Course 4: Order of Reading?

[bg23516]

In the event that you are using the texts and the study notes as reading material for Course 4, where do you start?

Many of us here are probably working, and most companies won't pay for the next set of study materials until they see a pass slip. That said, many of us here are fairly confident in passing course 3, and I personally would like to get a slow start on Course 4, merely using the books I already have, as well as the study notes freely available online.

So, I have Simulation, Loss Models, and the Credibility Study Notes. Which do I start with? Loss models seems to jump right into difficult topics assuming an understanding of estimation. We are told to read only 5.4 & 5.5, yet they seem to build on concepts in 5.1-5.3. Though I have some background in estimation from college stats (MLE, MME, RCLB, Fisher information, etc.) I'm not sure I want to go re-learning concepts in Ch5 if they aren't necessary.

Any recommendations?

[Anonymous]

I am starting with the intro credibility note from Philbrick, and then jumping into a co-worker's 4th edition of ASM. I may read the credibility study note first?[/u]

[bg23516]

I actually got a copy of the econometrics text from a coworker, so I think I'm going to start with that, after the intro to creibility note. I have a fair background in econometrics, so this should get my mind working.

[StephenLL]

Wait for the new studynote to come out before attacking the loss models piece. I know several people that were editing it and they promise it is a great improvement over the loss models book and the survival analysis book.

Personally, I have decided to attack the material in the following order:
1) Econometrics (while reading the background material for credibility)
2) Credibility
3) Loss Models/Survival Analysis
4) Simulation

Hit the hard or not well written stuff first (1 & 2) and tackle the well written/easier or less tested topics second (3 & 4).

Just my 2 cents.

Stephen

In the event that you are using the texts and the study notes as reading material for Course 4, where do you start?

Many of us here are probably working, and most companies won't pay for the next set of study materials until they see a pass slip. That said, many of us here are fairly confident in passing course 3, and I personally would like to get a slow start on Course 4, merely using the books I already have, as well as the study notes freely available online.

So, I have Simulation, Loss Models, and the Credibility Study Notes. Which do I start with? Loss models seems to jump right into difficult topics assuming an understanding of estimation. We are told to read only 5.4 & 5.5, yet they seem to build on concepts in 5.1-5.3. Though I have some background in estimation from college stats (MLE, MME, RCLB, Fisher information, etc.) I'm not sure I want to go re-learning concepts in Ch5 if they aren't necessary.

Any recommendations?

[Itsasecret]

I usually don't read the texbooks. I am planning on using the ASM for Course 4. I want to get an early start, but Actex says that my ASM won't ship until December 20th. I can borrow a co-workers manual from the May 2001 sitting. Does anyone know which material from that manual is "least likely to change"?

[New at pd]

ASM is all you need for this exam. It sounds like the Survival Analysis material has been dropped from the syllabus, wiping out sections (~) 46-56 of ASM.

I'd go with regression, credibility, loss models, time series, simulation

good luck

[glenn]

The Survival Analysis hasn't been dropped, it's been changed. Instead of the old yellow SA book, there will be a new study note by Klugman on the subject (not yet available at the SOA website as of yesterday).

[New at pd]

my bad -- i really don't care to think of that exam b/c it has had absolutely no contribution to my knowledge, just my paycheck.

[glenn]

:lolup:

[Girl]

All of the earlier section of the loss model book have been dropped? Like Sections in chapter 2 and 3 or am I missing something?

[StephenLL]

Sections 2 and 3 are now covered in the new study note.

Here is the table of contents from the new studynote as of Nov 1. There are 142 pages in total. I was given this as a reference. Unfortunatly I do not have the actual studynote.

Contents
2 Model estimation 3
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.2 Estimation using data-dependent distributions . . . . . . . . . . . . . . . . . . . . . 6
2.2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2.2 The empirical distribution for complete, individual data . . . . . . . . . . . . 6
2.2.3 Empirical distributions for grouped data . . . . . . . . . . . . . . . . . . . . . 9
2.2.4 Empirical distributions for censored and truncated data . . . . . . . . . . . . 12
2.2.5 Kernel smoothing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.2.6 Estimation of related quantities . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.3 Estimation for parametric models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.3.2 Method ofmoments and percentile matching . . . . . . . . . . . . . . . . . . 23
2.3.3 Maximumlikelihood . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3 Sampling properties of estimators 37
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.2 Measures of quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3.2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3.2.2 Unbiasedness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3.2.3 Consistency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.2.4 Mean squared error . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
3.3 Variance and confidence intervals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
3.3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
3.3.2 Methods for empirical distributions . . . . . . . . . . . . . . . . . . . . . . . . 44
3.3.3 Informationmatrix and the deltamethod . . . . . . . . . . . . . . . . . . . . 54
4 Model evaluation and selection 61
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
4.2 A review of hypothesis tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
4.3 Representations of the data andmodel . . . . . . . . . . . . . . . . . . . . . . . . . . 66
4.4 Graphical comparison of the density and distribution functions . . . . . . . . . . . . 67
4.5 Hypothesis tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
4.5.1 Kolmogorov-Smirnov test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
4.5.2 Anderson-Darling test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
4.5.3 Chi-square goodness-of-fit test . . . . . . . . . . . . . . . . . . . . . . . . . . 77
4.6 Selecting amodel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
4.6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
4.6.2 Judgment-based approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
4.6.3 Score-based approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
5 Models with covariates 85
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
5.2 Proportional hazards models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
5.3 The generalized linearmodel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Appendix
A Solutions to Exercises 95
B Using Microsoft ExcelTM 131

All of the earlier section of the loss model book have been dropped? Like Sections in chapter 2 and 3 or am I missing something?

[Michael]

I'm intrigued by the last part -- using Microsoft Excel -- curious what that will teach us....anyone know of readings specifically geared towards using Excel in Actuarial applications?

[StephenLL]

Some of the proceedings, which discuss the use of fast fourier transforms in aggregate distributions, give a "real world" example using excel. I believe Excel's solver comes up once in a while too.

I'm sure it won't be tested.

I'm intrigued by the last part -- using Microsoft Excel -- curious what that will teach us....anyone know of readings specifically geared towards using Excel in Actuarial applications?

[glenn]

Just curious - who wrote this study note?

[aces219]

Stuart Klugman

[glenn]

Duh. I guess I knew that. For some dumb reason I thought this was a second new study note. :duh: