Morgan Stanley Traders Lost $390 Million in One Day in August

GTS,

the role was for trading, and we were discussing statistical theories for applying to market. My answer YES is for the market, and anyone with my frame of mind, in holding winning stocks won't switch.

I don't understand why this has become such a big argument.

I did not contradict my original post. in my original post I give clear calculation as to why switching has higher probability.

But we're discussing applying this approach to trading - and it's really about decision making.

I would personally stick to my selection given that it 'survives' a exclusion round.

Please try to read my posts and grasp what I'm getting at, I know I might explain this well but it's not that simple.
 
Quote from GTS:



And lastly, again, this problem has nothing to do with statistics (its probabilities) and your calculations above are incorrect. The original odds of making the right choice was 1/3. After one cup no containing the coin was removed, switching your choice improves your odds to 2/3. Now explain again where 0.16 comes from? [/B]
More...

Probability falls under topic of Statistics last time I checked.

1/3 (round one) x 1/2 (in round 2) = 0.16.. do you know how to calculate probabilities of events? this is basic stuff. if you can't don't be so hyped thinking you understood what I said in the first place.
 
Quote from Batman28:
the role was for trading, and we were discussing statistical theories for applying to market. My answer YES is for the market, and anyone with my frame of mind, in holding winning stocks won't switch.
More...
There is no analogy between the Monty Hall problem and the market. In order for such an analogy to exist you would need someone with god-like powers to remove a losing choice from consideration. If that did indeed happen (perhaps inside knowledge) then the answer to the question would remain the same - you should switch your choice.

Quote from Batman28:
I did not contradict my original post. in my original post I give clear calculation as to why switching has higher probability.
More...
You gave the wrong probability
 
OK. Someone clear this up for me. The original post about the cups doesn't say anything about removing a cup that does not hide the coin. It just says, I pick a cup, you remove one of the remaining cups.

The Monty Hall example from Wiki clearly states that no matter which door you pick, he will reveal one of the doors with a goat. So if you originally pick the door with a car and switch, you lose, but is you pick a door with a goat, then switch, you win. Since 2 doors have a goat, you have a 2/3 chance of wining by switching.
 
Quote from Batman28:
Probability falls under topic of Statistics last time I checked.
More...
Fair enough...I disagree but I'm not going to argue over semantics.

Quote from Batman28:
1/3 (round one) x 1/2 (in round 2) = 0.16.. do you know how to calculate probabilities of events? this is basic stuff. if you can't don't be so hyped thinking you understood what I said in the first place.
More...
There are not two rounds in this example. What exactly do you think has a 16% chance of happening? Seriously, if you are going to act like a know-it-all its best that you get your shit straight first.

Probability that the first guess is right 33% (at any time during the exercise)

Probability that switching cups will yield the right cup 66%

16% does not come into play.
 
Quote from Arnie:

OK. Someone clear this up for me. The original post about the cups doesn't say anything about removing a cup that does not hide the coin. It just says, I pick a cup, you remove one of the remaining cups.
More...
Yes, batman did not type the example correctly.

If the person just randomly removes one of the two remaining cup (that may or may not contain the coin) then there is no point to the whole discussion and changing your pick doesn't increase your odds at all. The fact that the cup which is removed is guaranteed not to contain the coin (not a random selection) is why switching your choice is the right answer << information has been given to you. If you choose the ignore that information (and stick with your original uninformed guess) then you are foolish.
 
Quote from Batman28:


What I was arguing on this is that this does not necessarily hold true IN THE MARKET.
More...

Don't be silly. The monte hall problem is a classic illustration of Bayesian reasoning: the idea that you start with a belief about something. Then you observe something. You then figure out how likely you are to see what you observed given your belief. Finally, you update your belief. It is applied to everything from corporate finance to portfolio allocation decisions to risk management. More than that, it's a framework of reasoning that's rather useful for investment analysis.

You are just pretty arrogant. But that's okay. A few rounds of rejections will do wonders.
 
For those of you pondering the Monty Hall problem, I also had a hard time accepting the right answer until I looked at it this way:

Instead of 3 cups (or 3 curtains), imagine there are 100 curtains, and you choose 1 that you think has the great prize behind it. Monty then eliminates 98 curtains that did not have the big prize behind them leaving you with 2 curtains - one is your original choice, the other is the only other choice remaining.


Now do you switch? :D
 
Quote from IluvVol:

Well, what do we know? Really nothing. Its all rumors, news reporting. The number is out there. What contributed to this loss? A single trader, a group. Did the manager know? Traders are sometimes wrong on a bet. As long as it lies within the risk limits, it can happen. If the limits are violated action needs to be taken.
More...

Evidently you failed to read the link in the original post. This was not "rumors". This was a quarterly regulatory filing. Here, let me help you out:

"Oct. 10 (Bloomberg) -- Morgan Stanley, the world's second- biggest securities firm, said its quantitative strategy traders lost $390 million during a single day in August as their computer models failed to account for ``widespread'' investor selling.

The company's traders lost money on 13 days during the quarter ended Aug. 31, the New York-based firm said in a quarterly regulatory filing today. ``The largest loss days resulted from losses associated with quantitative strategies in early August 2007, when these strategies were adversely affected by widespread portfolio reductions,'' the company said.

Morgan Stanley said last month that the quantitative strategies group lost $480 million during the quarter after being caught off-guard when other investors sold securities to reduce borrowings. Separate areas of the equity sales and trading unit made up for the losses, enabling it to report $1.8 billion of revenue for the third quarter, up 16 percent from a year earlier."

OldTrader
 
Quote from GTS:

Fair enough...I disagree but I'm not going to argue over semantics.

There are not two rounds in this example. What exactly do you think has a 16% chance of happening? Seriously, if you are going to act like a know-it-all its best that you get your shit straight first.

Probability that the first guess is right 33% (at any time during the exercise)

Probability that switching cups will yield the right cup 66%

16% does not come into play.
More...

I believe batman28 was calculating the odds of each branch of the decision tree

1/3*1/2 = 1/6
1/3*1/2 = 1/6
2/3*1/2 = 1/3
2/3*1/2 = 1/3

1/6+1/6+1/3+1/3 = 1
 
You must log in or register to reply here.
Back
Top