- February 11th, 2008, 6:11 am
- Forum: Brainteaser Forum
- Topic: The Test Taker
- Replies:
**7** - Views:
**61985**

i disagree. all exam results are correlated: good students tend to be good at all exams. the multiplication needs the independent assumption, which does not hold here.

- January 22nd, 2008, 9:30 pm
- Forum: Brainteaser Forum
- Topic: another card game
- Replies:
**22** - Views:
**72081**

<t>the numeric procedure to solve this problem is to use the dynamic algorithm to backtrack the all possible states. it is sort of like the binomial tree but the probability of up/down is changing along the way. the best strategy is then to pick a new card if the expected pay of the move is larger t...

- January 22nd, 2008, 5:39 pm
- Forum: Brainteaser Forum
- Topic: Maximum permutations
- Replies:
**7** - Views:
**62696**

then the maximum m is be n!: it is to visit all possible states of the string reordering.

- January 11th, 2008, 6:20 pm
- Forum: Brainteaser Forum
- Topic: Uniform distribution and two points
- Replies:
**8** - Views:
**64145**

<t>for b, two commonly used estimators are MLE and MOM (of course, other estimators are possible as well): b_{MLE} = X_{max}, and b_{MOM} = X_1 + X_2.MLE is biased while MOM unbiased. the key is about how to assess the goodness of estimators. usually we use mean square error to evaluate estimators. ...

- January 10th, 2008, 11:14 pm
- Forum: Brainteaser Forum
- Topic: A question from Goldman Sachs interview
- Replies:
**11** - Views:
**68036**

<t>QuoteOriginally posted by: VassiliOption A has half option B's value. That's a restatement of the 'reflection principle.'Actually, we don't exactly know the price ratio because S is a geometric BM. With zero interest and divident, log(S) is a BM with downward trend. The reflective principle does ...

- November 14th, 2007, 6:12 am
- Forum: The Quantitative Finance FAQs Project
- Topic: How can I simulate correlated random numbers?
- Replies:
**34** - Views:
**296569**

<t>Quote Hi, I have same question as you. Do you have any solution now? I am eager to know your answer soon. Thanks!Theoretically, it is to compute the conditional distribution of the random variables conditioned on those known ones. For normal cases, the analytic solution of the conditional distrib...

- October 8th, 2007, 9:55 pm
- Forum: The Quantitative Finance FAQs Project
- Topic: What is convexity?
- Replies:
**34** - Views:
**278724**

<t>QuoteOriginally posted by: drdConvexity, at its root, means curvature, and ONLY curvature. In the case of non-contingent IR products, it has NOTHING to do with volatility or anything else, as such. It is purely a property of the pricing formulas. However, as there are many aspects in finance and ...

- September 17th, 2007, 2:36 am
- Forum: Student Forum
- Topic: which instrument to use, future, option or swap?
- Replies:
**3** - Views:
**65643**

thanks for the replies. it clarifies the issues for me.

- September 9th, 2007, 6:55 pm
- Forum: Student Forum
- Topic: which instrument to use, future, option or swap?
- Replies:
**3** - Views:
**65643**

<t>suppose a fund manager plans to hedge the company's financial portfolio, and various common instruments are out there. the question is how the manager determines which instrument is the right one to use. in other words, what are the pro and cons of these instruments?i know all instruments well (a...

- June 14th, 2007, 3:39 am
- Forum: Student Forum
- Topic: Box-Jenkins
- Replies:
**2** - Views:
**70287**

<t>QuoteOriginally posted by: tibbarIf fitting a CIR model to historic data only (I'm using it to model dividend yields), would it be reasonable to use a time series approach where dt = 1 day and then apply Box-Jenkins estimation?You need to check the model assumption. Likely it won't work because A...

- June 14th, 2007, 3:37 am
- Forum: Student Forum
- Topic: Box-Jenkins
- Replies:
**2** - Views:
**70287**

<t>QuoteOriginally posted by: tibbarIf fitting a CIR model to historic data only (I'm using it to model dividend yields), would it be reasonable to use a time series approach where dt = 1 day and then apply Box-Jenkins estimation?You need to check the model assumption. Likely it won't work because A...

- June 14th, 2007, 3:34 am
- Forum: Student Forum
- Topic: Basic chain rule question.
- Replies:
**14** - Views:
**70951**

QuoteWhat is dF/dG where:F = E[ f(X) ]G = E[ X ]E[ ] is expectation,X is any random variable.It is not right because both F and G are just two numbers.Furthermore, the original question is ill-posed because F is a function of G in general.

- March 24th, 2007, 11:17 pm
- Forum: The Quantitative Finance FAQs Project
- Topic: What are the major strategies used by hedge funds and prop desks?
- Replies:
**12** - Views:
**189353**

Thanks a lot for the list.

- March 7th, 2007, 9:35 pm
- Forum: Brainteaser Forum
- Topic: football score
- Replies:
**17** - Views:
**79858**

<t>we can also use the reflective principle to do the counting.let N be the difference of games brazil won and german won, the status of the game canbe denoted as a pair (t, N). given brazil won the first game, the number of total possiblepaths is C(13,6). the game starts at (1,1), and ends at (14,2...

GZIP: On