Powered by glolg
Display Preferences Most Recent Entries Chatterbox Blog Links Site Statistics Category Tags About Me, Myself and Gilbert XML RSS Feed
Saturday, Oct 25, 2008 - 18:59 SGT
Posted By: Gilbert

Eating Words

"When people call you an asshole, (technically) they are correct - since your asshole formed first as an embryo."

- LSM1301 lecturer on deuterostomes


"NUS Fights Climate Change"

- banner at entrance of Arts Canteen, next to LT9

"Buy Tiger Tooth Necklaces"

- temporary stall outside LT9, next to Arts Canteen


Yum yum.

It turns out that a person is quite often not wholly correct in his assertions, unless of course (s)he is a politician, in which case (s)he is always right and the response is to deny everything and sling mud at opponents. Or (some) teachers, in which case (s)he was only testing if you're alert (however I can sympathize with the [sometimes undeserved] lack of respect [some] teachers get). Or...

Okay, let's move on.


Puzzlecrack 2008

Team FourOKOJ/C.fckbsds solved all four daily puzzles in this year's edition of Puzzlecrack, sadly not fast enough to claim any prizes. But hey, it's for the challenge!

Keeping on theme, I can't count the number of times we went down the wrong track here, although that's only to be expected. Did most of the stuff for Monday, helped a bit on Tuesday (thanks to some nice tools), next to nothing on Wednesday (major credit: twm) and finally solved Thursday by myself (this is not saying much, since the winning team for Thursday settled it in seven minutes - it may be a good one to try if you're new though).

As team member C.Wenhoo said, "breadthwise search power increases exponentially with number of team members", so I guess we could be in with a shout with more (i) organization and (ii) participation. I'll see what I can do about a one-input decrypt-many-ways script one of these days.

Puzzlecrack worryingly resembled my HYP, which is my major concern after this semester's mid-term exam results were probably the best of all my semesters so far (again, that's not saying much - the marks from CS4213 reinforcing the impression that one can score well without actually knowing much in detail). Actually, research is probably worse, since for Puzzlecrack you know that there surely exists a solution, since somebody (sadistic) set the problem up. And this brings us to...


Grad School

Sought the opinion of my main mentor-professor, but didn't get any earth-shattering insights that could tip my decision firmly one way or the other. There seems to be a consensus that it would only be surely worth it to forsake NUS for a top-10+ ranked CS department in the USA, though, with the distinction in quality of education blurring between 10+ and 20+, with NUS probably a better choice once we hit the low twenties and thirties (which seems reasonable considering stuff like the Engineering & IT rankings from THES [criticism], since direct comparism between grad schools internationally appear hard to find).

ProsCons
NUS
  • Home comforts
  • Can save close to S$100k on stipend alone, over four years
  • Very high probability of acceptance, no worries (before starting the program anyway)
  • Average time needed probably only 4 years or so (though rumour has it that at least one guy took 8[!] years)
  • Quality of education not that bad. Really
    		•
  • The very top grad schools in the USA unequivocally better, in general (come on, lah)
  • NUS "brand name" not that developed yet, THES rankings notwithstanding
  • Low exposure to other (new) research cultures
  • Difficult to get a position at NUS in future if need be (universities usually don't like to hire their own graduates)
    • USA
  • It's the cutting edge of the world, baby!
  • Unparalled global recognition
  • It's where the top research jobs are at
  • See the world, adventure
    		•
  • Have to get used to whole new environment with little support
  • Unlikely to save much if anything, since stipends in the USA seem about the same as at NUS, with many personal expenses that could otherwise be avoided in Singapore
  • Realistically will take 5+ years in the American system
  • Hardly easy to get in (with funding!), in the first place

    • Actually, I'm teetering towards NUS (as I don't see myself as a professor-type, for now anyway). With each application to a USA grad school costing upwards of a hundred bucks, just shooting them off sounds rather unattractive... so, spend the cash, or eat the words?


    More on Minibonds

    Last month, I expressed the opinion that the general public/stakeholders should not be made to indirectly subsidize (i.e. pay) for the mistakes made by particular individuals who went in search of higher returns (but got unlucky).

    Well, it turns out that some who got caught up in Lehman-linked investments were old uncles and aunties, who with all due respect aren't the sort whom would be comfortable calculating portfolio risk variances. Quite a few of them were going for the tried-and-tested zero-risk (well, if the bank doesn't collapse) fixed deposits, but were persuaded into "very safe" structured investments instead.

