Actuarial Outpost
 
Go Back   Actuarial Outpost > Actuarial Discussion Forum > Careers - Employment
FlashChat Actuarial Discussion Preliminary Exams CAS/SOA Exams Cyberchat Around the World Suggestions


Reply
 
Thread Tools Search this Thread Display Modes
  #1  
Old 05-12-2019, 03:08 PM
Modigliani-Miller Modigliani-Miller is offline
SOA
 
Join Date: Aug 2018
Posts: 25
Default What actuarial fields use Predictive Analytics

So SOA is having every new ASA take the PA exam. But what employers actually hire actuaries to do predictive analytics work?

I know my company does not. I work for a health insurance company. PA is still new and is done by non-actuarial data science people. So I'm afraid I will face the "use it or lose it" problem.

If I want to do data science type of work as an actuary, what employers should I look for? I hope switch to P&C and work for Uber is not the only way out...

Last edited by Modigliani-Miller; 05-12-2019 at 03:42 PM..
Reply With Quote
  #2  
Old 05-12-2019, 03:18 PM
Westley Westley is offline
Member
 
Join Date: Nov 2001
Posts: 28,683
Default

Most fields can use PA; not all do, and not all that do use actuaries for it.

If your company is using PA and you have an interest, why don't you ask if they have an interest in including an actuary on the PA team?

PA def has a "use it or lose it" problem. I'm not really sure an exam structure is the way to go at all: from the time they determine the exam materials until you receive a passing score can easily be several years, at which point you're not even cutting-edge, and you continue to fall behind from there if you're not on a PA team. If you're going to be an actuary doing traditional actuarial work and working together with a PA team, that's probably ok; for working on a PA team, not so much.

I do think that there's a lot more on the P/C side, but you don't have to work for Uber.
Reply With Quote
  #3  
Old 05-12-2019, 03:41 PM
hostess's Avatar
hostess hostess is offline
Member
CAS
 
Join Date: Aug 2018
Posts: 621
Default

as a followup, what is the difference between predicitive analytics and modelling? is one predictive whereas the other is inferential?
Reply With Quote
  #4  
Old 05-12-2019, 03:47 PM
Modigliani-Miller Modigliani-Miller is offline
SOA
 
Join Date: Aug 2018
Posts: 25
Default

Thanks! I have definitely thought about asking the data science team that question. But, on the other hand, I feel what I learned from the PA exam is not nearly enough for doing data science work and I'm afraid my actuarial skills will not be valued in the DS team.

Quote:
Originally Posted by Westley View Post
Most fields can use PA; not all do, and not all that do use actuaries for it.

If your company is using PA and you have an interest, why don't you ask if they have an interest in including an actuary on the PA team?

PA def has a "use it or lose it" problem. I'm not really sure an exam structure is the way to go at all: from the time they determine the exam materials until you receive a passing score can easily be several years, at which point you're not even cutting-edge, and you continue to fall behind from there if you're not on a PA team. If you're going to be an actuary doing traditional actuarial work and working together with a PA team, that's probably ok; for working on a PA team, not so much.

I do think that there's a lot more on the P/C side, but you don't have to work for Uber.
Reply With Quote
  #5  
Old 05-12-2019, 03:49 PM
Modigliani-Miller Modigliani-Miller is offline
SOA
 
Join Date: Aug 2018
Posts: 25
Default

Quote:
Originally Posted by hostess View Post
as a followup, what is the difference between predicitive analytics and modelling? is one predictive whereas the other is inferential?
That sounds right! Really depends on the problem you're trying to solve. Both predictive and inferential use modeling techniques.
Reply With Quote
  #6  
Old 05-12-2019, 05:46 PM
Westley Westley is offline
Member
 
Join Date: Nov 2001
Posts: 28,683
Default

Quote:
Originally Posted by hostess View Post
as a followup, what is the difference between predicitive analytics and modelling? is one predictive whereas the other is inferential?
Hopefully some more knowledgeable people will speak up. I'll speak from a non-DS perspective. When HR, senior management, investors, and other non-technical people are speaking, all of the following are roughly synonyms: PA, modeling, data science, big data (I think most people now realize that using this term means you are a clown), and perhaps even business analytics.

