Actuarial Outpost
 
Go Back   Actuarial Outpost > Actuarial Discussion Forum > Risk Management
FlashChat Actuarial Discussion Preliminary Exams CAS/SOA Exams Cyberchat Around the World Suggestions

DW Simpson Global Actuarial & Analytcs Recruitment
Download our 2017 Actuarial Salary Survey
now with state-by-state salary information!


Reply
 
Thread Tools Display Modes
  #1  
Old 01-05-2018, 10:33 AM
ronaldy27's Avatar
ronaldy27 ronaldy27 is offline
Member
CAS
 
Join Date: Jan 2012
Studying for many many stuffs
Posts: 1,936
Default Economic Capital Modelling

Depending on how large a company is, there are many different roles that involve economic capital modelling. I believe the following are some of these roles:

-There are software engineers, data scientists who create their own ECMs and sell their models as vendors

-There are managers/analysts who just "use" these models and run them for results. Usually smaller companies buy cheaper models and require some time for model validation.

My question is, in which role do actuaries usually fit in?
I'm getting into ECM and I want to know which direction I'm headed and perhaps I could steer my future to a desirable area.


Thanks guys.
__________________
Spoiler - for size:



Reply With Quote
  #2  
Old 01-05-2018, 10:44 AM
JMO's Avatar
JMO JMO is offline
Carol Marler
Non-Actuary
 
Join Date: Sep 2001
Location: Back home again in Indiana
Studying for Nothing actuarial.
Posts: 36,776
Default

I think it depends on the type of company rather than the size. Companies in the insurance business mostly use models from vendors. There is a lot of actuarial judgement involved in setting assumptions and specifying the product features.

Software companies are the ones that mostly design the models.

So are you P/C or Life?
__________________
Carol Marler, "Just My Opinion"

Pluto is no longer a planet and I am no longer an actuary. Please take my opinions as non-actuarial.


My latest favorite quotes, updated Feb 15, 2018. Hmmm. It's been quite a while.

Spoiler:
I should keep these four permanently.
Quote:
Originally Posted by rekrap View Post
JMO is right
Quote:
Originally Posted by campbell View Post
I agree with JMO.
Quote:
Originally Posted by Westley View Post
And def agree w/ JMO.
Quote:
Originally Posted by MG View Post
This. And everything else JMO wrote.
And this all purpose permanent quote:
Quote:
Originally Posted by Dr T Non-Fan View Post
Yup, it is always someone else's fault.
MORE:
Quote:
Originally Posted by PeppermintPatty View Post
But sometimes you get someone who has that charming mix of "don't care" and "don't get it", and there's just not much you can do.
Quote:
Originally Posted by Bro View Post
I recommend you get perspective.
Quote:
Originally Posted by Enough Exams Already View Post
Dude, you can't fail a personality test. It just isn't that kind of test.
Reply With Quote
  #3  
Old 01-05-2018, 10:52 AM
ronaldy27's Avatar
ronaldy27 ronaldy27 is offline
Member
CAS
 
Join Date: Jan 2012
Studying for many many stuffs
Posts: 1,936
Default

Quote:
Originally Posted by JMO View Post
I think it depends on the type of company rather than the size. Companies in the insurance business mostly use models from vendors. There is a lot of actuarial judgement involved in setting assumptions and specifying the product features.

Software companies are the ones that mostly design the models.

So are you P/C or Life?
Hi JMO,

I'm in P&C. My company uses a smaller vendor's model but I'm interested in getting into "big vendor's" stuff like WTW's Igloo.

Do models vary a lot or is it fairly easy to learn to use a new model if you have prior experience using other models?
__________________
Spoiler - for size:



Reply With Quote
  #4  
Old 01-05-2018, 11:34 AM
Corn Row's Avatar
Corn Row Corn Row is offline
Member
 
Join Date: May 2003
Posts: 861
Default

Which smaller vendor? The market share leaders, WTW and Aon, are far inferior, IMO, to Ultimate Risk Solutions' Risk Explorer.

WTW's Igloo is good if you 1) have a large team supporting the product, and 2) have a model mostly used for regulatory (e.g. Lloyd's) purposes. The reason Igloo is good for #2 is that you can't change your model quickly or easily which lends itself to a product that is difficult to change. With Risk Explorer, you can redesign your entire approach in weeks, if desired.