    It turns out that some of them may have been badly advised by their relationship manager, and thus probably should get something back (though the worst-case scenario remains to get nothing). Taking DBS as an example, as they expect to pay out $70-80 million in compensation to investors who put in about $360 million, it seems safe to say that for each fully-compensated person, four will go empty-handed. Quite clearly these likely younger/better-educated persons aren't too happy at losing out to the "vulnerable" group, with some among them claiming to be quite as "ignorant" too.

    Now, such investments are inherently risky to an extent - the fact that it offers a higher interest rate illustrates that by the universal No Free Lunch rule, and really, investors should be expected to know that at the very least. Of course, if a relationship manager explicitly told an 80 year-old granny that there was "absolutely no chance" of default, then it's the bank's fault.

    Continuing on to investors who recognized the existence of risk - did they bother to ask about the worst-case scenario, and did the relationship manager volunteer this information (probably not)? Opinions can differ on this, but really I feel that anybody who is going to entrust a substantial portion of his wealth to a financial institution should at least have taken the responsibility to ask that one question (and also weighed the chance of the institution failing).

    The alternative is of course to quietly collect the higher returns (as doubtless some structured products have managed to give) if things go well, and make a lot of noise and demand one's money back when things go wrong. Obviously if authorities cave in each time this happens, there would be no downside at all to buying risky investments - heads I win, tails I can't lose. Therefore, in principle and in general, such bailouts should not be (and in reality usually are not) done.

    The solution, as often happens, is knowledge. As I have suggested, the worst-case scenario for any financial product could be printed prominently at the top of each contract or brochure (like those lung cancer pics on cigarette boxes). If mandated by the government to all banks, no bank would be worse off under this regulation.

    Yes, a bank might argue that hey, though the worst-case is to lose everything, we calculate that there's only a 0.1% chance of it happening, so making us reveal this makes us look unfairly bad. To this my answer is that if the risk is that low, why not have your quant wizards work their mathematical magic and whip up a product which guarantees a minimum of like 90% (backed by the bank) in the worst case, by reducing returns when good things happen?

    In effect, this shifts the work of portfolio balancing from clients to banks, eliminating doomsday outcomes, while leaving those who want to take their own risks in stocks or other derivatives (with or without private non-bank funds) with no excuses. One can argue that while this might reduce the choices of informed investors, these investors should be okay with going it themselves, instead of desiring the pseudo-assurance of having their funds administered by a big bank with all the connotations of trust and responsibility that brings. Allowing banks to double as investment agencies is problematic this way (but of course when the going is good, "free market" and "liberalization" are the iron-clad mantras).

    I would gather that most investors are risk-averse anyway, and would be happy with such an "innovation". Then again, there will always be the greedy breed who will be attracted solely to the highest returns with "reasonable-looking" risks, but then there are inveterate gamblers too, and neither have my sympathy.


    Misc. Cool Links

    Nothing particularly wrong with these - just had to post them before I forgot about them totally.


    Nifty treemap by WinDirStat
    • Kimchi I helped to make, from the LSM1301 Respiration lab. I may eat my words, but I'm not going to eat that!

    Sure it's safe?

    Rooney, Rooney

    Quite contrary - I bashed him for underperformance last month too, and am now happily munching on my comments after his resurgence. Not that Tevez deserved to be dropped, but Ferguson knows best.

    $537/$800 with United saving me from a wipeout last week - Yes, accuse me of bias, but Liverpool simply have ridiculous luck nowadays. And how did City contrive to go behind and then struggle to draw with Newcastle?

    Tottenham losing is about the only thing one can count on these days, and their plight has gone beyond funny by now. If they don't get anything (or even a point) from Bolton on Sunday, they face Arsenal (which will really hurt for Spurs fans), Liverpool and Man City next, and I seriously don't see them netting anything from these fixtures. This would leave them with 2 or 3 points from 12 games, which is nearly a third of the season, and probably worse than even Derby last season.

    $50 on Man Utd (-1.5) vs Everton (at 2.35) - over evens sounds good
    $50 on Sunderland to beat Newcastle (2.15) - looks worth a try



    comments (0) - email - share - print - direct link
    trackbacks (0) - trackback url


    Next: Do Do Do


    Related Posts:
    On The Money
    Open Bookame
    Economics Thus Far
    On Economics
    A Pain, Explained
    		Back to top

    Copyright © 2006-2025 GLYS. All Rights Reserved.