When technical people use those terms, they might mean something more specific, and I think it's worth asking what a specific group does.


Quote:
Originally Posted by Modigliani-Miller View Post
Thanks! I have definitely thought about asking the data science team that question. But, on the other hand, I feel what I learned from the PA exam is not nearly enough for doing data science work and I'm afraid my actuarial skills will not be valued in the DS team.
If you learned enough from one exam to be valued on the DS team, that's not a good sign for your DS team. Your value to the DS team would be
1) I have a beginner's knowledge of what you do
2) I have interest and ability to eventually be a contributor
3) my knowledge of other areas of the company can also be a valuable part of what you're currently doing.

Will they value this? You don't know until you have a conversation. Also, if the DS team is somewhat insular, they may not see the value, but somebody higher up might still see the value. Whether you want to fight that battle (being put on a team that doesn't value your overall abilities) is a separate question to consider.
Reply With Quote
  #7  
Old 05-12-2019, 06:07 PM
hostess's Avatar
hostess hostess is offline
Member
CAS
 
Join Date: Aug 2018
Posts: 621
Default

AFAIK insurers have made use of modelling since the start. It never made much sense to me when people say PA has just started to make its move into blank industry (if PA is the same as modelling). Auto insurers have always used rating variables and life insurers have always used mortality tables. Surely these are models of their own.

My guess was PA referred to the new wave of modelling (transitition from traditional excel raters to GLM's and more dynamic models).

Feel free to correct me as I'm fairly new to the profession....

Last edited by hostess; 05-12-2019 at 06:21 PM..
Reply With Quote
  #8  
Old 05-12-2019, 06:10 PM
campbell's Avatar
campbell campbell is offline
Mary Pat Campbell
SOA AAA
 
Join Date: Nov 2003
Location: NY
Studying for duolingo and coursera
Favorite beer: Murphy's Irish Stout
Posts: 87,614
Blog Entries: 6
Default

A big part of this is the amount of the data you have......and the quality of that data.
__________________
It's STUMP

LinkedIn Profile
Reply With Quote
  #9  
Old 05-12-2019, 09:05 PM
Colonel Smoothie's Avatar
Colonel Smoothie Colonel Smoothie is offline
Member
CAS
 
Join Date: Sep 2010
College: Jamba Juice University
Favorite beer: AO Amber Ale
Posts: 48,805
Default

Quote:
Originally Posted by campbell View Post
A big part of this is the amount of the data you have......and the quality of that data.
The good news is I think carriers, particularly personal auto are realizing this and beefing up the rest of their DS ecosystem, by hiring data engineers and taking a more active approach to acquisition.
__________________
Recommended Readings for the EL Actuary || Recommended Readings for the EB Actuary

Quote:
Originally Posted by Wigmeister General View Post
Don't you even think about sending me your resume. I'll turn it into an origami boulder and return it to you.
Reply With Quote
  #10  
Old 05-13-2019, 03:38 AM
The_Polymath's Avatar
The_Polymath The_Polymath is offline
Member
CAS SOA
 
Join Date: Jun 2016
Posts: 1,689
Default

All the big reinsurers use it now.

It's quite interesting. For example, for motor claims in rural areas, you can factor in the density of deer populations in the general area (even adjust them for their seasonal migration
patterns) into your model, as that is a potential claims vector.
__________________
Quote:
Consistency is the hobgoblin of little minds
Reply With Quote
Reply

Tags
data science, machine learning, pred. anal.

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off


All times are GMT -4. The time now is 11:12 AM.


Powered by vBulletin®
Copyright ©2000 - 2019, Jelsoft Enterprises Ltd.
*PLEASE NOTE: Posts are not checked for accuracy, and do not
represent the views of the Actuarial Outpost or its sponsors.
Page generated in 0.16189 seconds with 9 queries