In the world of Machine Learning, is Actuary Redundant ?
Updated: Feb 2, 2019
The core activity of an Actuary is:
1. Number crunching 2. Assumption setting
But this will no longer be the case, in the world of powerful processors and highly evolved machine learning algorithms powered by huge data.
With more powerful machines and programming models, number crunching has become very easy.
With the advent of the models like MapReduce, most complicated calculations for millions of records can be distributed across multiple networked machines and can be completed in few minutes or hours.
Once the code is written and testing done, the machine is all set to do the calculations.
The other core activity is assumption setting. A lot of judgement is required in setting assumptions like mortality, morbidity, persistency etc. Using machine learning algorithms and analytical tools, the future trends can be predicted more accurately. Using machine learning algorithms like classification and clustering, a lot of slicing and dicing of data can be done, and assumptions can be made and validated based on such slicing, which results into a lot more accurate calculations.
But is it good or bad for an actuary?
This is good for customer. This minimises cross subsidy and helps customer pay what she deserves to pay as premium. But the debate is, what does it do to an actuary? Is it good or bad for an actuary?
In olden days, the Actuarial calculations were done manually using commutation functions. That’s no longer the case. The calculations are done by machines now. Although, a lot of effort goes into writing the code and checking the results. With more and more advanced machines, some of this effort may come down.
Similarly, analytical tools may help making assumptions more accurate. This minimizes errors in assumption setting.
In other words, it will only help an actuary. He can now spend more time in making sense out of numbers and give critical feedback to management on the strategy. Depending on trends, he might, for example, suggest different benefit structure for different age groups of customers. He might come up with some new products to certain niche segments.
On the reporting side, the unnecessary extra margins can be reduced in calculating liability. This can make the company more capital efficient. Reducing cost of capital will ultimately reduce the price paid by a customer.
All this will help actuary in getting deep insights of the business. Hence he will become a key contributor to the business strategy of the company. He will also be able to contribute significantly on key decisions related to operations, finance and other departments. Already, IFRS 17 seems to be creating a lot of space for Actuaries in the finance area. Another important area is communication. Every actuary should necessarily put in effort to de-mystify the complexities of insurance business to non Actuaries. This will be a big contribution to the business.
In other words, from a number cruncher, the role of an actuary will transform into a key contributor to various aspects of the business. In fact, Actuaries would be encouraged to work in other departments, given their knowledge and understanding. Though, Actuaries, rather most of them, feel comfortable in working in Actuarial department, days are not far that they are seen everywhere, in an insurance company.
I would be curious to know your views on this. Please leave your thoughts below.