I’m going to pull out some quotes from the document – this relates to the comment letters from the earlier draft.
Area of practice—Several commentators said it was unclear what was meant by “area of
practice,” which is referenced in several sections of the current USQS. The COQ notes that
section 1.4 states, “The traditional areas of actuarial practice are casualty, health, life, and
pension.” This language appears in both the 2008 USQS and this exposure draft. The lack of
clarity relates to references or circumstances when the same terminology “area of practice” was
used to refer to an area that may be more specific than these four practice areas. In response to
the comments, references to “area of practice” in sections 2.1(d), 2.1(d)(1), 2.1(d)(2), 2.1(d)(3),
and 2.1.1(b) where the COQ recognized more specificity was needed, were revised.
Specialty tracks and examinations—The COQ revised the language to clarify that section
2.1(d) sets qualifications for issuing SAOs “on a particular subject within an area of actuarial
practice,” rather than “an area covered by a specialty track offered by the Society of Actuaries or
in an area of practice covered by the exams for the Casualty Actuarial Society or the American
Society of Pension Professionals and Actuaries.” The goal was to keep this section of the U.S.
Qualification Standards clearly focused on the basic education component in a way that was both
sufficiently particular about basic education as well as sufficiently general so as not to become
outdated solely because of a change in the name of a specialty track or the addition of a
particular specialty track offered by a basic education provider. In addition, the applicability of
this section to actuaries with CAS credentials was unclear because the CAS does not currently
offer specialty tracks.
Diversity, Equity, and Inclusion (DE&I)—Many commentators requested that the USQS
clarify that professionalism topics include DE&I, and several of those commentators suggested
that DE&I continuing education be required. The COQ considered all these comments and
agreed with the need to clarify that some DE&I topics could be considered professionalism CE.
Thus, Frequently Asked Question #48 has been updated to clarify that DE&I topics may be
considered either general business skills CE or professionalism CE depending upon the content
of the particular CE offering. This change immediately addresses some of the comments received,
with the understanding that the FAQ may be further amended if appropriate upon final approval
of an updated USQS.
New Requirement for CE related to Bias Topics—This second exposure draft includes a new
one-hour minimum CE requirement dedicated to what is referred to as “bias topics.” Section
2.2.6 specifies that bias topics “include content that provides knowledge and perspective that
assists in identifying and addressing biases that may exist in data, assumptions, algorithms, and
models that impact Actuarial Services. Biases may include but are not limited to statistical,
cognitive, and social biases.” The COQ included bias topics in the CE requirements to focus on a
topic that provides knowledge and perspective that is relevant to the Actuarial Services that form
the basis of an SAO.
Should be interesting to see what people will come up w/ for bias-related CE.
Of course, in the statistical sense, a biased estimator is one for which the expected value differs from the value of the thing you’re trying to estimate.
Wasting a chance to get some CE!
Well, there’s that.
I will give you a TOC, so you can decide what to read:
pp 2-6 – background & summary of comment letters from first draft
6-10 – changes made in this draft from prior draft
11 – committee members
12-13: intro to new QS
14-19: general QS (including CE requirement)
20-21: specific QS
22: change in practice/field
23: acknowledgement statement
24: recordkeeping
25-30: Appendix 1 - examples of SAOs
31: Appendix 2 - alternative basic qual
32-36: App 3 - procedures for adopting new QS
37-38: App 4 - procedures to determining specific QS needed
39: App 5 - recordkeeping form
The response letters to the first draft can be found here, just to pop out this link:
https://www.actuary.org/node/13857
I went to a session last year about bias in measuring expected losses, and how the "red lines"of the past affect today’s rates. It was quite good. I’m sure there’s more where that came from.
