Numericals

[peter333]

Dear David,

I read last year post by Jyoti in editgrid forum and your posts as well. It looks like that you emphasize very much on numericals. I agree that numericals are really very important but when I see e.g. FRM exams (provided by GARP for illustration) they donot have many numericals, i think a ratio of 3 logical or theoretical questions to 1 numerical.

So are these exams provided by representative of actual FRM exam? and what is the ratio of numericals?

Like many course providers emphasize too much on numericals for CFA exams but in the final exam, ratio of such questions turned out to be less than expected.

peter333

 

[David Harper CFA FRM]

Hi Peter,

I do say, "don't underestimate the quant, it counts for a lot more than its formal 10%." I do want to warn/prepare my customers to the quantitative nature of the exam (I disagree with your 3:1 analysis). Last year, over two dozen passing customer thanked me for this emphasis, and none (zero), to my knowledge, said i took the quant too seriously.

But I don't really think too hard about this distinction you speak of, to tell the truth. My emphasis is the list of 550 - 600 AIMs. I feel responsible to personally master every one and to convey every one. The distribution of numerical/non is incidental. The movies embed the AIMs (in order), the notes embed the AIMs, and the new questions that I write are 1:1 to track with the 2008 AIMs. (To focus on this numerical/non numerical distinction, IMO, is the wrong focus. It is like preparing for a beauty pageant by trying to figure who will be the judges. Better to work on your concepts. In working concepts, what difference is this ratio you speak of? I can see course providers, if they entertain this false dichotomy, may be led astray).

If you refer to the early birds, yes, they are free numerical preparation b/c many asked for a quantitative "pre FRM primer." If you refer to the EditGrids, even there i don't even think about numerical/non numerical distinction. The spreadsheets are an excuse to clarify the hardest ideas in a most concrete way; e.g., it is hard to engage qualitative queries around duration or derivatives or even Basel or OpRisk (especially as GARP tilted even more numerically this year with OpRisk) without knowing the "numerical" basis. The advantage of quantitative illustrations is that they do not allow much for fuzzy thinking, which can sometime pass in a qualitative approach. Also, since you raise an interesting point:

* My real focus, for what it's worth: because there are almost 600 AIMS (things to know), I try to bundle and link connected ideas. A disproportionate share of the exam can be reduced to two dozen (+/-) key themes and variants on them. What i am really trying to do here at BT is draw a focus on the reduced number of key themes that will serve well for the exam. So, yes i may seem numerical to harp, for example, on the "first derivative" but (here i go again) look at the number of testable AIMs it subsumes: bond duration, option delta, Liquidity VaR, Jorion's marginal/component VaR, Ong's risk contribution.

* And, I disagree with your analysis of prior exams. (I've been thru all of them) You are treating too many as non-numerical that in fact have a numerical (quant) premise or link or foundation. I would say, easily, more than 50% have a quantitative link/underlying premise. Look at the FRM handbook, most pages are numerical. The more i think about this line (numerical/not) the less i think it counts. Elsewhere on this forum, is our delta discussion numerical or qualitative? I don't know myself!

* You would think OpRisk would be qualitative, but in the last few years, GARP has inserted more quant type material and, unfortunately IMO, not beefed up the case studies. I do wish the exam had more case studies. The FRM seriously lacks in actionable case studies

* True, in the screencasts, i do spend less time on qualitative lists than case study quants because I feel, if your task is to memorize a list, I can be less helpful than somewhere else. Some of the qualitative items, to be candid, is list memorization. In these cases, I try to draw a picture. For example, lots of Basel is non quant (I don't even here like this distinction) so i have a big picture map...

* Finally, I offer this: the thing about risk measurement/management is that often times quantitative understanding is the prequisite for qualitative insights. In many cases, without the quant foundation, you simply cannot speak the language. I am sorry, i have written too much, but consider Basel's IRB for credit. Probably you may count the sample questions here as non-numerical. In my screencast, we will go over all the lists, etc, but at the end i will try to convey a full understanding of the formula, here it is:

http://learn.bionicturtle.com/images/forum/irp_equation_4.png


I submit to you, Peter, for your consideration: this formula impounds (embeds) most of the the "qualitative" notions that underly Basel IRB. In learning/teaching this formula, you must travel a bit of a path and along the way contend with all or almost all of the important "non numerical" issues/trade-offs along the way (and, btw, it employs a copula which was newly this year introduced in John Hull credit derivatives. Another connection, two birds with one stone!). Maybe there are two approaches: Candidate A tries to memorize the lists without understanding the formula. Candidate B is willing to try to understand the formula (and ANYBODY can understand the formula, if they try, with help. If i cannot help you understand this, that's my fault for being a lame teacher). Candidate B will know what candidate A knows and more, being well prepared. Candidate A, perhaps placing too much on exam technique (?), leaves something to chance, and hopes to be quizzed about the stuff he/she happens to memorize.

David

append:
I forgot. I did get this one point here about too much quant: http://forum.bionicturtle.com/viewthread/201/
so, yours is (strictly speaking) the 2nd potential float of this concern.

 

[peter333]

Thanks David,

It makes a lot of things clear to me. Actually, I havenot read BT notes or heard your lecture so far. So my question was not very specific but its nice that you gave very comprehensive comments.

one last point that is perhaps missed: are FRM exams sold by GARP representative of actual exam or they are rather easy?

Thanks again,

Peter

 

[David Harper CFA FRM]

"I have not read BT notes or heard your lecture so far."
LOL. I mistakenly thought you were a customer with more of a fact base

I didn't sit for the last exam, of course, but based on feedback/comments left here: various views. Definitely some candidates felt the samples were not representative.

I can say: the further back you go, the less relevant. On this forum, many from the FRM Handbook were posted for examination/review and, collectively, we found several dubious older questions. Sometimes it is just the topic has changed (e.g., hedge accounting applied when the handbook was published, but is not on the 2008). I would say: stick to the recent samples mostly. David

 

[peter333]

Thanks a lot David.

Actually, I have seen the notes samples and heard one screencast as well. They are actually really very nice. Two weeks back I was wandering from pillar (FRM core readings) to post (Jorion's Book) but now I think you have got a decent stuff.