Re “schools you’ve never heard of”
I hire a lot. Finding ways to interpret educational background of overseas candidate has dramatically improved my hiring results.
Most people put international backgrounds into the “too hard” bucket.
It takes 2 minutes to understand a university more from google. Even better if you can ask someone from that country for advice.
This helps me find undervalued talent that other people won’t hire. Highly recommend.
I guess I’m wondering if it’s better to have GMAT / GRE optional as opposed to required. The biggest problem I see is that these tests don’t really tell you much after a certain score (maybe 600 to 650 range) other than how well someone has managed to game the test. Having taken both, I think it’s especially true of the GMAT which has more of a gaming aspect than any other test I’m familiar with. The downside, as I’ve pointed out, is that it can become rather expensive to acquire high scores with a nice advantage baked in for anyone willing to throw a lot of money at it.
More on the gaming aspect of the GMAT:
People have simulated outcomes using the official practice tests and discovered you can more or less blindly guess on the hardest questions and still get a great score as long as you don’t flub the easy ones. That’s easier said than done of course, but the main takeaway is that accurately identifying difficult questions (to skip them) is more important than answering them correctly, and it’s because the IRT calibration penalizes you much more for missing easy questions. There are numerous threads where people have done this in various ways, but here’s a recent one:
My five scores were: Q49 (74th percentile), Q49 (74th percentile), Q27 (10th percentile), Q31 (14th percentile), and Q49 (74th percentile).
On my first practice test, I intentionally answered difficult questions (as noted by GMAT Club) incorrectly, and this resulted in my Q49 score.
For test #1, you can see that getting only the difficult questions wrong resulted in a solid Quant score of 49. The same can be said about tests #2 and 5, in which most of the incorrect responses were for difficult questions.
https://www.reddit.com/r/GMAT/comments/qc11sg/the_gmat_scoring_algorithm_how_to_get_21/
So he intentionally missed 11/31 questions and still produced a Q49. Knowing this, if I was hypothetically offered a scenario where I could see the model’s difficulty rating for each question with the tradeoff that I’d be forced to randomly guess on 1/3rd of the questions, I’d snap accept. The major thing you gain there is all of the additional time to use on other questions with +2.2 Sklansky correct on the hard ones to boot. Depending on how well you dice roll, I assume you could guess your way into a Q50 pretty often and maybe even a Q51 before your 8 takes are up, all without having to know anything about the hard stuff other than how to identify them.
(For those who aren’t familiar, GMAT quant is 31 questions in 62 minutes. If you’re good at it, you’ll see more of the hard questions because the test is adaptive. Many of the harder questions are time traps unless you’ve memorized some silly trick to solve them easily. There’s an enormous penalty for not finishing a section in excess of simply answering incorrectly.)
I took the GMAT on paper, I believe the last group to do so before they computerized it. I’m glad I didn’t have to worry about any of that shit, just get the answers right and be done with it. I taught Kaplan GMAT test prep for a little bit, too, as a side gig (did like one classroom course and had a few tutoring students), but that was basically just re-teaching adults how to do math.
I actually got really lucky on my timing on two fronts:
I took the GMAT before I graduated college, just in case I applied to MBA within five years. I did apply within five years, thanks to losing my job when the dot com bubble popped, so I didn’t have to retake it.
The school I ended up going to was a top-25 program (or at least borderline at the time, don’t remember), but happened to be trying to game the rankings by increasing their average GMAT score. I had lower than average years of work experience and nothing impressive (PwC consulting yay I’m exactly like everyone else), but my GMAT score was great, so they gave me a fucking scholarship. MBA applications were funny. I either got rejected outright or was given money, no in between.
I suspect that we would approximately never admit an applicant without a standardized test score. On one hand, you could say, “Ok, have that be your informal rule, but let people go ahead and apply without a test score if they want.” On the other hand, the application itself costs the applicant money (the same people who request test waivers often request fee waivers), and it would cost me as well - I’d feel obligated to at least skim through the non-test applications.
So ¯\_(ツ)_/¯
While you’re here, thoughts on this:
That is so interesting. Wonder if this was true 20 years ago when I took it.
I’m going to take the PMP exam in a few months. Will be looking to see if there are similar angles.
If you took it on a computer then yes. Are you familiar with IRT / Rasch models?