Agreed. These seem like mostly modest changes that improve the US qualification standards.
comment period is ending soon
via email…
Amended Qualification Standards for Actuaries Issuing Statements of Actuarial Opinion in the United States Released
| To: | Members of Actuarial Organizations Governed by the Qualification Standards of the American Academy of Actuaries | |
|---|---|---|
| — | — | |
| From: | The Board of Directors of the American Academy of Actuaries & The Committee on Qualifications of the American Academy of Actuaries | |
| Re: | Qualification Standards for Actuaries Issuing Statements of Actuarial Opinion in the United States |
The amended Qualification Standards for Actuaries Issuing Statements of Actuarial Opinion in the United States (U.S. Qualification Standards or USQS) promulgated by the American Academy of Actuaries (the Academy) are now available. This document supersedes in its entirety the Qualification Standards (including Continuing Education Requirements) for Actuaries Issuing Statements of Actuarial Opinion in the United States that took effect January 1, 2008. The amended U.S. Qualification Standards take effect January 1, 2022, indicating that this document applies to actuaries issuing statements of actuarial opinion starting on January 1, 2023, and that such actuaries will need to meet the continuing education (CE) requirements before issuing any Statement of Actuarial Opinion (SAO) in 2023.
The Academy will host a webinar on Dec. 20 at noon EST to discuss the amended standards and the Committee on Qualifications is working to update the Frequently Asked Questions. Registration for the webinar will open soon.|
|Having trouble viewing this post? View it in another browser tab or window, depending on how you have your settings set up.|
I saw the 1 hour bias topics requirement made it in — any other major changes or just the stuff around generalizing the definition of qualified actuary?
re: bias topics – from the doc
b. Bias topics include content that provides knowledge and perspective that assist in
identifying and assessing biases that may exist in data, assumptions, algorithms,
and models that impact Actuarial Services. Biases may include but are not limited
to statistical, cognitive, and social biases.
Statistical bias means when the expected value of a statistical estimator is not the “real” value of what you’re trying to estimate.
Bias means a lot of different things, depending on context.
Ok, reading through, the bias topics looks like the only major change.
CE Requirements are:
30 total CE hours per year
At least 6 credits in organized activities
At least 1 in bias topics (which I think is a subset of professionalism?)
At least 3 in professionalism
At most 3 in business
FWIW, if you wanted to, you could have all 30 CE hours in professionalism in organized activities on bias topics, and that would fulfill requirements.
Unless you are signing the capital SAO, then the specific requirements apply and you can’t get away with that I think.
I tend to think this confusion leads to real errors.
Bias in the statistical sense is a practically inevitable consequence of learning. It can be good or bad depending on the environment.
Bias in the ordinary sense of the word is always bad.
So as I understand the words, statistical bias that is an inevitable consequence of the finite nature of our brains can also lead to racial bias in our everyday lives or in our conceptual understanding of things. Different biases also allow us to turn 2D images into 3 dimensional concepts, and help us immensely.
Its not clear to me how this learning objective distinguishes between the two. For example, random effects models or actuarial credibility have classical statistical bias built into them from a particular perspective. Does studying these meet this objective? Probably not by intent, but maybe so by wording.
Gah. I really wish they had a minimum in “specific” topics, and not a maximum in “business”. I just feel like it gives the wrong message that there’s a max, when many actuaries would be more effective with more business training. Sure, you shouldn’t have only business, but this just feels like the wrong way to construct the requirements.
What specific topics though? These are requirements for the entire profession, not just life, or health, or P&C.
No, statistical bias has a very specific mathematical definition, which has nothing to do with human brains.
For example, the sample mean is an unbiased estimator for the actual population mean. There’s no statistical bias.
The sample median might be a biased estimator for the population mean, if the actual distribution is not symmetric.
I could do another (free) CE video, but I’ll do it later, and y’all can learn what statistical bias means if you don’t already know what it means. Statistical estimators are random variables… yadda yadda yadda.
I don’t need to throw this thread off course re: what has changed for those of us who have satisfied basic education requirements and need to think about what has changed for continuing education requirements.
To be blunt, I get far more than 30 CE “hours” each year. I could satisfy the requirement, other than the “organized” aspect, within January.
If I were really nasty, I would just disallow this “business” CE. Make it all actuarial-related.
Hell, make it all professionalism.
Make it all ASOP 56.
Now we’re talking.
Right, but use of imperfect models tend to lead to biases in that sense. And incorporation of prior information does too, which is most easily seen when bayesian priors on parameters lead to estimates that are biased in the classical sense.
The only time i’m aware that biases don’t exist in real world measurements is when you have the right scientific theory, and then make unbiased estimate of parameters in that theory. And even then true unbiasedness is only a kind of ideal limit. But our finite brains cannot formulate or evaluate such a theory in our everyday lives, even if such a theory does exist. So we use biased approximations instead.