Diebold, Chapter 5, 7 and 8

[Dr. Jayanthi Sankaran]

Hi David,

I notice that while there are Study Notes encompassing Diebold, Chapters 5, 7 and 8, there are no corresponding PQ sets.

Does that mean that this material is not that relevant or that you are working towards producing a PQ set...Please let me know

Thanks!
Jayanthi

 

[Nicole Seaman]

@Jayanthi Sankaran,

If you look at the Daily Practice Questions in the forum, you can see that we do not yet have a complete set for Diebold. Once we have posted all of the the practice questions in the daily PQ thread, then we publish the complete set in the study planner.

Thanks,

Nicole

 

[Dr. Jayanthi Sankaran]

Thanks, Nicole - appreciate it

Jayanthi

 

[lRRAngle]

Hi there - How testable are the Diebold readings, I just watched the video on chapters 7&8 and David said the material has low testability, but what about chapters 5&6?

 

[emilioalzamora1]

Hi @lRRAngle

basically everything is testable based on my experience. Just do not make any compromises and disregard some topics/chapters completely. Even footnotes (in Dowd's book) have already been tested when I sat for the exam.

Chapter 5 in Diebold is highly relevant: especially selection criteria (AIC, Schwarz).

See my post about selection criteria here a few weeks ago:
https://forum.bionicturtle.com/threads/p1-t2-505-model-selection-criteria-diebold.8381/#post-58838

Chapter 6 also but perhaps to minor extent! less questions than from Chapter 6,7,8.

You can be rest assured that there will be questions about white noise and autocorrel (also calculation questions!).

Hope that helps.

 

[lRRAngle]

Thanks @emilioalzamora

 

[valenski]

Regarding Diebold EOC questions: I am now covering topic about Diebold’s chapter 5-8 and wonder if it is necessary to do these questions which seems different from real FRM question.. THX

 

[Nicole Seaman]

Regarding Diebold EOC questions: I am now covering topic about Diebold’s chapter 5-8 and wonder if it is necessary to do these questions which seems different from real FRM question.. THX

Hello @valenski

Please note that I moved your question here to this thread that discusses the Diebold reading. We include the EOC questions in our practice question sets to give our members more questions to study from. It is really up to you if you want (or have time to) cover those EOC questions. They would be helpful if you feel that you need more practice on the concepts covered in this reading, but not everyone chooses to complete them.

 

[valenski]

I

Hello @valenski

Please note that I moved your question here to this thread that discusses the Diebold reading. We include the EOC questions in our practice question sets to give our members more questions to study from. It is really up to you if you want (or have time to) cover those EOC questions. They would be helpful if you feel that you need more practice on the concepts covered in this reading, but not everyone chooses to complete them.

Regarding to Diebold’s chapters I did not see any EOC included in the PQ...... Can you double check it? You do include the EOC of Hull in the PQ though.

 

[Nicole Seaman]

I


Regarding to Diebold’s chapters I did not see any EOC included in the PQ...... Can you double check it? You do include the EOC of Hull in the PQ though.

Hello @valenski

We do not always include the EOC questions. I didn't check the Diebold PQ set when I responded to your post. I only meant that if they are published in the PQ set, it is not because they are required. It is for extra practice for our subscribers. This is true for the EOC questions in any of the readings. They are very helpful if you feel that you need more practice with certain concepts.

 

[David Harper CFA FRM]

Hi @valenski Diebold's EOC questions are atypical (sui generis) for syllabus reading assignments in the FRM. Although we do have the solutions manual, Diebold's EOC questions are difficult--often very difficult--and lab-type such they often require working with a dataset. Many of them go beyond--or far beyond--anything that could be asked on the exam, where the average response time cannot be much more than ~ 3.0 minutes per question (in Part 1, the average is only 2.4 min per question). Some of Diebold's EOC question require research and might take 30 minutes or more than an hour. So I just did not want to "bog down" candidates with such ponderous exercises. Frankly, the Diebold readings are (in my opinion) a bit "inconvenient" in the syllabus: ideally they require better "scaffolding" (aka, setup). As I have argued before (and fed back specifically and concretely to GARP), the assigned Diebold almost requires the reader to read all previous Chapters in Diebold. Some of the EOC questions are similarly this way, they require mastery beyond the syllabus assignment. In this way to be totally candid, I also do not expect Diebold to "survive" the syllabus long-run. In the meantime, I've written complete question sets for Diebold that I'm proud of: i think they are exam-representative and better than Diebold's EOC, at least in terms of the exam (don't get me wrong, time series is an important topic). Eventually, I predict that a more congruent assignment will be installed. I hope that's helpful as an explanation as to why specifically I decided to omit Deibold's EOC questions. Thanks,

 

[valenski]