There are other models, such as GC MetaRisk and Conning's ADVISE. I've never heard one positive thing said about ADVISE, though Conning's ESG, GEMS, seems to be a rockstar over the past few years. MetaRisk is good for reinsurance evaluation, but severely lacks when it comes to full financial statement needs.

All the models, excluding Risk Explorer, are pretty awful at asset modeling. I've seen Igloo models regularly treat MBS as corporate bonds. ADVISE may be good at asset modeling - I would hope it is since it's now owned by Conning.

Regarding ease of learning, Risk Explorer is likely the "easiest" as it has an intuitive interface. Igloo is probably the most difficult as it's essentially a programming language.

Actuaries are often involved at all stages, though there are plenty of statisticians and MBA types that fill certain roles. I see capital modeling as a jack of all trades and master of none. Your role is often to understand all different aspects of the company, from asset modeling, reinsurance, volatility/correlation, financial statement analysis & accounting, possibly tax, and more. An actuarial background helps, but being curious, analytical, and intelligent are more important.
__________________
Lord, beer me strength.
Reply With Quote
  #5  
Old 01-05-2018, 11:40 AM
ronaldy27's Avatar
ronaldy27 ronaldy27 is offline
Member
CAS
 
Join Date: Jan 2012
Studying for many many stuffs
Posts: 1,936
Default

Quote:
Originally Posted by Corn Row View Post
Which smaller vendor? The market share leaders, WTW and Aon, are far inferior, IMO, to Ultimate Risk Solutions' Risk Explorer.

WTW's Igloo is good if you 1) have a large team supporting the product, and 2) have a model mostly used for regulatory (e.g. Lloyd's) purposes. The reason Igloo is good for #2 is that you can't change your model quickly or easily which lends itself to a product that is difficult to change. With Risk Explorer, you can redesign your entire approach in weeks, if desired.

There are other models, such as GC MetaRisk and Conning's ADVISE. I've never heard one positive thing said about ADVISE, though Conning's ESG, GEMS, seems to be a rockstar over the past few years. MetaRisk is good for reinsurance evaluation, but severely lacks when it comes to full financial statement needs.

All the models, excluding Risk Explorer, are pretty awful at asset modeling. I've seen Igloo models regularly treat MBS as corporate bonds. ADVISE may be good at asset modeling - I would hope it is since it's now owned by Conning.

Regarding ease of learning, Risk Explorer is likely the "easiest" as it has an intuitive interface. Igloo is probably the most difficult as it's essentially a programming language.

Actuaries are often involved at all stages, though there are plenty of statisticians and MBA types that fill certain roles. I see capital modeling as a jack of all trades and master of none. Your role is often to understand all different aspects of the company, from asset modeling, reinsurance, volatility/correlation, financial statement analysis & accounting, possibly tax, and more. An actuarial background helps, but being curious, analytical, and intelligent are more important.
The more I get into this stuff, I'm starting to think everyone fits that description. Either we dive into the programming aspect and be very good at that or as someone in management, just learn to use the model and know the business well.


Thank you for the reply. What is your take on Aon's model? I believe it's called ReMetrica?
__________________
Spoiler - for size:



Reply With Quote
  #6  
Old 01-05-2018, 12:17 PM
Corn Row's Avatar
Corn Row Corn Row is offline
Member
 
Join Date: May 2003
Posts: 861
Default

Quote:
Originally Posted by ronaldy27 View Post
The more I get into this stuff, I'm starting to think everyone fits that description. Either we dive into the programming aspect and be very good at that or as someone in management, just learn to use the model and know the business well.


Thank you for the reply. What is your take on Aon's model? I believe it's called ReMetrica?
Not impressed. Igloo and ReMetrica are market leaders because 1) they had the best products years ago when companies were starting their capital modeling journey and 2) large, and very expensive, consulting benches to support new users.

If all products were released right now, there is no chance that Igloo or ReMetrica would be leaders. However, it's not an easy task to change capital models (regulatory hurdles, staffing hurdles, if it's not broken why fix it) so they maintain their market share. Since they have dominant market share (and large companies behind them pushing their product) they are the safe choice.

