By Beyers Strydom.
These data points don’t lie...
What happens when an insurance company gets accused of underpaying its reinsurer?
Well, after you’ve let the situation cool down, you go look at the data. Or at least, that’s what I did as an aspiring actuary.
I recently worked with a client who wanted to reconcile premiums paid to their reinsurer and settle an underpayment dispute.
My client felt like they had been paying the right amounts each year and didn’t want to fork out thousands of additional dollars (these treaties can get quite big at times). The problem was that their data was not well maintained and they had no means of checking whether the premiums they paid to the reinsurer were correct. Without the necessary evidence, their case wasn’t very strong.
That’s where my expertise came in. To help them build their case, I developed a model to calculate monthly reinsurance premiums for each policyholder throughout the period of the reinsurance treaty.
But it wasn’t all plain sailing. The terms of the reinsurance treaties were subject to change over time, which meant that the reinsurance premium calculator had to dynamically allow for these changes. I ended up building several individual calculators to process the monthly data and calculate the reinsurance premiums.
It was an extensive project that involved data cleaning, model development and extensive back testing - checking whether the models were predicting the premiums correctly by comparing the calculated premiums against the historical data.
In the end, using my model and its outputs, my client and its reinsurer found a way to settle their differences. All’s well that ends well and this just shows again that the data doesn’t lie.
Commentaires