Computer skills, math courses, etc.

[wkh]

Well, it seems as if actuarial recruiters can't stop singing the praises of SAS/APL (not Perl, Java, C/C++), Microsoft Access (not Oracle, MySQL, etc), etc. I'm also surprised that I haven't seen anyone mention Linux/UNIX even once.

Could a lack of skills in something obscure like SAS be mitigated by demonstrated ability to learn new things quickly? A lot of what someone in IT is most familiar with seems to not even transfer over to actuarial work.

Also, it seems that a math degree is often desired. If actuarial work isn't that mathematical (in the sense that advanced math is not used), does it matter what courses one takes? In my view, it would be developing formal reasoning skills that'd be more important than knowledge of any one area. What'd be wrong with, say, taking a year of algebra, a year of analysis, some number theory classes, and a couple topology courses (and having this be one's mathematical education)? That is, if in fact relatively little math (lower-division calculus, some basic statistics, some basic probability, etc.) is used--it's all the same, right?

[DW Simpson]

Actuarial employers sing those praises. Actuarial recruiters are the disk jockeys.

Originally Posted by wkh

Well, it seems as if actuarial recruiters can't stop singing the praises of SAS/APL (not Perl, Java, C/C++), Microsoft Access (not Oracle, MySQL, etc), etc. I'm also surprised that I haven't seen anyone mention Linux/UNIX even once.

Perl/Java/C/C++ would be used by someone in the IT department to do low level manipulation. Typically as an actuary your manipulation of the data will be at a higher level. In other words, you would normally be expected to know the tools to manipulate data, not develope new software packages. Perl/Java/C aren't the best tools for doing this type of work.

Same thing with Access. It's a step up from Excel but still typically a tool for the end user. And it's an MS product, the platform that everyone in this industry uses.

As for Linux/Unix, if an employer has data running on these platforms it would normally be an IT person who would grab the data and convert it to an access db for you. Linux/Unix is again the bailiwick of the IT person, you wouldn't normally be expected to be proficient in any OS other than Windoze. (and most employers will likely be running an MS system to handle their user network. They might have a *nix system somewhere, but it would be buried and again, not accessed by users).

Typically your job would be to manipulate the data and form conclusions. The actual gathering of the data into a format that can be manipulated is left to IT, whose job it is to do this.

[Westley]

Originally Posted by wkh

Could a lack of skills in something obscure like SAS be mitigated by demonstrated ability to learn new things quickly? A lot of what someone in IT is most familiar with seems to not even transfer over to actuarial work.

Yes, there is little that transfers directly over, but if you have a lot of IT skills and show that you can learn new stuff, that's good enough for most employers

[Westley]

Originally Posted by wkh

Also, it seems that a math degree is often desired. If actuarial work isn't that mathematical (in the sense that advanced math is not used), does it matter what courses one takes? In my view, it would be developing formal reasoning skills that'd be more important than knowledge of any one area. What'd be wrong with, say, taking a year of algebra, a year of analysis, some number theory classes, and a couple topology courses (and having this be one's mathematical education)? That is, if in fact relatively little math (lower-division calculus, some basic statistics, some basic probability, etc.) is used--it's all the same, right?

number theory, topology - these are way past what an actuary needs. If you want to take them, fine, but not necessary.

"math degree is often desired" - ask DTNF about this. Oh, heck, I'll tell you: pass an exam or two, then give me a call.

[Buru Buru]

Okay, not to sound ignorant, but what is SAS/APL? I don't think that I have ever used them.

Fortran has not been mentioned. I use Fortran all the time.

[Westley]

SAS is (was?) a pretty standard academic statistics program. Basically (from what I recall) it was designed to do a lot of statistics functions (regressions, ANOVA, r-squared, etc) really quickly and easily. APL is for use with huge data sets on a mainframe? SAS is pretty commonly used, I thought, outside of actuarial. APL is archaic.

[SamChevre]

SAS--Statistical Analysis System. It's a combination of a database with a ton of powerhouse statistical tools--it's commonly used in academics (we used it a lot in econometrics courses when I was in school) and in business, where it's used for data-mining and statistical work.

APL--A Programming Language. This is probably the original language for mathematical work on computers--I think it was written in the 1950's. It was the standard on mainframes in the late 1960's. It is very efficient for storage and easy to program, but programs aren't real easy to read--the symbol-set is around 100. It is vector-based, so matrices and matrix algebra are simple. It seems to be common in actuarial and engineering work. I use it daily and am a big fan of it.

[wkh]

Originally Posted by Westley

number theory, topology - these are way past what an actuary needs. If you want to take them, fine, but not necessary.

clearly those aren't necessary. they were just examples. if by chance, say, one has a particularly unpleasant internship and decides the actuarial profession is not the way to go, then grad school is still an option -- generally the minimum preparation needed is a year of algebra and analysis (differential equations, linear algebra, etc. would probably be better than topology, but whatever).

and doing things that way doesn't constitute making an actuary career a backup plan, it means having a backup plan in case actuary work doesn't work out.