I recently sat with Nick Gerhart to discuss the regulatory environment for U.S. insurance carriers. Nick offers a broad perspective on regulation based on his experience: after roles at two different carriers, Nick served as Iowa insurance commissioner and currently is chief administrative officer at Farm Bureau Financial Services.
Nick is recognized as a thought leader for innovation and is regularly called on to speak and moderate at insurtech conferences and events. During our discussion, Nick described the foundation for the state-based regulatory environment, the advantages and challenges of decentralized oversight and how the system is adapting in light of innovation.
This is the last installment of a three-part series. The first focused on the regulatory framework insurers face (link
). In the second part (link
), Nick provided the regulator’s perspective, with a focus on the goals and tactics of the commissioner’s office. Here we discuss the best practices of the insurers in compliance reporting as well as future trends in compliance reporting.
From my experience in speaking with carriers, I’ve been struck by the challenges of reporting data in various different reports to so many different entities. A lot of carriers struggle just with the process, and the quality of the data reported suffers. So, to dive into the quality of the filings for a moment, what are you looking for?
Garbage in, garbage out, obviously.
The most obvious issues start with the outliers. And it would come back to the state catching the company filing some bad data. So, for instance, on the life and annuity side, how you define “replacement” can trigger a percentage up or down that maybe you shouldn’t have in there.
If you think about it, from the company side, a lot of MCAS data is probably gathered on an Excel spreadsheet, or in Sharepoint, or a shared drive, and it’s someone’s job to pull the data. And, he or she is often not the subject expert of the report to be filed.
Overall, companies make a commendable effort in terms of timeliness and accurate data. But, to the extent that a carrier does not pay close attention to what’s going into the file, it can be a problem. You really don’t see the output very well from a 30,000-foot view; a carrier is far more likely to have issues unless it has a really solid data entry process in place or someone who owns it on the executive team who actually knows what is going into the report.
Any examples you can share?
One that comes to mind was a company that reported an unbelievably high replacement ratio. And when we dove into it, we realized they had pulled the wrong file to calculate the rate. Now, it worked itself out, and the ratio was actually much lower, which is a good thing, but again I think companies need to pay more attention to how they are filing this data and where they’re pulling it from.
And that’s where every company could do a little bit better job. I’ve had roles in three insurance companies now, and you can look at something as a check-the-box exercise, or hey-let’s-do-it-right. In my view, if you’re a bigger company, all of this does build into your ORSA filing in some respect.
See also: Why Risk Management Certifications Matter
Your Own Risk and Solvency Assessment is just a picture of where you are on a risk basis. But a lot of your risks are related to market issues. Every company can probably do a little bit better job of making sure the data you submit is timely, relevant and the right data.
And, when you’re looking at specific data with a report, the replacement rate within MCAS, for instance, how do you come up with that benchmark data? Are you looking at trending analysis in the context of industry benchmark data or trending within the company?
That’s a really good question. It’s more art than science; there isn’t one right way to do it. If you had a 75% replacement ratio, but you only sold four annuities, that may or may not mean anything. If you have a 75% replacement ratio, and you sold 25,000, that’s a different issue.
You start to look at it from a benchmarking of industry, a standard across the industry. Whether you can get that data from LOMA, LIMRA or WINK. Regulators have all of those same data points and benchmark studies, so you have a gut feel for what is an industry number.
Then beyond that, to your point, you’d have to dig down for context. For example, Transamerica sells a lot more life insurance and annuities than EMC National Life. A benchmark is a benchmark, but it doesn’t differentiate from a small mutual carrier or small stock carrier.
This is why context is really important. If you see a disturbing relationship or ratio develop on complaints, you have to look at the line of business, how much business they write, whether or not it’s an agent issue, or a producer issue, or home office issue, or a misunderstanding issue. You really have to dig in. Benchmarking is a start, and it’s certainly helpful.
Iowa has 216 carriers, and the vast majority are small or midsize, sometimes just county mutual carriers. You have to look at each carrier on its own, as well. The benchmark helps, but it’s not the end all and be all.
Did you look at consistency of data? For instance, premiums written is a component, in some form, of the financial reporting, market conduct and premium tax filings.
Certainly. Our team would look for consistency of data across filings. Our biggest bureau at the division was on the financial side. And that’s really where I spent a lot of my time to develop staff.
If we start to realize that a premium tax number doesn’t line up with premiums written, they start to ask questions. And sometimes there are good answers, and, other times, it’s a miss. And so, again, it’s data consistency and quality across all the reporting to make sure we have a clear picture.
Because oftentimes, it’s something we didn’t understand, or the carrier filed but didn’t pull the right number. The sophistication of the models that the companies use – as well as the sophistication of the reporting – varies greatly from small carriers to big carriers. Some have home-grown systems; some have ad hoc processes. It’s all done differently.
Do you have a sense – both from your time in industry as well as your role as insurance commissioner – how feasible it is to have a meaningful review process? To put this question in concrete terms: If you’re the CFO, you’re signing off on a lot of reports. Based on the volume of reports you’re signing, are you truly reviewing the data that’s being reported?
