Morgan Stanley Traders Lost $390 Million in One Day in August
- Thread starter ASusilovic
- Start date
I remember arguing with the professor in statistics 101 for economists when he explained exactly this problem. "implying from the start the coin was under one of the other cups" = total BS to me, back then and still now. If this is a condition that is important, then why don't they include it in the problem right from the start?
not to me...
Seems to me most traders would have chosen the "other" cup and drastically incresed their odds of picking the correct one. Isn't this what trading is all about? Whatever your setup happens to be, hopefully there's some sort of theoretical edge. If so, over the long term you should make money (assuming your risk management, etc. is in order)... simply put, you're trading prob & statistics.
No. Afraid you don't make any sense. The Monty Hall problem is well established. It always pays to switch. it's not a statistical argument - it's one that, if you subscribe to the same basic axioms of logic that everyone else in the reasonable universe do, then you should come to the same conclusion. Could you share with us what argument you can possibly offer that shows it's better not to switch?
That doesnt make any since with the cups. Why would you switch cups? Wouldn't that just be second guessing yourself? If you picked that cup in the first place and then you have the opportunity to switch how can you determine that you have better odds picking the other cup? Its a 50-50! That is probably the most retarded thing I have ever heard lol! Now im gonna be honest im a C student in math and absolutely hate it and find no use in calculus or algebra. Remember trading is a social science at best not a pure science like math, physics and such. So it will never be mastered by those nerds!
Sorry to be so direct, but you must be the biggest idiot under the sun. Why? Because either a) you fabricated the whole story because nobody of a sell-side firm would sit with you for another second after you got this question wrong. Its called the Monty Hall paradox and its such an old shoe that anyone who does not know about it has surely NOT DONE HIS HOMEWORK and does not deserve another minute of interviews. I actually believe you made up the whole interview story. Or b) its true and you must have run into the dumbest trader alive. THis is not some statistical bullshit, it is very simple and straight forward that it pays twice as much in terms of probability to switch. You can think about it, derive it with the Bayes probability formula, you can simulate it by writing a program in C++, you can do it in Excel. You can do all kinds of things but fact remains that it pays to switch. Come on you cannot really be that stupid and are given an opportunity to interview for such position. Have they really only looked at your school name and nothing else? Thats why I think a) applies and not b). (I say those things as a rates options trader for a large sell side firm, knowing what is going on in those interviews and what questions are usually asked)
Good luck in the interviews but I am afraid we will have another big time loss by a trader such as you in the future with such statistical knowledge or lack thereof.
No, you dont make sense because you dont understand probabilities at all. Its a question of conditional probabilities. By the way, statistics itself does not prove or disprove anything and nobody cares about proving anything with statistics. Statistical techniques prove or disprove something after specifying a confidence level. Nobody is right all the times, but the best traders are very happy with the odds in their favor. That is plenty enough.
Here another question that is not too closely related but also confuses people such as you: You flip a coin twice and you know that on the first flip it came up heads. What is the probability you get 2 heads in total after the second flip?
Ano another one: 3 dice. Given one shows a 5. What is the probability that the sum of all 3 is 14?