J**********g
发帖数: 213 | 1 The interview from MS was about fifty minutes, though it was supposed to be thirty
minutes long. My performance was...I don't know, not as good as expected. I
learned one thing from it, as said on Heard on the street:
ATQ: answer the quesiton.
I learned most thing by myself, so when I heard the question, it took time for me
to get what he wanted to know exactly. I asked him to repeat quesitons
several times. This is not good at all.
Again if this were my third or fourth interview company, I would for sure
get it...
here are questions I was asked (they want someone playing the role between
modeler and developer, I think).
1. explain your thesis (I am PhD in math)
2. what course you took to strengthen your financial background
3. what's call option. given several call option prices with same
expiriation date but different stike prices, how to the distribution of
underlying stock price (at time T?)?
4. what's the virtual functions, give an example (using the syntax)
5. when is a virtual destruction function is used?
6. what's abastract class? example
7. define two dimensional array dynamically in C++
8. what's the variance of brownian motion?
9. find the variance of \int_0^T f(t)dW, where W is the standard brownian
motion
10, find the variance of \int_0^T Wdt, where W is the standard brownian
motion
11. what's the difference between explicit methods and implicit methods for
finite difference methods.
12, (I don't quite understand this question, so I just give the version in
my ears)for heat equation, using implicit methods, what's the complexity of
solving the question, or finding u^{n+1}, or whatever...I don't know what he
is trying to ask. What a shame, my major in master is computional math...
He was nice and patient, and I can understand his English pretty good. I
asked him to repeat himself a lot. Again, this is really not good.
Also, many people fail the interview with Credit group, and they are still
interviewing people. I know many people in FID went to other banks/funds, so
they do need people to fill those position. Another thing is to try to meet
/talk to/get contact info of recruiting people, especially HR. This is how I
got my interview with MS. Well it just worked again. The HR lady scheduling
the interview for me is really nice. I told my friends about her. One of my
friends talked to her after the info session, and got an interview from MS.
You can apply online but you don't know when you would hear from them. This
also helps you build a good relation the the HR person too. this is VERY
important. I got second interview with Credit group in FID two hours after
my first round interview with my HR lady :-) I got the second round
interview with commodities modeling group next morning. I guess, or I know
she was there helping.
If you are the friend I mentioned in the post, I wish you would not mind. I just
wanna help people get the interviews at least. Thanks.
The last point I wanna make is that, your strength, ability, knowledge,
skill set, personality are the key reasons for you in getting a desired job. I believe I can get a job, a good one, but I need to work on things, and practice. I believe this is true for you, at least most of you, too.
Again good luck to everyone...
Jonathan
PS1. I got messages about HR contact information. I My suggestion is same as before: try to talk to them, make a good impression, and then send a follow-up email. Also I ever emailed the HR lady about giving her email address. She seems not to like it and encourage to apply online instead.
http://www.mitbbs.com/article_t/Quant/31263515.html
PS2: I try to be nice, but I received to a lot of messages about contact info. I ever asked Her for permission of giving her info out when one of my good friends wanted it too. She replied politely but recommended to apply online directly. I would simply refer the above link by "simple1987 (追忆似水年华)" and I think it makes a lot of sense. I would appreciate for your understanding... | c****o
发帖数: 1280 | 2 Thank you for sharing the information~
I
for me
【在 J**********g 的大作中提到】
: The interview from MS was about fifty minutes, though it was supposed to be thirty
: minutes long. My performance was...I don't know, not as good as expected. I
: learned one thing from it, as said on Heard on the street:
: ATQ: answer the quesiton.
: I learned most thing by myself, so when I heard the question, it took time for me
: to get what he wanted to know exactly. I asked him to repeat quesitons
: several times. This is not good at all.
: Again if this were my third or fourth interview company, I would for sure
: get it...
: here are questions I was asked (they want someone playing the role between
| d*j
发帖数: 13780 | | k*******d
发帖数: 1340 | 4 Good luck!
Thanks for sharing.
These questions are not hard. | J**********g
发帖数: 213 | 5 Thanks. Can you let me know how you answer the third and the last question?
those are the ones I had a hard time.
Also, questions were easy, but when he asked me, I did not know how to
answer the question. Like defining an two demensional array dynamically, it'
s not hard, but it seems he was looking something else. It's simple to use
things like
new int a[3][4];
but this seemed not to be the one he wanted.
【在 k*******d 的大作中提到】
: Good luck!
: Thanks for sharing.
: These questions are not hard.
| w*****e
发帖数: 158 | 6 Just my opinion,
If in C programming, (int *)malloc(sizeof(int)*3*4) maybe a better answer.
If in C++, vector maybe better.
?
it'
【在 J**********g 的大作中提到】
: Thanks. Can you let me know how you answer the third and the last question?
: those are the ones I had a hard time.
