How and why to leave legacy actuarial systems behind

How and why to leave legacy actuarial systems behind

Let’s face it - your existing actuarial technology is hopelessly outdated.

Slow adoption of modern alternatives means every actuary is working with one hand tied behind their back thanks to:

  • Exceedingly long run times, which severely limit productivity
  • The need for highly specialized legacy software skills to develop and run models
  • Hard or impossible to validate models with inadequate process control
  • The inability to integrate with data sources and data science tooling

So the real question is: aren’t you tired of using tools from the 90s to do your job in 2022?

Actuarial workflows are frustratingly inefficient. Throughout the actuarial process, both in valuation and commercial teams, there’s clear opportunities for cloud-native actuarial technologies to improve the process, its outputs, and general productivity. At Montoux, we call this Actuarial Automation, a key capability enabled by our Platform

Use Case 1: Reducing model run times

Generally, productivity as an actuary is measured in the number of runs you can do in a day or a week. Because existing actuarial models run so slowly and the process is sequential, legacy systems place hard limitations on productivity. 

The quicker you can complete runs, the more productive you are. Most actuarial tools run on on-premise infrastructure and are inherently limited in terms of compute capacity. Montoux’s modern, cloud-based infrastructure enables elastic and scalable computations, reducing run times from hours to minutes. This automation also lets you run scenarios simultaneously to visualize what works best.

Use Case 2: Model flexibility and clarity (getting away from the black box)

When it comes to actuarial models, few things are more frustrating or risky than the black box issue. If you are the only actuary who can change models and understand how they work, the rest of the team is running scenarios in the dark. This can lead to a massive backlog of work, model validation issues, material reporting errors, and even significant fines.

To make matters worse, legacy software models are so hard to understand that it’s difficult to find auditors to provide positive assurance on models. All you can do is hope the models don’t have unfavorable, material errors.

Actuarial automation, coupled with a model viewer, and a highly transparent modeling framework provided by Montoux’s Platform, provides your actuarial  team with a transparent view of how the model is working. Everyone can understand the model and complete their own runs to submit for review, reducing mistakes and time wasted.

We're working with AWS on some exciting developments in this space, and in actuarial model replication. Check it out here.

Use Case 3: Clear and concise outputs

Traditional systems and tools are not designed for reviewing, organizing, and comparing actuarial model outputs. Using tools like Excel, you’re forced to ask for less information than you may want, due to long run times, and can’t easily compare and sort the results. 

Montoux’s actuarial automation capabilities makes it easy for you to compare outputs and prioritize the ones you need. It also lets you use any amount of data and information you want, aligning different scenarios for easy comparison. 

Use Case 4: Replacing highly complex table assumptions structures with an integrated data science routine

Increasingly, carriers are applying data science to assumption setting, resulting in significantly more accurate value and risk estimates as well as more targeted business decisions.

Data science tooling enables your actuarial team to create complex functions which provide significantly more reliable estimates. However, shortcomings in legacy software means these functions can’t be embedded or integrated. Instead, you and your data scientist colleagues must go through a painful process of translating these series of complex tables.

Montoux’s Platform enables you to integrate the results of the data process (like a Python function and its input tables) directly into your actuarial model, removing the painful part of the process and reducing the risk of mistakes in the translation process.

Use Case 5: Integrating data science tooling to create powerful solvers and goal seek functionality

Have you ever been asked to identify how much premium rates can be reduced while ensuring the IRR does not drop below 8 percent? If not, chances are you’ve been asked something similar which requires manual goal seek and solver processes. 

Numerous, out-of-the-box data science algorithms can automatically solve this problem, but data science tooling, like R and Python, aren’t compatible with legacy actuarial models. Extremely long run times on traditional models make the use of additional tooling even more time consuming.

Actuarial automation reduces model run times to minutes, if not seconds, and enables the integration of data science tools. Just like that, goal seeking and solving is automated, significantly more powerful, can run highly complex goal seek exercises, and gives the business deep insight into how to create the most value from existing options.

Use Case 6: Democratizing model access

The usability of actuarial models, aka how easy it is to define the next run, varies dramatically; for example, does adjusting every premium involve changing 23 tables or two numbers?

Depending on your familiarity with a particular model, it can be frustratingly difficult to adjust model inputs. What’s more, existing models limit experimentation and model scenarios because it takes so long to complete one analysis. That’s before you add external data sources.

Montoux’s Platform enables intuitive and clear UI design across actuarial models, enabling you to comfortably use any models required to analyze, adjust, or illustrate the data you need.

Use Case 7: Use proxy models to provide additional checks on your valuation runs

As a Chief Actuary, do you lie awake at night thinking about how you rely on your expert actuaries to go through high-level results and identify patterns that look ‘’off’’?

While these manual checks are very important, they are not bulletproof. It’s much easier for your actuarial team to complete a thorough review when they have access to proxy models, which offer another estimate of the key results they are expecting from a run.

Accurate proxy models can be constructed quickly through a combination of model point reduction techniques and low-sample proxy modeling techniques. On Montoux’s Platform, creating an accurate proxy model takes at most one or two hours, and provides significant comfort at a modest level of investment.

Use Case 8: Analysis change control

Actuarial valuation teams are required to go through highly complex processes, and any process errors can have significant consequences. Legacy actuarial technology fails to provide analysis change control out-of-the-box, and as a result many teams still rely on Excel to manage the process.

Modern actuarial software should directly integrate with process control tools to ensure every step happens in the right order, gets properly reviewed, and all actions are logged. Montoux’s Platform and actuarial automation capabilities enables this.

Automation improves results and sends actuarial productivity through the roof

When it comes to wasted actuarial resources, the problem is with everything but you, the actuary. Legacy technologies create wasteful siloes and slowdowns which limit productivity. 

With Montoux’s Platform, insurers can leverage actuarial automation, enabling actuaries to work smarter, not harder. With quicker runtimes, user friendly models, and clear outputs, you can work more collaboratively, productively, and put your skills to work on projects which deliver strategic, high-value, and impactful results for the company. 

Want to set up a demo or discuss any of these use cases in more detail? Reach out to our team here.

Back to blog home