“What gets measured gets managed – even when it’s pointless to measure and manage it, and even if it harms the purpose of the organization to do so.”
– V.F. Ridgway’s “Dysfunctional Consequences of Performance Measures”
The title might be a bit controversial for actuaries, but let’s start off by clarifying the “mostly” part. When I speak of rate change, I’m talking about E&S, reinsurance, large account pricing, and everything else that doesn’t conform to a nice, cookie cutter, pure, algorithmically driven price. I do have beef with rate change as a whole, but I’ll reserve those thoughts later for when I ramble about rate indications and why they are easily broken.
If you’re a pricing actuary, you’ve been asked to calculate the rate change on your account, or to roll up the rate change on a whole set of accounts. Remember the time that request was asked of you. It was something along the lines of “how much rate did we get on account X?” or “for portfolio Y, did we achieve Z amount of rate?”. There isn’t anything crazy about these questions, except they lack two key bits of information. WHY is that number important, and WHAT are you going to do once they have that number?
As actuaries we want to calculate that number, number crunching is where we get our cheap thrills. But we need to step back for a second, and define terms before we get deep into things. How is rate change defined? If you went the CAS route you first encountered rate change studying for Exam 5. Digging through the canonical text1 you’ll find that “rate change” is never directly defined. It’s loosely used when talking about changes to base rates or rating factors. In general, most actuaries would agree that rate is calculated premium as a ratio to exposure.
For a large account, treaty, or portfolio, then, the rate change could be defined as (P_r/E_r)/(P_o/E_o), where the “o” subscript is the original and “r” is the renewal. If you slightly rearrange this it’s just the change in premium amounts scaled by an exposure adjustment.
What does this actually tell you? By itself, it only tells you whether or not you are getting more dollars for a given amount of exposure. So a positive rate change is good, right? It depends. What if you just discovered the risk is right over some newly mapped fault line? In that case, the expected loss on the account just shot straight up; if the losses are up 50% and your rate change is +5%, you have a problem.
Given all the changes that can occur in a renewal, people now start talking about “risk-adjusted rate change”. This is just a fancy way of saying we want to somehow account for this updated view on loss potential. Ignoring how we could mathematically define this number, we’ll focus on what we’re trying to answer. What’s important is whether we think we’re going to be making more profit, after we’ve adjusted for the more or less risk we’re taking on. This starts to get crazy hard. How do you account for new coverages, different structures, contractual language, or any other standard change? You need to pick your battles on this one, and thankfully there are some actuaries out there who’ve thought about this quite some bit.23
Reviewing the historic literature will highlight some key themes.
- Defining and calculating rate change, or risk-adjusted rate change is not a settled issue.
- You can make the calculations as complicated as you want it to be.
After you’ve done your extensive literature review and decided on a method, you’re now racking your brain on how to fit it all into a nice dashboard or Excel workbook. However, what haven’t we answered in all of this? What is anyone going to DO with this number?
Even if you had the Platonic ideal of a rate change perfectly calculated4, it’s not actionable because the account or portfolio is already bound and written. It’s in the past, it’s collecting dust, it’s now in the rear-view mirror. All you’ve done is successfully describe what happened over the past renewal cycle. In that sense, sure, maybe there is a sliver of usefulness there because the bonus target was a function of achieved rate change. You can pat yourself on the back, unless the rate change was negative, because the state of things improved according to one metric.
By now, you have a sense for my feelings on rate change. Is it an interesting actuarial puzzle to noodle over? Yes. Is it a number that’s going to drive meaningful action? Not really. So am I here to complain, or do I think there is some hope on where we spend our time?
Us actuaries should be spending way more time focusing on thinking about the future, and forecasting the pricing dynamics before the renewal cycle begins. Think about how valuable it would have been if an actuary had even a slight view that the post-COVID era was going to show a massive spike in inflation. You would of had real action items on the table to try and manage the risk that was headed your way. But once accounts are renewed that’s it. The reserving actuaries and time will tell if the portfolio or account performed better than last time, but dwelling on the details of the rate change isn’t going to make things better.
What I’m pondering about:
- Rate & price adequacy rather than rate change
- How to better incorporate different market drivers and their impact on price & rate movement
- A colleague brought up an interesting idea of looking at historic figures and getting a sense for the macro cycle. While our individual estimates might be wonky, are their nuggets of wisdom to be gained by looking over longer time horizons? Cycle management is notoriously difficult, but it’s probably where there is most to be gained.
- https://www.casact.org/sites/default/files/old/studynotes_werner_modlin_ratemaking.pdf ↩︎
- https://www.casact.org/sites/default/files/database/forum_09wforum_bodoff.pdf ↩︎
- https://www.casact.org/sites/default/files/database/forum_09wforum_robbin.pdf ↩︎
- My favorite is when you do calculate some horrendously complex rate change number, and management’s reply is “that doesn’t seem right”. Clearly, they already had a number in their head anyway so what’s the point? ↩︎
It's Gonna Lose Money, Jim 