Actuarial Outpost
 
Go Back   Actuarial Outpost > Search Forums
FlashChat Actuarial Discussion Preliminary Exams CAS/SOA Exams Cyberchat Around the World Suggestions



Showing results 1 to 25 of 300
Search took 0.10 seconds.
Search: Posts Made By: Heywood J
Forum: Exam 8 - Advanced Ratemaking 07-02-2019, 11:42 AM
Replies: 453
Views: 44,055
Posted By Heywood J
You probably don't want to be fumbling with the...

You probably don't want to be fumbling with the NCCI or ISO manuals during the exam, and figuring them out for the first time. It takes some practice to jump to the right page quickly, or to know...
Forum: Property - Casualty / General Insurance 06-18-2019, 02:11 PM
Replies: 21
Views: 1,842
Posted By Heywood J
Why can't loss trend be negative? When I hear...

Why can't loss trend be negative? When I hear actuaries say that, it reminds me of real estate analysts in 2005 assuming that housing price appreciation cannot be negative.
Forum: Property - Casualty / General Insurance 06-18-2019, 02:06 PM
Replies: 67
Views: 8,360
Posted By Heywood J
There is nothing that comes close to R Studio in...

There is nothing that comes close to R Studio in Python, as far as I'm aware. The ease with which R Studio allows you to inspect intermediate output on demand is another reason why actuaries should...
Forum: Property - Casualty / General Insurance 06-18-2019, 11:34 AM
Replies: 67
Views: 8,360
Posted By Heywood J
I think R vs. Python debate is really a debate...

I think R vs. Python debate is really a debate between statistically-minded data scientists and CS-minded data scientists. If you're dealing with a flood of data, and your challenge is to make...
Forum: Property - Casualty / General Insurance 06-17-2019, 08:40 AM
Replies: 6
Views: 622
Posted By Heywood J
One nitpick: I don't think it's correct to refer...

One nitpick: I don't think it's correct to refer to Amount of Insurance factor in property insurance as an ILF. Limits for liability insurance are controlled by the policyholder, but AOI for...
Forum: Property - Casualty / General Insurance 06-17-2019, 08:06 AM
Replies: 6
Views: 622
Posted By Heywood J
These kinds of relativities can indeed be...

These kinds of relativities can indeed be indicated by the LER approach, if you completely cut up your data by amount of insurance ranges, so that nothing is tying your results for one range to your...
Forum: Property - Casualty / General Insurance 03-21-2019, 10:32 AM
Replies: 8
Views: 908
Posted By Heywood J
It seems to me that there isn't much need for a...

It seems to me that there isn't much need for a credibility procedure for premium trend. You need a credibility procedure for pure premium, since few policies have losses, and when they do happen,...
Forum: Property - Casualty / General Insurance 03-07-2019, 01:21 PM
Replies: 5
Views: 551
Posted By Heywood J
Which part do you find wrong? If you model some...

Which part do you find wrong? If you model some kind of a ratio, putting whatever is in the denominator as a weight is the default thing to do. That would give more weight to observations with...
Forum: Property - Casualty / General Insurance 02-22-2019, 10:10 AM
Replies: 57
Views: 2,894
Posted By Heywood J
That's not the reason. Standard errors are more...

That's not the reason. Standard errors are more a function of the estimate of your dispersion parameter; the higher the dispersion, the higher the standard error.

For standard Poisson,...
Forum: Property - Casualty / General Insurance 02-18-2019, 11:53 PM
Replies: 57
Views: 2,894
Posted By Heywood J
What is the difference between ODP and Poisson...

What is the difference between ODP and Poisson model when it comes to fitted values in GLM?
Forum: Property - Casualty / General Insurance 02-18-2019, 11:34 PM
Replies: 31
Views: 2,175
Posted By Heywood J
As far as R packages, I suggest sticking with...

As far as R packages, I suggest sticking with tidyverse, unless you have a good reason to go with data.table. The simplicity and compactness of syntax in tidyverse for data manipulation is unmatched.
Forum: Property - Casualty / General Insurance 02-14-2019, 11:09 AM
Replies: 57
Views: 2,894
Posted By Heywood J
In general, GLMs are simplistic enough that they...