I know nothing about this one but am curious now. The only “professional” test I’ve taken is the first actuarial exam (P) which was so different than the college admissions tests that they feel like carnival games by comparison. Exam P is 30 questions in 3 hours and not adaptive, and the recommended study time is typically 300 to 400 hours with a suggested course load of calc I-III, linear algebra, probability theory, and maybe number theory. Compare that to GMAT/GRE which top out at high school algebra in required knowledge but are intentionally tricky with purposefully tight time constraints and weird scoring algorithms. On only one of these can you spend a lot of time (and money) preparing and learn tons of useful stuff in the process.
This actuarial exam sounds legit.
I looked up some “tough” GMAT questions and they really weren’t bad at all. I barely even math any more, and I’m sure I could solve them all. Timing would be my only issue. But that sounds like it might be the only issue for most.
You can see the 1P sample exam here by just clicking right into it. No need to sign up for anything.
You can take free GMAT practice tests too if you create an account on the MBA site although I’m not sure why you’d want to. You’re correct that most of the questions are solvable by a ton of people given enough time and that time is the major constraint. That’s why jijing is such an effective cheating method. Once you crack 700, the amount of effort required to substantially improve your score increases dramatically.
The actuarial exams are no joke. I knew several future actuaries in university and it seemed like they were forever stressed from studying for them on top of their course work.
And yes, that’s exams plural, if memory serves there’s at least half a dozen and each one is harder than the last.
I seem to recall people taking 2FM before 1P based on the idea it was slightly easier. No clue if that is accurate though. Only reason I took 1P is because I had just passed quals so was super fresh on probability theory and figured it was as good a time as any. We had a prof who offered a “course” to prepare for it which was just a few one hour sessions giving heads-up on the type of questions they ask with one of the major points being to expect to pull a deductible out of everything:
Solution:
The nice thing about these is that they don’t expire because there’s no chance in hell I could sit this cold right now and pass it. There are anywhere from 7 to 10 exams depending on which track you choose. I know a few people who’ve gone through most of them and it seemed pretty awful which is why I never seriously pursued it.
Yeah if you don’t have your head buried in it constantly it can start looking an awful lot like alien hieroglyphics. That’s what some of my old grad school work looks like when I pull it out trying to find a note or reference.
I’ve got a question about the question.
Are you supposed to assume that the policy will pay out the amount of damage less the deductible? I’d definitely never assume that in real life.
I suppose so. They obviously aren’t going for realism here given that you’re modeling automobile damages as a uniform distribution. It’s basically just a “can you do this particular math?” question dressed up like an insurance thing.
Re: actuarial science
I can’t imagine why anyone would want to do this. It used to be touted as a high-paying job with the highest job satisfaction when I was an undergrad, and I had several classmates in econ/math that were trying for it. I only took this exam on a whim after finishing my master’s because someone pointed out that all of the problems were things we knew cold from quals. It was basically a backup Plan D and thing to put on my resume.
This was back in 2007. I got a job offer from one of the major health insurance companies within the week. I didn’t even call them, they called me! The pay sucked imo (~$50k), and that’s precisely when I started to question this being a great career. The pay jumps don’t really happen until you hit 5 (~$80k) and then 7+ ($120k+) exams, and it’s anything but a guarantee to get that far. (These are old salaries and I’d hope they’re higher now but wouldn’t be surprised if they aren’t.)
Fast forward to 2021, and now people are saying you need 3 exams, maybe an internship, and high GPA (lol), and/or relevant work experience to have a great shot at that same entry-level job. If you check any of the actuarial forums, it’s basically just a bunch of people bitching about how ultra competitive the job market is and/or how completely miserable the work grind is. Common claims of being underpaid and overworked with work deadlines on top of exam deadlines.
Also, actuarial work is a highly-specialized field with not that many jobs that are heavily concentrated at the relatively few insurance companies, so it’s not like you have tons of options for career switching or geographic location switching. We have a few actuaries here on the forum, and maybe they have a different take having survived this stuff already, but I’m failing to see a good pitch for young people to take this career path.
Additionally, the fact that this has become a highly-competitive labor market looks like a bad sign in general? It always seemed to me like one of the appeals of this career path was that the high bar of passing even one exam was a significant barrier to entry that controlled the applicant pool effectively. Like, just in terms of payoffs / incentives it doesn’t make any sense to me because you’re signing up to do some of the hardest shit possible with no guarantee that you’ll even land the most basic low-paying job. Why are so many people rat racing into that?