: Also, questions were easy, but when he asked me, I did not know how to
: answer the question. Like defining an two demensional array dynamically, it'
: s not hard, but it seems he was looking something else. It's simple to use
: things like
: new int a[3][4];
: but this seemed not to be the one he wanted.
| l******e
发帖数: 12192 | 7 MS?
I
for me
【在 J**********g 的大作中提到】
: The interview from MS was about fifty minutes, though it was supposed to be thirty
: minutes long. My performance was...I don't know, not as good as expected. I
: learned one thing from it, as said on Heard on the street:
: ATQ: answer the quesiton.
: I learned most thing by myself, so when I heard the question, it took time for me
: to get what he wanted to know exactly. I asked him to repeat quesitons
: several times. This is not good at all.
: Again if this were my third or fourth interview company, I would for sure
: get it...
: here are questions I was asked (they want someone playing the role between
| J**********g
发帖数: 213 | 8 I did mention using the vector class, and he asked me about the syntax staff
, and then I changed to the old C-style one...
I am working on C++ and actually I just need one or two more days....at
least be able to answer the interview questions...
【在 w*****e 的大作中提到】
: Just my opinion,
: If in C programming, (int *)malloc(sizeof(int)*3*4) maybe a better answer.
: If in C++, vector maybe better.
:
: ?
: it'
| M********t
发帖数: 163 | 9 I think this is not dynamic at all. I think you need to make some class.
?
it'
【在 J**********g 的大作中提到】
: Thanks. Can you let me know how you answer the third and the last question?
: those are the ones I had a hard time.
: Also, questions were easy, but when he asked me, I did not know how to
: answer the question. Like defining an two demensional array dynamically, it'
: s not hard, but it seems he was looking something else. It's simple to use
: things like
: new int a[3][4];
: but this seemed not to be the one he wanted.
| M********t
发帖数: 163 | 10 IMO it is not that easy. It is a long time accumulation. Just try to learn a
little bit everyday.
staff
【在 J**********g 的大作中提到】
: I did mention using the vector class, and he asked me about the syntax staff
: , and then I changed to the old C-style one...
: I am working on C++ and actually I just need one or two more days....at
: least be able to answer the interview questions...
| | | k*******d
发帖数: 1340 | 11 3. 我觉得他在问Volatility smile. Hull书上有，那个distribution不是严格log
normal的，在K大的地方，tail比Log normal小，K小的地方，tail大，这个是equity
option。foreign currency option是两边都是fat tail
7.
int** Darray = new int*[nRow];
for (int i = 0; i
Darray[i] = new int[nCol];
int a[3][4] is not dynamic.
?
it'
【在 J**********g 的大作中提到】
: Thanks. Can you let me know how you answer the third and the last question?
: those are the ones I had a hard time.
: Also, questions were easy, but when he asked me, I did not know how to
: answer the question. Like defining an two demensional array dynamically, it'
: s not hard, but it seems he was looking something else. It's simple to use
: things like
: new int a[3][4];
: but this seemed not to be the one he wanted.
| k*******d
发帖数: 1340 | 12 版上前阵子有人说过，温水煮青蛙。真正到面试了基本上也没啥准备的，鬼知道会问什
么。就是现在觉得年纪大了，记忆力不好使了，看过的老忘记。唉。
a
【在 M********t 的大作中提到】
: IMO it is not that easy. It is a long time accumulation. Just try to learn a
: little bit everyday.
:
: staff
| w*****e
发帖数: 158 | 13 Jonathan,
In C++ vector is container implemented by dynamic array. So, it maybe a good
answer.
New/Delete is part of C++, that's why I prefer answer Malloc/Free if you
using old C-style one.
Just my two cents, correct me if I am wrong.
For my experience, C++ interview questions may focus on different stuff from
what we use everyday. There are some materials may help you refresh your
memory.
C++ primer plus.
effective C++,
more effective C++
staff
【在 J**********g 的大作中提到】
: I did mention using the vector class, and he asked me about the syntax staff
: , and then I changed to the old C-style one...
: I am working on C++ and actually I just need one or two more days....at
: least be able to answer the interview questions...
| J**********g
发帖数: 213 | 14 Thanks man, I am reading C++ primer. I learned C++ nine years ago but I
seldom use it. I can pick it up very quickly, that's why, if I could have
two or three more days, I would not worry about it at all...
good
from
【在 w*****e 的大作中提到】
: Jonathan,
: In C++ vector is container implemented by dynamic array. So, it maybe a good
: answer.
: New/Delete is part of C++, that's why I prefer answer Malloc/Free if you
: using old C-style one.
: Just my two cents, correct me if I am wrong.
: For my experience, C++ interview questions may focus on different stuff from
: what we use everyday. There are some materials may help you refresh your
: memory.
: C++ primer plus.
| J**********g
发帖数: 213 | 15 Thanks and I need to read John hull book now...hehe.
【在 k*******d 的大作中提到】
: 3. 我觉得他在问Volatility smile. Hull书上有，那个distribution不是严格log
: normal的，在K大的地方，tail比Log normal小，K小的地方，tail大，这个是equity
: option。foreign currency option是两边都是fat tail
: 7.
