that actually has to do the final mismatch analysis have the information at their fingertips instead of having to go pull it all together manually.
A lot of the other stuff can be completely automated, like regulatory reporting, the creation of board rows, or the “meat and potatoes” types of things like premium trends and earned premium through time. A lot of that can be completely automated. However, it is important that we create some type of check and balance. Whenever you do not have individuals involved in a process and you’ve automated something, then it’s even more important that you have some way of auditing your own automation. Regression testing or some type of ongoing checks that are in place to make sure the automation stays in sync.
And finally, I want to talk about the types of data that we often consume. When we talk about all these reporting requirements, data feeds, automation of manual processes, and exception reporting, it all relies on certain data sources being available. The types of data sources that we most commonly work with when it comes to insurance reporting, building insurance reportopia, is policy administration systems. This is where the policies are originating. Oftentimes rating is done within these systems and the modifiers are used to create the coverages. The core root level premiums and all the characteristics of those policies are stored in these systems. So, the policy admin system is often super important in what we do. Not always, but most of the time it is the policy admin system that’s the center of this data source universe.
Now, from there, it really varies quite a bit. If you’re doing your claims processing within your policy administration system, that’s probably where you’re getting your claims from. If you’re using a third party to do your claims administration, they’re likely sending you a feed. The feed can come in different forms they could be a flat file, an Excel file, or a connection directly into their data source. When it comes to loss runs there’s another category of feeds.
Often goals, when we start out, are in someone’s head. They might be written down in Excel. They might be in various places, but goals are also something that we need to consume so we can determine if we’re tracking against these goals. Goal can come from different types of data sources; it really depends on each company. It could be an Excel file, or it could be something very sophisticated, like a reference data management system.
One type of standard feed that comes to mind right now is the workers comp standard. We pick those types of feeds up as well. There are all sorts of standards out there, of course, but these standards when it comes to partners really help us. When you have a standard in place, we don’t have to come up with a spec that defines exactly what we’re going to be providing our partners or what they’re providing us. Partner feeds, especially the standardized versions, are also a different type of data source that we often get into.
Then you have public data. We talked a minute ago about the need to keep track of insureds’ obligations. The example was reporting trucks that are added to a fleet in the case of commercial auto. In that case, you’ve got DoT data that’s being aggregated to show inspection information. That type of information is also important to have available to you. And it’s not just auto, there’s all sorts of public information that can be useful all the way down to the physical environment. If you’re talking about property and you’re insuring property in an inclement weather zone, something as simple as weather information could be really important in understanding the risk associated with a particular exposure.
This has been a brain dump of various types of reporting that we’ve helped our insurance clients. Nearly every client we start working with, we learn something. There’s just so many ways of looking at this. Ultimately what we try to do here at LeapFrogBI is help our clients build their reportopia.
When we talk about reportopia, you’ll often hear me say it’s about capturing value. It’s about delivering reporting so that it improves a particular business process. It’s about assisting business owners, process managers operate more effectively, efficiently. All those things are true, but at the end of the day what we’re all trying to do is build a successful business. And we must have the information required to make decent decisions in order to have any chance of success. I’ve said this before, but I’m going to say it again. If you believe that the cumulative decision-making ability of everyone in your organization is going to determine your organization’s success, if you believe all our decisions and the actions that follow that are going to determine our success, that’s pretty much saying the things that we do are what’s going to determine if we’re successful or not.
The only alternative to that is saying that we all have no control of our destiny. Everything is predefined. If you’re thinking along that way, like you have no control over things, then yeah maybe this is not for you. Most of us are trying to find a way to understand the environment that we’re operating in. We’re trying to improve the decisions that we’re making and ultimately the actions that we’re taking so that we can be competitive. So, we can grow our business. So, we can deliver a better product to our customers. That’s what reportopia is about. Regardless, if we’re talking about insurance, healthcare, or any other industry, it’s just about that. It’s about growing the business, delivering something that a customer wants. Something that someone values.