Delayed reply, because I didn’t want to type this on a phone:
I technically disagree with it, but I think it ends up being right a lot of the time. Here’s why:
[Important note: I think it’s generally very difficult to predict which applicants are going to be good students/successful faculty. So I’m not actually very confident in my opinion about the attributes of successful candidates.]
Sticking to the accounting/finance/econ world (there’s a lot of overlap, so I’m basically thinking of the group of people who will end up studying some aspect of capital markets), the PhD program has 2 components:
This coursework is very quantitative and requires a good background in math and stats. The classes are more difficult than undergrad, obviously, but the structure is similar - you have exams and projects and milestones and you generally know the things that you need to do in order to successfully complete the class.
It doesn’t require a masters in econ or math, but those would be helpful. And it would certainly be good to have taken (and retained) multiple calculus courses, linear algebra, and probably other classes as well. Generally the first two years of the program, although students will probably take a couple of classes in years three and beyond.
This stage is much less structured in terms of both time and material. You don’t necessarily have milestones, and you definitely don’t have someone making sure that you’re making progress on a day-to-day basis. If you have a good advisor, you have regular meetings set up to make sure that you’re moving towards a dissertation, but this might be an every other week or every month thing.
More important, there’s not an obvious cookbook recipe towards completing the dissertation. You need some degree of creativity to come up with a new research idea.
So with that as background, applicants are typically strong in one of two areas that (imo) correspond to these two stages of the program:
You have academically strong candidates who have taken the hard-core masters courses. These students often do very well in the coursework for the first two years. But they struggle in the dissertation stage, because they have no real world experience. (The typical applicant in this group went straight from undergrad to a masters program, then straight to applying to a PhD program.) So that means if they’re interested in studying something like investment banking activities, IPOs, sell-side analyst behavior, firms’ disclosures, tax rules, SEC enforcement, executive compensation, etc. then they’ve never had the exposure to those things in practice, and have no idea what’s actually interesting or unknown or weird about those areas. So it’s hard for them to identify something worth studying. On top of that, their lack of exposure to these practical areas often lead them to saying things that absolutely make no sense to people who do have experience. (For example, blindly asserting that sell-side analysts are entirely compensated based on the performance of their stock recommendations, and that those recommendations are based on theoretically-grounded discounted cash flow models, using CAPM to estimate a discount rate.)
You have candidates who did not take advanced classes, but instead worked for a while (as a debt or equity analyst, a consultant, an auditor, ectc.). These candidates are in the reverse situtation. They often struggle with the hard-core courses in the first two years, but conditional on surviving those courses are in a better position to complete their dissertation stage, for two reasons. One, like I mentioned before, is that they’re better able to identify interested problems/questions to research. The other is that they’re typically better suited to handle the unstructured nature of the dissertation stage - they’re less likely to need those weekly assignments and milestones. I think this group is likely to end up being better at teaching. But again, a lot of this is unsupported opinion of mine.
So going back to the original question, an undergrad in finance or accounting or whatever isn’t necessarily a red flag on its own. But those undergrad majors often point to people who don’t have the math/econ coursework to do well in the first two years of the program, which is obviously a red flag.
(I was clearly in the second group. I had virtually no advanced math background and was not at all prepared for the doctoral coursework. In fact, I had no idea what the PhD program involved - I thought it just taught you to be a teacher, and that a college professor seemed like a pretty great job. I probably shouldn’t have been admitted into the program at all.)
I don’t see how that’s disrespectful and seems like a good question to ask about any field really. I don’t really have an informed answer though and imagine one of the actuaries here would have a better response.
Some actuarial people I knew used to tell me that certain tracks of actuarial work were mostly solved, particularly life & annuities I think? I’m sure that’s oversimplified and somewhat hyperbolic, but the insinuation was basically that there wasn’t much interesting work left in that field compared to others.
However, actuarial work isn’t just doing math models. It’s math models within the framework of insurance policy, regulatory, and legal stuff that’s probably constantly changing or subject to change. For example, CAS Exam 6U is called “Regulation and Financial Reporting.”
I meant finance bachelor’s degree and nothing else. There are extremely few of that particular profile seated in finance PhD programs aren’t there? Just quickly looking for something that’s top ~50 from UT Dallas sort_by_finance_A_hits I saw this:
They have 13 people with 12 profiles listed. One guy has just the finance bachelor’s and it’s from UF, so I assume the faculty are familiar with him and had confidence he would succeed. Everyone else has engineering / math / statistics / econ and many of those are master’s degrees which is how I expect this shakes out just about everywhere?