: int** Darray = new int*[nRow];
: for (int i = 0; i: Darray[i] = new int[nCol];
: int a[3][4] is not dynamic.
:
: ?
| M********t
发帖数: 163 | 16 我觉得还是要多平时积累，面试的时候调整好心理。也许临阵磨枪可以躲过一时，但是
躲不过一世。
【在 k*******d 的大作中提到】
: 版上前阵子有人说过，温水煮青蛙。真正到面试了基本上也没啥准备的，鬼知道会问什
: 么。就是现在觉得年纪大了，记忆力不好使了，看过的老忘记。唉。
:
: a
| J**********g
发帖数: 213 | 17 I try to be nice, but I received to a lot of messages about contact info. I
ever asked Her for permission of giving her info out when one of my good
friends wanted it too. She replied politely but recommended to apply online
directly. I would simply refer the link by "simple1987 (追忆似水年华)" in my
post and I think it makes a lot of sense. I would appreciate for your
understanding... | B*********s
发帖数: 306 | 18 No I think he meant to ask for the risk-neutral distribution of the stock
from option prices. This can be recovered by differentiating the option
price wrt to the strike twice.
【在 k*******d 的大作中提到】
: 3. 我觉得他在问Volatility smile. Hull书上有，那个distribution不是严格log
: normal的，在K大的地方，tail比Log normal小，K小的地方，tail大，这个是equity
: option。foreign currency option是两边都是fat tail
: 7.
: int** Darray = new int*[nRow];
: for (int i = 0; i: Darray[i] = new int[nCol];
: int a[3][4] is not dynamic.
:
: ?
| J**********g
发帖数: 213 | 19 Yes, this is the one he mentioned...the reason I was confused is that risk-
neutral distribution is a continous function, and we have only a few prices
for call options with different strike prices. How can we get the
distribution from several discrete numbers??
【在 B*********s 的大作中提到】
: No I think he meant to ask for the risk-neutral distribution of the stock
: from option prices. This can be recovered by differentiating the option
: price wrt to the strike twice.
| B*********s
发帖数: 306 | 20 Can't you fit the few available option prices using splines and then
differentiate? This can be easily done in Matlab.
risk-
prices
【在 J**********g 的大作中提到】
: Yes, this is the one he mentioned...the reason I was confused is that risk-
: neutral distribution is a continous function, and we have only a few prices
: for call options with different strike prices. How can we get the
: distribution from several discrete numbers??
| | | J**********g
发帖数: 213 | 21 This is easy, but I thought to get the exact, theoretical solution, rather
than an approximation. I think it is good and workable too, but my question
is how well this could approximate the original distribution, and whether
the error we would have using this method would fall in the certain range, I
mean, acceptable for the practice. I don't have the actual working
experience, and I wonder if this is indeed how the quants use to get the
distribution of stock price from finite available option prices with
different prices. thanks.
【在 B*********s 的大作中提到】
: Can't you fit the few available option prices using splines and then
: differentiate? This can be easily done in Matlab.
:
: risk-
: prices
| B*********s
发帖数: 306 | 22 yes quants do this all the time. they also study approximation errors. For
example, Jiang and Tian's paper on model free implied vol.
【在 J**********g 的大作中提到】
: This is easy, but I thought to get the exact, theoretical solution, rather
: than an approximation. I think it is good and workable too, but my question
: is how well this could approximate the original distribution, and whether
: the error we would have using this method would fall in the certain range, I
: mean, acceptable for the practice. I don't have the actual working
: experience, and I wonder if this is indeed how the quants use to get the
: distribution of stock price from finite available option prices with
: different prices. thanks.
| B*********s
发帖数: 306 | 23 also you can read papers by jackwerth and Rubinstein.
【在 B*********s 的大作中提到】
: yes quants do this all the time. they also study approximation errors. For
: example, Jiang and Tian's paper on model free implied vol.
| c**********o
发帖数: 213 | 24 great post, thank you for sharing. | J**********g
发帖数: 213 | 25 Thank you. definitely will read it later...
【在 B*********s 的大作中提到】
: yes quants do this all the time. they also study approximation errors. For
: example, Jiang and Tian's paper on model free implied vol.
| J**********g
发帖数: 213 | 26 Yes, it seems to be the original one. guys in business world really have
interesting ideas...
Thanks again...
【在 B*********s 的大作中提到】
: also you can read papers by jackwerth and Rubinstein.
| B*********s
发帖数: 306 | 27 These are both academics though. But good academics are really never very
far from the industry.
【在 J**********g 的大作中提到】
: Yes, it seems to be the original one. guys in business world really have
: interesting ideas...
: Thanks again...
| J**********g
发帖数: 213 | 28 actually that's the taste I like. hehe.
【在 B*********s 的大作中提到】
: These are both academics though. But good academics are really never very
: far from the industry.
|
|
