b***k
发帖数: 2673 | 1 ☆─────────────────────────────────────☆
Portfolio6 ($)^^大虫晚成^^( 于 (Sat Mar 29 09:40:01 2008) 提到:
My boss asked me to compare BA and DB, I found that they have a very high
correlation in 5 year, 2 year, and even 1 year stock prices.
From fundamentals, what is the explanation?
一个是欧洲钱商, 一个是美国铁匠(铝匠)
☆─────────────────────────────────────☆
littletree (本版超级乱指(Mr. 50% right,ohyeah)) 于 (Sat Mar 29 10:45:03 2008) 提到:
is BA the Boeing company? why is it 美国铁匠(铝匠)? just because of the
material |
|
n*******o
发帖数: 18 | 2 Two non-stationary processes X_t, Y_t. Prove that conditional correlation
corr_r(X_r+d,Y_r+d) converges to 1 as d goes to infinity.
Is there any paper I can read? |
|
i*****r
发帖数: 1302 | 3 举个例子
Y X
10% 5%
5% 10%
-10% -5%
-5% -10%
10% 5%
5% 10%
-10% -5%
-5% -10%
10% 5%
5% 10%
-10% -5%
-5% -10%
过程结果
downside correlation, downside beta |
|
c**********e
发帖数: 2007 | 4 You have a time series of stock price. What will result in the
auto-correlation if the series is actually uncorrelated? |
|
k******w
发帖数: 269 | 5 I was studying the valuations of some structured products and confused with
one thing.
My professor told me that in the U.S., the correlation of any two stocks can
't be negative. Is this true? Why?
Thanks. |
|
m********0
发帖数: 2717 | 6 刚算了一下，NASDAQ.
17193 out of 93496 pairs are negative correlated. |
|
r*******r
发帖数: 1014 | 7 you need to explain what's need of correlation for real-time appllication.
online alg in eecs is always about prediction, correction, ...
at here, people usurally talks about sdae, too expensive (price and cost) for you. |
|
w********r
发帖数: 9 | 8 How to define the correlation of two variable X and Y defined in different
domain? For example, X is a continuous function in [0,1], but Y is in [0,1/2
]. |
|
t**********a
发帖数: 166 | 9 correlation doesn't mean too much for dependency structure other than
Gaussian. You might alway calculate
some number, but it doesn't mean it is a good measure of dependency.
/2 |
|
a*********a
发帖数: 3656 | 10 getting into the pricing of a CDO, one first note that the price of a CDO
depends on the correlation in the default processes.
to illustrate this consider CDOs on a basket of 2 assets in a one period
world. each asset has notional of 1, the total notional of the basket is 2.
recovery rates are both 0.
A and B. each has 50% prod to default between time 0 and T. coupon and
protection payments only happen at the end of the single period T.
we have two tranches, equity 0 - 50%, and super senior 50-1 |
|
w********u
发帖数: 328 | 11 请您在解释清楚一点可以吗？emv难道是属于empirical correlation吗？
而且这个怎么manage downside和maximize profit啊？
=(Today’s Last – Yesterday’s EMA) x (Smoothing constant) + Yesterday’s
EMA
Smoothing constant=2/（1+N)
这是我查到的，可是什么叫today's last? 而第一天的ema又是什么啊?
我只有知道第一天的才能做递推啊，谢谢！ |
|
w********u
发帖数: 328 | 12 我理解您解释的empirical correlation的定义了，但是就这道题而言，我就单凭a,b,c
每天的数值，我也不知道他们的权重啊；而且就算算出来了portfolio的variance, 又
跟downside和profit有什么关系? 谢谢！ |
|
m****r
发帖数: 59 | 13 请教统计高人：
计算两个time series的correlation，需要要求这两个time series 都normally
distributed的吗？volatility clustering, heteroskedasticity, time-
dependence 这些是不是问题？需要做Dickie Fuller之类的测试并去除trend和
autoregressive lag吗？
多谢多谢！！！ |
|
D**u
发帖数: 204 | 14 Randomly (also uniformly) choose 3 points A,B,C on a 2-dimensional sphere.
Let x be the spherical area of the spherical triangle ABC;
and y be the volume of the (Euclidean) 四面体 formed by A,B,C and O
(the center of the sphere).