In general, GLMs are simplistic enough that they are fairly robust to the choice of distribution, within reason. With more sophisticated models, that may no longer be the case, but then you'd...
Forum: Exam 8 - Advanced Ratemaking 02-08-2019, 10:50 AM
Replies: 3,094
Views: 257,044
Posted By Heywood J
The correct answer is that if you're looking for...

The correct answer is that if you're looking for work, you're counted as unemployed. The length of your unemployment has nothing to do with it.
Forum: Exam 8 - Advanced Ratemaking 02-06-2019, 03:31 PM
Replies: 3,094
Views: 257,044
Posted By Heywood J
That's not true.

That's not true.
Forum: Property - Casualty / General Insurance 01-25-2019, 08:52 AM
Replies: 31
Views: 2,175
Posted By Heywood J
The other basic one is that you will almost...

The other basic one is that you will almost certainly be excluding some data from the GLM analysis, either because it's bad or because it's just too out of the ordinary. There is no guarantee that...
Forum: Property - Casualty / General Insurance 01-24-2019, 10:09 PM
Replies: 31
Views: 2,175
Posted By Heywood J
I think that's dangerous. There are multiple...

I think that's dangerous. There are multiple reasons why your model predictions may not quite hit the premium level target in aggregate, even if you go to the (unnecessary) trouble of doing all...
Forum: Property - Casualty / General Insurance 01-22-2019, 09:39 AM
Replies: 32
Views: 2,888
Posted By Heywood J
I think my experience goes much farther than your...

I think my experience goes much farther than your imagination. I can indeed do every single thing you listed, and my conclusion holds.
Forum: Property - Casualty / General Insurance 01-18-2019, 06:45 PM
Replies: 32
Views: 2,888
Posted By Heywood J
In my experience that's only true for small...

In my experience that's only true for small datasets or very noisy lines of business. Above a certain point, there is really nothing you can do to keep up with the boosting methods, at least on...
Forum: Exam 8 - Advanced Ratemaking 01-04-2019, 02:16 PM
Replies: 21
Views: 4,496
Posted By Heywood J
Am I reading this correctly, does it essentially...

Am I reading this correctly, does it essentially say that there are two pass marks for exams where there is a defective question? And if I am reading this right, does it mean that there are 2^n pass...
Forum: Exam 8 - Advanced Ratemaking 12-29-2018, 11:02 PM
Replies: 453
Views: 44,055
Posted By Heywood J
I would say it depends on the source, and depends...

I would say it depends on the source, and depends on your learning style. I personally find it very hard to practice problems without sufficient familiarity with the "color" of material. Study...
Forum: Exam 8 - Advanced Ratemaking 12-27-2018, 04:52 PM
Replies: 3,094
Views: 257,044
Posted By Heywood J
I don't think that's valid reasoning. As long as...

I don't think that's valid reasoning. As long as Z > 0, the recent 10 years will get more weight. Z just determines how much extra weight the recent 10 years get in addition to the equal weight...
Forum: Careers - Employment 12-27-2018, 01:48 PM
Replies: 327
Views: 26,869
Posted By Heywood J
I think this cynicism is a bit too far. There...

I think this cynicism is a bit too far. There are actuaries who are very good at their job, and it's hard to be good at your job if you're not passionate about it.
Forum: Exam 8 - Advanced Ratemaking 12-26-2018, 02:37 PM
Replies: 3,094
Views: 257,044
Posted By Heywood J
It's absolutely CAS's fault for writing a bad...

It's absolutely CAS's fault for writing a bad exam. That's what sucks about the situation: if CAS makes a mistake, their punishment is having to write a letter with some test-taking advice, while...
Forum: Exam 8 - Advanced Ratemaking 12-26-2018, 02:26 PM
Replies: 3,094
Views: 257,044
Posted By Heywood J
To use Exam 8 material for an analogy, you can't...

To use Exam 8 material for an analogy, you can't fix a model with low AUC by tinkering with the threshold, you can only trade one kind of mistake for another by doing that.
Forum: Exam 8 - Advanced Ratemaking 12-26-2018, 01:00 PM
Replies: 3,094
Views: 257,044
Posted By Heywood J
It all depends on which error you're more...

It all depends on which error you're more comfortable with. If you'd rather fail a prepared candidate than pass an unprepared candidate, then bad exams will lead to lower pass ratios. You can't set...
Showing results 1 to 25 of 300

 

All times are GMT -4. The time now is 01:26 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 0.28837 seconds with 7 queries