Hi @valenski Diebold's EOC questions are atypical (sui generis) for syllabus reading assignments in the FRM. Although we do have the solutions manual, Diebold's EOC questions are difficult--often very difficult--and lab-type such they often require working with a dataset. Many of them go beyond--or far beyond--anything that could be asked on the exam, where the average response time cannot be much more than ~ 3.0 minutes per question (in Part 1, the average is only 2.4 min per question). Some of Diebold's EOC question require research and might take 30 minutes or more than an hour. So I just did not want to "bog down" candidates with such ponderous exercises. Frankly, the Diebold readings are (in my opinion) a bit "inconvenient" in the syllabus: ideally they require better "scaffolding" (aka, setup). As I have argued before (and fed back specifically and concretely to GARP), the assigned Diebold almost requires the reader to read all previous Chapters in Diebold. Some of the EOC questions are similarly this way, they require mastery beyond the syllabus assignment. In this way to be totally candid, I also do not expect Diebold to "survive" the syllabus long-run. In the meantime, I've written complete question sets for Diebold that I'm proud of: i think they are exam-representative and better than Diebold's EOC, at least in terms of the exam (don't get me wrong, time series is an important topic). Eventually, I predict that a more congruent assignment will be installed. I hope that's helpful as an explanation as to why specifically I decided to omit Deibold's EOC questions. Thanks,

Hi Daivd

Thanks so much for resolving my question. I am reading the core reading of Diebold’s Chapter but to be frank I could not understand the text very well. So I watch your video about these 4 chapters but still feel weird with the topic of forecasting and time series. It seems like your video covers only a little about the whole topic. I wonder if there is any other textbooks or YouTube video of you to access this topic. You mentioned “complete question sets for Diebold”. Does this mean the PQ? Btw Can I have the full solution manual of Diebold?
Thx very much!

 

[David Harper CFA FRM]

Hi @valenski I don't have any great textbooks in mind to "replace" Diebold as a introduction; in my estimation, if a holistic introduction is desired, then a fine thing to do is simply start with Diebold Chapter 1 and continue (it's not a terrible book on the topic. In my opinion, the disadvantage of this approach is that you are forced to use eViews). The practical issue is that this topic (time series) really needs to be introduced in the context of code, in addition to the mathematical prerequisites which frankly lack scaffolding in the FRM syllabus: time-series is inherently about operating on time-series data and it's weird to try and communicate time series without any reference to code/software. Deibold uses eViews, which is not my favorite choice for students who want realistic exposure, I think it probably should be learned with python or rstats (aka "R") simply because that's what the rest of the world seems to be using.

I'm sure there is probably a good book that introduces time series in the context of Excel, i just have never pursued that angle sorry. Here is a great blog and they often review time series https://www.analyticsvidhya.com/blog/2019/01/introduction-time-series-classification/

Re: complete question sets: yes, I was referring to our (BT's) practice question sets for Diebold. I am proud of them, I think they cover the LO's "completely" at an exam-level type difficulty.

Re: Can I have the full solution manual of Diebold?
Let me see if we can add it to the Resources. Currently Resources has the book but let me look into sharing the solutions to Resources. BTW, in my recollection, they won't be maybe as helpful as you would expect, consistent with the lab-type approach in the book. Many of the answers are terse (e.g., "go do this"); in my recollection, it's an underwhelming solutions manual.

Thanks!

 

[valenski]

Thank very much! Can I have the link to the Diebold textbook resource? I couldn’t find the by myself.

Hi @valenski I don't have any great textbooks in mind to "replace" Diebold as a introduction; in my estimation, if a holistic introduction is desired, then a fine thing to do is simply start with Diebold Chapter 1 and continue (it's not a terrible book on the topic. In my opinion, the disadvantage of this approach is that you are forced to use eViews). The practical issue is that this topic (time series) really needs to be introduced in the context of code, in addition to the mathematical prerequisites which frankly lack scaffolding in the FRM syllabus: time-series is inherently about operating on time-series data and it's weird to try and communicate time series without any reference to code/software. Deibold uses eViews, which is not my favorite choice for students who want realistic exposure, I think it probably should be learned with python or rstats (aka "R") simply because that's what the rest of the world seems to be using.

I'm sure there is probably a good book that introduces time series in the context of Excel, i just have never pursued that angle sorry. Here is a great blog and they often review time series https://www.analyticsvidhya.com/blog/2019/01/introduction-time-series-classification/

Re: complete question sets: yes, I was referring to our (BT's) practice question sets for Diebold. I am proud of them, I think they cover the LO's "completely" at an exam-level type difficulty.

Re: Can I have the full solution manual of Diebold?
Let me see if we can add it to the Resources. Currently Resources has the book but let me look into sharing the solutions to Resources. BTW, in my recollection, they won't be maybe as helpful as you would expect, consistent with the lab-type approach in the book. Many of the answers are terse (e.g., "go do this"); in my recollection, it's an underwhelming solutions manual.

Thanks!

 

[David Harper CFA FRM]

Hi @valenski Our Resources are organized by topic, so a PDF copy of Diebold (not the solutions, yet ...) is located in T2 @ https://forum.bionicturtle.com/resources/categories/frm-p1-t2-quant.4/ ... and specifically https://forum.bionicturtle.com/resources/elements-of-forecasting.109/ I hope that's helpful!