If we treat x and y as 2 random varibles,
then what is the correlation between x and y? |
|
a*********r
发帖数: 139 | 15 Of course not. You can take two independent Brownian motions, then their
correlation is zero. |
|
t***l
发帖数: 3644 | 16 correlation确实我定义搞错了，我一直想的是covariance。
但是你说的那个 2*W_{t/4}是不是martingale呢？可以说是，也可以说不是，这得看你
取得F_t是什么了，具体你看我之前回avidswimmer的一个证明。
至于你说X_t是不是martingale，这还比较显然，两个martingale的linear
combination一定是martingale的。
martingale; |
|
g******r
发帖数: 29 | 17 嗯 就这个意思
我开始就是问 两个布朗运动的 quadratic variation是不是常数xt
也就是correlation的意思
我只是不确定我后面的例子是不是布朗运动
你判断是 dX_t = a(t)dW_t + b(t)dB_t 是布朗运动是用levy判定吧
其实我那个例子就是 a(t) = 1_{t<=1} - 1_{t>1}
那应该就没问题了 |
|
t***l
发帖数: 3644 | 18 convention吧，你也得看看这个例子啊。老抱着convention一成不变有意思吗？
难道说例子中给的两个随机过程在不同的filtration里，随后来算他们的correlation？如果这也是你的convention我没话说。
还有，楼主的问题哪里ill-posed了？我本把corr想成cov了，所以觉得问题提的不对。现在看看提的没问题。
the |
|
a*********r
发帖数: 139 | 19
That's you! From the beginning, you assume a very arrogant position towards
all others. Even if I told you {tB(t)}_t is not a martinagle. You still
resisted thinking about what others say until somebody did this for you. You
should rein in your arrogance and anger and learn the stuff seriously. This
may hurt, but frankly, you know very very little about stochastic calculus
(even the simplest Brownian motion version).
啊。
Exactly the opposite. When you write a paper, you don't write every single
... 阅读全帖 |
|
t***l
发帖数: 3644 | 20 你读书读太多了，如果就这种思维方式，做quant是不适合的。为了你要的严谨：这里
我假设了你想做quant才来这个版的。
correlation process从字面意思就一目了然了吧。为了你我定义一下吧：给定两个随
机过程A_t和B_t，我们可以定义随机过程C_t := corr(A_t,B_t)，明白了？
没什么好争的，就为了那两个不知所云的例子。。。反正楼主出的问题解决了。
towards
You
This
calculus
need |
|
B*******t
发帖数: 135 | 21 按照美国眼下的形式，IT业正在形成一场新的泡沫。在泡沫形成初期进入泡沫是一条投
资策略。但问题是通过什么vehicle进入？
我的想法是找一个跟泡沫correlation最强的vehicle，然后trade这个vehicle的跟
SP500的spread。
目前想到的有tech的ETF（比如VGT，QTEC等等），所以可以trade比如说VGT跟SPY的
spread。
但是这些ETF基本上还是hold的大IT公司，比如GOOG， AAPL什么的。虽然说跟dot com
泡沫也有关联但不是直接的。真正的泡沫应该来源于Facebook，Twitter这些即将IPO的
公司。但ETF似乎找不到跟这块市场关联很强的。
不知道各位知不知道什么vehicle可以直接的跟这些泡沫中心的公司挂钩的？ |
|
m*****e
发帖数: 37 | 22 比如说有两个option A, B, 他们的vol return correlation是0.9. 现有$100 vega的A,
需要卖多少钱的B vega来使得总体的vega risk最小化? |
|
m*****e
发帖数: 37 | 23
多谢回复.
我的问题和portfolio theory中的问题不一样, 你的答案是将A和B的vega portfolio的
standard deviation最小化, 而我的问题是如何最小化 100*vol_A_change +
x*vol_B_change (即最小化vega return).
这个问题也可以改成一个类似的问题: 有两个stock A, B, 他们的stock return correlation
是0.9, 现在买了$100的A, 需要买多少钱的B, 使得股票价格变动时portfolio的return最接近0? |
|
b*****n
发帖数: 143 | 24 I have an impression that the correlation coefficient of (X,Z) is in a
certain range. But according to the triangle rule, it will be a fixed value?