That’s a great question.
You’ve got reporting requirements for Sarbanes-Oxley if you’re public. You’ve got other reporting requirements under corporate governance at the state level. It’s impossible to dig into every single report for every single data point. So, you do have to rely on your staff, on your auditors and your chief accounting officer. And that’s why you have those controls in place leading up the reporting structure of those organizations.
That being said, a CFO would want to have a clear picture from a benchmarking dashboard. There are a lot of tools for people in the C-Suite for tracking and visualizing data that call out for attention when a metric is out of place or not reported.
The CFO relies on the team and the controls in place for the data to be correct in order to sign off. But, having a snapshot that showed what is filed, and when, and different data points and sources would be of immense help.
What are the consequences, from a regulator’s standpoint, of poor quality or inconsistent data? Is it reputational? Does it add to question marks around a company?
There are several things. Yes, it’s possibly reputational. But that’s in the longer term. Most immediately, the carrier is going to have to commit resources to resolve the issue.
If a commissioner’s officer is asking questions, he or she has found something. You’ve got to commit resources to adjudicate and resolve the issue. And, it could very well lead to a targeted exam, which, in turn, could end up as a full-blown market conduct exam.
It could also create a number of other issues during the triennial exam or the five-year deeper dive exam, which would require additional resources. These exams can cost quite a bit of money. And so, that’s a hard dollar cost. But, there is also the soft dollar cost of staff time, resources expended and opportunity cost in that it kept the carrier from have done something more productive.
How does this work in practice?
I can think of when I was commissioner once or twice when we had targeted exams based on filings that ultimately led us to say, “Okay, there is a problem here.” Both times were out-of-state companies.
To your point earlier, you can call an exam on any company that is doing business in your state, certainly on the market side. On the financial side, you’re going to have more deference. But, on the market side, every commissioner’s office is reviewing the data, as well.
Often for us, we would start with the complaints that are coming in, and then identify a trend with a carrier. And if you start to see a number of complaints, then you pull the data.
Some insurers have a cynical view of regulators, particularly in some states. I’ve heard them refer to this as “the cost of doing business.” They feel that, if you’re going to write policies in some states, you’re going to get fined from time to time. And then, if you get fined by one state, then you’re going to see fines from other states as well. How does this work in practice?
A carrier has an obligation to report a fine in all states in which it’s licensed. On top of that, there is this thing called the internet. When a state issues a fine – Commissioner Jones or Director Huff was famous for this – it would be followed by a press release, as well.
So, there is some truth to the idea that if an insurer has trouble in one state, it might have it in multiple states. But there is some right to have a level of cynicism. There are some states where you’re much more prone be fined. Whether this is a cost of doing business, that’s a decision for that management team. But, if there is a fine in one state, the chances that of it in multiple states is high
Our view of the world, in the Iowa division, was not necessarily to gang tackle but rather how to resolve the issue in our state. If there was a problem, we asked, “Did you make customers whole?” I would look at a systems issue with billing differently from an issue in which someone was ripped off. We tried to use judgment and look at the issues based on the facts and circumstances.
Currently, data flows from carriers to commissioners in a defined cadence. What do you think of the promises of regtech – the concept that software and system automation will allow for data to flow to regulators seamlessly, in real time and without the need for insurers to prepare and curate data for filings?
Right now the NAIC is the hub of a lot of this. And the idea that a state would get this directly from the insurer is a stretch.
What about through the NAIC?
Through the NAIC, I could see it happening. They’ll go to a cloud-based system, I’m guessing. As they make that shift, could that happen? Possibly.
I always joke that for the state of Iowa, and most states, you have the best technology from 1985. Some states are ‘95. It is a stretch to think that this could happen without the NAIC leading.
See also: The Current State of Risk Management
The NAIC really is the hub. If you’ve been to Kansas City, you’ve seen how impressive their system is, and their folks are. NIPR, for instance, I would always joke, is a technology firm. It’s not a producer licensing firm. The NAIC has tremendous resources. Their CTO has ideas on how to streamline it further. I could see this happening in 10 years or less. The reality is that a state could never do this.
So, a state has to rely on the NAIC. Going back to why this system works, well it works because you have an association - the NAIC - that has the ability to upgrade and transform quicker than any state ever could.
Is it possible that the states could innovate on their own, outside the NAIC?
It would be hard, at best. If you think about the state-based system, if Iowa doesn’t transform as quickly as California, or Montana as Wyoming, that starts to be a problem.
The NAIC can take care of that in one fell swoop and we, as state regulators, all benefit from that work.
I could see data delivery and reporting being quicker, more meaningful, real-time. I could even see, down the road, machine learning processes put in place to help on policy review form, financial review form. I think you could get there. I don’t know if it’ll be five years, 10 years or 15 years, but it will certainly happen in my career, where it’s going to be a continuously improving process.
The NAIC is the best way that regulators keep up with the demands that are happening, through leveraging the NAIC tech and personnel.