The result is that you have a whole industry, especially in London, that has these very expensive capital models which do one thing: produce regulatory reports. There is no true "use" of the model. There is no evaluating reinsurance on the fly during renewal time (or if there is then it's an overly simplified version of that reinsurance product because you try programming a complex inurance structure on a top-and-drop coverage where different perils have different attachments and there is an AAD... nope, in Igloo/ReMetrica you just pretend it's a basic XoL). Basically, the capital modeling industry, because of these models, is seen as a cost center and not a source of business intelligence/value producing analytics.

The easiest thing in the world to do is to take an AIR YLT file and include it in your capital model. You literally just have to read the file... the first "year" is all the events that happened in the simulation 1. See if it's that easy in ReMetrica (in my experience it's not and it's a convoluted process that is overly complex). I can't imagine how difficult they've likely made including a RMS ELT or doing relatively straightforward cat model blending.
__________________
Lord, beer me strength.
Reply With Quote
  #7  
Old 01-16-2018, 03:11 PM
ronaldy27's Avatar
ronaldy27 ronaldy27 is offline
Member
CAS
 
Join Date: Jan 2012
Studying for many many stuffs
Posts: 1,936
Default

What's the underlying programming languages used in Igloo?
Is it Python?

C++?
__________________
Spoiler - for size:




Last edited by ronaldy27; 01-16-2018 at 03:14 PM..
Reply With Quote
  #8  
Old 02-11-2018, 09:45 PM
PAC PAC is offline
Member
CAS AAA
 
Join Date: Jan 2003
Posts: 230
Default

General comments...

A well-rounded actuarial practitioner should be able to build and model ALM/ECM from scratch in excel (w/VB or @Risk or I suppose R) and the choice of modeling tool/software should obviously make sense for the task at hand.

IGLOO or REMetrica or RiskExplorer or GEMS/ADVISE are all designed to effect the same result including a bench of consultants to facilitate the process.

Other contenders - in ALM and ECM generally - include QRM and polypaths.

If interested, the CAS produced a Public Access DFA model that may instructive.
Reply With Quote
  #9  
Old 02-12-2018, 10:11 AM
Corn Row's Avatar
Corn Row Corn Row is offline
Member
 
Join Date: May 2003
Posts: 861
Default

Quote:
Originally Posted by PAC View Post
General comments...

A well-rounded actuarial practitioner should be able to build and model ALM/ECM from scratch in excel (w/VB or @Risk or I suppose R) and the choice of modeling tool/software should obviously make sense for the task at hand.
That would be one hell of a basic model unless you're giving the well-rounded actuarial practitioner a number of years to create something from scratch. Off the top of my head they would need to develop a robust solution for complicated reinsurance treaties and structures, non traditional asset valuation (bonds with options) and more complicated traditional assets (structured products), all sorts of accounting knowledge, potentially including tax accounting, etc. They better be familiar with distributed computing techniques unless they want their model to take days or months to complete a simulation.

Sure, someone could use @Risk to create a small model with a few attritional and large loss sources of risk, attach a quota share or simple XoL and produce some output, but they would be a decade behind other technologies. I've seen plenty of Excel-type capital models and they are not adequate.
__________________
Lord, beer me strength.
Reply With Quote
  #10  
Old 02-12-2018, 10:57 AM
jas66Kent's Avatar
jas66Kent jas66Kent is offline
Member
SOA
 
Join Date: May 2012
Location: London
Favorite beer: Corona :)
Posts: 21,819
Default

You can't do this sort of thing in excel.

Excel was only ever meant for mild-moderate financial modelling.

Please stop abusing it.
__________________
Quote:
“Stupidity is a more dangerous enemy of the good than malice. Against stupidity we are defenseless. Neither protests nor the use of force accomplish anything here; reasons fall on deaf ears; facts that contradict one’s prejudgment simply need not be believed – in such moments the stupid person even becomes critical – and when facts are irrefutable they are just pushed aside as inconsequential, as incidental. For that reason, greater caution is called for when dealing with a stupid person than with a malicious one".
Reply With Quote
Reply

Thread Tools
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 03:47 PM.


Powered by vBulletin®
Copyright ©2000 - 2018, 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 0.17100 seconds with 9 queries