Correct me if I am wrong. Thanks. |
|
p********6
发帖数: 1802 | 25 correlation matrix is positive define?
coefficient |
|
d********t
发帖数: 9628 | 26 把其中一个a*X+b，把另一个c*Y+d，新的distributions的correlation还是rho吗？
谢谢！ |
|
l**********n
发帖数: 303 | 27 Correlation between X and Y is 0.5. X increases by 5 units, by how much Y
will increase. |
|
f********k
发帖数: 136 | 28 N个的情况，是不是N很大的时候只能全部重合？correlation只能是1？
仍然不是太清楚。 |
|
s*******0
发帖数: 3461 | 29 How to calculate the correlation of two stock A and B if we know the odd day
price of stock A and even day price of stock B |
|
s*******0
发帖数: 3461 | 30 How to calculate the correlation of two stock A and B if we know the odd day
price of stock A and even day price of stock B |
|
l********g
发帖数: 18 | 31 这个事高频数据的问题，我学这个的，有个estimator 叫，Hayashi-Yoshida cross-
correlation function。 可以看一下，估计楼主面hedge fund 吧。 |
|
s*******0
发帖数: 3461 | 32 不知道 还是看一下Hayashi-Yoshida cross-
correlation function 这个吧 这个应该是这个问题的官方答案 |
|
A***A
发帖数: 98 | 33 Patrick E. Shrout & Joseph Fliess (1979). Psychological Bulletin, Intraclass
Correlations: Uses in Assessing Rater Reliability.
Go to the library and look it up. |
|
x*****i
发帖数: 22 | 34 Can anyone recommend some references for competing risks models with
correlated outcomes? Thank you. |
|
l******1
发帖数: 292 | 35 If two variables have a strong positive correlation, does that mean that
the means of the two variables will be similar? |
|
l******1
发帖数: 292 | 36 这个我作业题里的一到，我个人的理解是是否two means是相似的？意思是说是不是因
为这两个variables是similar，所以他们才positive correlation？谢谢 |
|
p********a
发帖数: 5352 | 37 随便举个例
y=10x
这个不能说是SAME吧
Correlation(r) = NΣXY - (ΣX)(ΣY) / Sqrt([NΣX2 - (ΣX)2][NΣY2 - (ΣY)2])
就算是1的话，难道你能推出X=Y? |
|
x******n
发帖数: 173 | 38 I do not have an idea available now
maybe you can try the nonparametric method, Spearson's correlation? |
|
o****o
发帖数: 8077 | 39 sample correlation has an approximatly normal distribution after Fisher's z-
transformation |
|
c**********e
发帖数: 2007 | 40 It is a little bit more complicated because the two correlations
hence the two t-values are not independent.
z- |
|
g********7
发帖数: 1145 | 41 通过采样后得到两个的离散信号（采样的时间点全部相同），我能不能通过correlate
coefficience来判断这两个信号的相似程度？
我用的是matlab中的corr2函数，但不太清楚得出的值能不能真实反映这两个信号的相
似程度。 |
|
|
f**g
发帖数: 18 | 43 有两组数据，想知道correlation between serum level>100 and presence of AB？
serum level, presence of AB
23 0
46 0
34 1
165 1
145 0
196 1
。。。。 |
|
q**********0
发帖数: 335 | 44 I have two data set as following (E.activity &. (V.load)) In excel scatter
analysis, they have linear relationship but the R2 is very low ( R2=0.239).
Now I don't know whether this two data set is real correlated to each other.
What else statistical method I can use to analyze them and get a P value?
Thank you very much!
E.activity (V.load) |
|
q**********0
发帖数: 335 | 45 What stastistic method used in correlation regression analysis in software
GraphPad Prism 5? Thanks a lot !!! |
|
s*****n
发帖数: 2174 | 46 The reason why many people have this confusion is the definition of
similarity. Actually, correlation is never is measurement of similarity, but
a measurement of (linear) relationship. In other words, it measures the
similarity of change of X1 and change of X2, we can call it "relative
similarity".
Other things, such as Euclidean distance, could be a measure of similarity (
or "absolute similarity" in my words.) |
|
P****D
发帖数: 11146 | 47 “试着”（当然这不可能）用一条直线把(1,-1),(-1,1),(-1,-1)这三个点连接起来就
知道了啊。显然这个correlation coefficient是负的。 |
|
s******g
发帖数: 193 | 48 FA和PCA本质上是不一样的，FA is to find how the independent variables are
correlated while PCA is simply constructing new independent and uncorrelated
variables.
Since you suspect there might be multicolinearity in your data, then PCA
should be a good method in your situation. It can generate a few new
meaningful uncorrelated variables.The questions remains: is loss of variance
information acceptable? |
|
c*******o
发帖数: 8869 | 49 GOOGLE correlation vs causation.... |
|
s********k
发帖数: 6180 | 50 我想过这个，不过能够把c0或者c1和rho（correlation）套上关系吗？第二如果a，b不
一定是线性的话这样做还可以吗？ |
|
