I saw this too and was thrown off. I don’t think I remember ever seeing it in any of the papers, but I could have also completely missed it. I wonder if it’s from the ERM section of the old exam syllabus??
Can somebody help me out here? I’m doing the opposite of what this problem is doing. The exchange rate is 1.07 $/Euro.
The solution MULTIPLIES euros by 1.07 to get to USD.
Shouldn’t 1.07 dollars per euro mean we divide euros by 1.07 to get USD?
(If I’m breaking a rule by sharing a portion of one problem, tell me and I will remove it. I see no ethical conundrum with sharing a single problem.)
I found this online
“The exchange rate gives the relative value of one currency against another currency. An exchange rate GBP/USD of two, for example, indicates that one pound will buy two U.S. dollars.”
So an exchange rate of USD/Euro = 1.07 means 1 dollar buys 1.07 euros.
So, converting euros to USD, divide by 1.07.
I think the problem is just wrong.
Okay, I think I understand better today. The notation could be better.
USD/Euro = 1.07 means USD is 7% more than Euro.
Someone asked your question on the RF dashboard and the answer is basically as you said, that the conversion rates are inverted.
1 Like
Thank you. I’m just remembering it as though it’s straight division of the value of the currently. 1.07 USD-Euro means USD is 7% more than Euro.
What are people doing to help themselves prepare for the new format questions?
really enjoy how there are only 2 practice questions for Exam 7 and both are free response… Is the “new format” just the inclusion of these free response essay boxes now or am I missing something?
and one of the questions don’t even seem to be on the syllabus.
Yeah, I’m starting to feel like there’s no way I’m going to be able to pass this time around.
Started early, but now it’s less than 2 months away and I don’t feel even remotely prepared.
Plus seeing those two sample questions doesn’t fill me with hope.
so do we think the CAS will go more in depth on problems this time around or just add more topics/questions from what’s left on the syllabus?
I think that won’t ignore any papers like they did in the past.
They might also ask more in depth for some papers that previously did not lend itself to complicated question such as the Mack 94 paper.
If anyone from the exam committee wants to chime in feel free!
When calculating Pearson residuals, what purpose does the scale parameter have? (This is in the Shapland paper.) If it is adjusting the variance so that the triangle is homoscedastic, then how is it different than the hat matrix adjustment?
I’m worried they’re going to ask some very difficult topics like the s.e.(R) of an entire reserve rather than a single year, etc.
I expect to have a few points that are “wow, what BS”, strictly testable, but never done before.
I’m making sure to brush up on the more foreseeable twists like a Brosius caseload question. Anything that’s been tested before but infrequently, or hasn’t been tested but seems quite reasonable to do. Not planning to learn the Mack s.e.(R) calculations. If that happens I’ll just calculate the s.e.(Ri) for each year and add them together, accept that I’ll lose points but maybe get partial credit.
1 Like
Did anybody who already sat for 7 take the new rising fellow exam? How do you think it compares to the actual exam in terms of difficulty and asking on random topics like the CAS often does?
Hi Everyone! I’m going through Brosius and Im kinda stuck at table at page 10 (non - linear example of ultimate and reserves). There is a column Q(x) in the table which shows the expected number of claims incurred given that x claims have been reported.
The values in the column are:
83/31, 256/88, 390/99, 308/64, 156/29, 46/8, 6/1
I recalculated it and got different results:
88/31, 286/88, 390/99, 308/64, 156/29, 46/8, 6/1
This is a small difference, but in the paragraph they describe it as an example of a reserve not being monotonic (it increases between x=1 and x=2), whereas as the result of my calculations it is monotonic (it is kinda intuitive for me that it should be monotonic).
I can’t find any errata, it’s possible that I’m making some mistake, could anyone have a look? thanks
1 Like
In the Hurlimann paper what would happen if the ELR was greater than 100%? That would make the NH Z greater than 1 for certain ages.
The reason why I think that the ELR has to be less than 100% is that if it is greater then the NH estimate will be greater than the Individual and the collective method which seems counterintuitive to the idea of the Hurlimann paper, which is to weight the individual and collective methods.
Would that break the credibility weighted method or would Hurlimann ignore that and proceed as normal?
I would limit Z at 1. I think Neuhaus Reserve would still work then.
My understanding is that by definition of credibility Z must be between [0,1]. The question is how much credibility are we willing to assign to R(ind) (the most we can is full credibility) and the rest goes to the complement: R(coll).
I understand that multiplying Z factor from Benktander by ELR to get to Neuhaus Z is like an additional penaIty, when the ELR is low (like saying the claims are “scarcer”, so we have less faith in R(ind)). When ELR is above 1, the claims are abundant relative to the premium and we can rely on R(ind) more, but the most we can assign is Z=1 (full reliance).
1 Like
TIA Practice Exam 1 definitely assigns Z > 1.0 to Rind, and Z<0 to Rcoll (subtract part of the expected loss.) I expanded the D55 in the (1-D55) to better show.
1 Like