SERVING THE QUANTITATIVE FINANCE COMMUNITY

Search found 53 matches

March 2nd, 2004, 8:08 am
Forum: Brainteaser Forum
Topic: Hats . . .
Replies: 7
Views: 190091

Hats . . .

Ob (with the help of Dima) has the answer.What's interesting is this would work with N colours, assuming the prsioners had exceptional memories (especially under such stressful curcumstances).
March 1st, 2004, 2:37 pm
Forum: Brainteaser Forum
Topic: Hats . . .
Replies: 7
Views: 190091

Hats . . .

<t>100 prisoners are to be executed. However, the prison governer is persuaded into allowing at least some of the prisoners a chance of reprieval.They are each to be given a hat that is either red, green or blue, each prisoner obvioisly not knowing the colour of his own hat. They are then to line up...
January 6th, 2004, 3:47 pm
Forum: General Forum
Topic: Changes in MTM of a CDO
Replies: 0
Views: 189089

Changes in MTM of a CDO

<t>I am trying to come up with a simple model to calculate what would have to happen to the subordination of a tranche in CDO after one of the assets has been switched, so as the MTM did not change.I considered, to first order, assuming MTM = MTM(Spread(S),Beta(B),Recovery(R),Attachment Point(V)), t...
November 18th, 2003, 9:13 am
Forum: General Forum
Topic: Correlation Risk
Replies: 13
Views: 191721

Correlation Risk

<t>A simple correlation tweak I came across was each name is tweaked up and down using the rule:b +- (1-b)fwhere b is the beta of the name wrt the portfolio, and f some factor (here was 2%).However, I couldn't see the rationale of tweaking names with a large beta by a small amount, and vice-versa.Do...
September 5th, 2003, 7:06 am
Forum: Student Forum
Topic: CDOs: Correlation
Replies: 9
Views: 189878

CDOs: Correlation

<t>Thanks guys, that's been very helpful. I think for me there's been some confusion between correlation of defaults and correlation of outcomes: in the former case, if it's high, then then the value of lower tranches increases (to the seller of protection). When looking at correlation of outcomes, ...
September 4th, 2003, 4:05 pm
Forum: Student Forum
Topic: CDOs: Correlation
Replies: 9
Views: 189878

CDOs: Correlation

<t>The graph I managed to grasp as it's intuitive, I just can't tally it with the statement that to a buyer of protection, the equity tranche has more value because in a low correaltion case there is a higher probablility of a low number of defaults (as opposed to in a high correlation case). This s...
September 4th, 2003, 2:50 pm
Forum: Student Forum
Topic: CDOs: Correlation
Replies: 9
Views: 189878

CDOs: Correlation

<t>Thanks Aaron.The reason I ask is the following statement I read in a presentation a friend had:"Default correlation affects the shape of the portfolio loss distribution:As default correlation increases the portfolio behaviour converges thus making senior tranches more risky and junior tranches le...
September 4th, 2003, 2:03 pm
Forum: Student Forum
Topic: CDOs: Correlation
Replies: 9
Views: 189878

CDOs: Correlation

<t>Could anyone explain the effect of correlation in a CDO?More specifically, why it is better to receive losses on an equity tranche of a CDO if the portfolio has a low correlation? I would have thought that a low number of defaults is a high correlation situation as you have a correspondingly high...
August 20th, 2003, 12:49 pm
Forum: Student Forum
Topic: A question of Correlation?
Replies: 2
Views: 191070

A question of Correlation?

I'll try normalizing like you say. I just need a ready-reckoner to so that I get a subset of names, but don't eliminate those with good (high) spread, and low corr.
August 20th, 2003, 12:04 pm
Forum: Student Forum
Topic: A question of Correlation?
Replies: 2
Views: 191070

A question of Correlation?

<t>If I had a portfolio of Names, each with 5Y spreads and their respective corr. %ges, what would be an informative metric to use to find the names with high spread and low corr.? I'm just looking for a rough guide.I'm trying simplistic methods at the moment, eg Spread *(1-Corr.)^4, to try and filt...
August 15th, 2003, 6:43 am
Forum: Brainteaser Forum
Topic: Another bloody brainteaser . . .
Replies: 2
Views: 190193

Another bloody brainteaser . . .

<t>A finite number of points are drawn on the plane. Each is colouredblack or white. Between ( on the line joining them ) every two blackpoints there is a white point and every two white points there is a blackpoint.Prove that the points must all lie on a straight line - i.e. cannot havethis structu...
August 6th, 2003, 7:16 am
Forum: Off Topic
Topic: London attractions
Replies: 26
Views: 190646

London attractions

That's ok, you can take a day to get a joke . . .
August 5th, 2003, 5:18 pm
Forum: Off Topic
Topic: London attractions
Replies: 26
Views: 190646

London attractions

<t>Quotewhen I'm in London, I buy a bunch of pills, drink a lot of pints, and search the red phone booths for the perfect one night stand. Doesn't everyone do this?It's amazing what people will try to sell in these phone boxes - electrical equipment, everything. I saw a '19 year-old TV' for sale onc...
July 31st, 2003, 11:28 pm
Forum: Off Topic
Topic: Roulette Arbitrage
Replies: 47
Views: 193361

Roulette Arbitrage

<t>I think I saw something like this in Scarne's Guide to Gambling. A high-roller playing roulette at Harrah's in Lake Tahoe was continually making money - within 5 days he had made $400,000. The casino manager consulted Scarne who informed him that it wasn't his max bet ($7000) that was the problem...
July 27th, 2003, 10:36 pm
Forum: Off Topic
Topic: Einstein's Quizz
Replies: 55
Views: 192348

Einstein's Quizz

You're not thinking fourth dimensionally!It's as simple as going up to any guy and asking "If I were to ask you yesterday if this is the door to freedom, what would you have said?"If the answer's "Yes" you know that must be the door to freedom, regardless of whom you're talking to.

GZIP: On