x**********t
发帖数: 45 | 1 遇到一个问题，大家帮我看看我想得对不对？
generate random number from a distribution which is skew to left.
假设生成了500个数据，Y_{500}.
现在我在每个数据上都乘上一个positive number,这些positive values都是从
lognormal distribution中生成的。但是 每一个数据是从不同的lognormal dist生成
的（mean和SD都不同）。 用X_{500}表示。
那么Y_500 * X_500有可能skew to left,也有可能skew to right.是不是这样？
谢谢了。 |
|
|
t********t
发帖数: 1264 | 3 这个vol skew的positive和negative到底有没有准确的定义？哪种图形是positive，哪
种是negative？查了半天也没有个严格说明的
positive put-skew是指OTM的puts implied vol随着strike下降而上升，对么？那
positive call-skew是指OTM call implied vol随strike上升而上升？ |
|
P****d
发帖数: 369 | 4 Depends on asset class. For equities, buying skew means buying downside vol
as vol is negatively skewed. Skew is cheap usually refers to the fact that
downside vol is low. |
|
x**********t
发帖数: 45 | 5 遇到一个问题，大家帮我看看我想得对不对？
generate random number from a distribution which is skew to left.
假设生成了500个数据，Y_{500}.
现在我在每个数据上都乘上一个positive number,这些positive values都是从
lognormal distribution中生成的。但是 每一个数据是从不同的lognormal dist生成
的（mean和SD都不同）。 用X_{500}表示。
那么Y_500 * X_500有可能skew to left,也有可能skew to right.是不是这样？
谢谢了。 |
|
|
r**u
发帖数: 130 | 7 【 以下文字转载自 Programming 讨论区 】
发信人: rubu (rubu), 信区: Programming
标 题: 求助Photoshop里的skew transform的算法
发信站: BBS 未名空间站 (Mon Dec 13 20:27:55 2010, 美东)
有没有大侠了解Photoshop里的skew transform的算法,能否指点一下.自己想重写一下.
或者了解任意四边形的内置椭圆算法也请指点一二,多谢. |
|
c*w
发帖数: 4736 | 8 sorry for using English...
does anyone know how sas computes its skewness?
it looks like sas computes variance by: sum(x - xmean)^2/(N-1)
for skewness, it looks like it has something to do with N as well,
not simply sum(x - xmean)^3/(sqrt(variance)^3). |
|
r****y
发帖数: 1437 | 9
In some software package, skewness is defined with " - 3" so that
the skewness of the normal distribution is 0. In others, it might be without
this "- 3". |
|
c*w
发帖数: 4736 | 10 en... maybe you are talking about kurtosis.
for skewness, it's third moment, so skewness of standard normal is 0,
because standard normal is symmetric. |
|
N******r
发帖数: 642 | 11 skew risk = skewness
smile risk = kurtosis |
|
m*****e
发帖数: 37 | 12 比如说SPX, 当market goes down时, SPX volatility skew会因为更多的人买OTM puts
变大吧? 反之, 当market goes up时, skew变小? 多谢. |
|
发帖数: 1 | 13 SKEW：大盘的价外put已经6连涨了，大家多小心 |
|
E***r
发帖数: 1037 | 14 据我观察，skew在极端值才是反向指标，中间值一般没有指示意义。 |
|
m*********7
发帖数: 293 | 15 彪哥的意思应该是skew peak伴随着short term底部，或者反之。参见二月大跌。 |
|
|
r**u
发帖数: 130 | 17 有没有大侠了解Photoshop里的skew transform的算法,能否指点一下.自己想重写一下.
或者了解任意四边形的内置椭圆算法也请指点一二,多谢. |
|
发帖数: 1 | 18 看到我身边的一个中国女group leader，又一次骂走了学生，听说以前多少次骂哭学生
，不禁有感而发。
带领团队需要的是管理能力、知道怎么激发下属的兴趣和能力，根据下属的特点安排各
有侧重的工作，并且要眼界宽广、大气、能容人能团结人。
反观大部分生物PI，特别是中国PI，就是发了几篇半捏造半灌水的文章，甚至有不少人
十多年发了3篇10分的也当上了group leader，还自命不凡、认为没有自己不会的实验
，对下属尖酸刻薄，做不出他／她要的结果，就出言不逊，辱骂斥责，把自己的压力发
泄到学生的身上。学生在PI面前本来就是弱势，动辄呵斥辱骂，真的好吗？
为什么公司招聘需要做心理测试，PI发了文章就能当group leader？有太多人自己根本
没有能力带领团队，只会skew up 别人的人生。
垃圾行业充斥着垃圾人！ |
|
x****e
发帖数: 1780 | 19 有人问我怎么arbitrage option implied volatility skew。
我不懂这个问题。
有人能在这个问题上面指导一下吗？
谢谢！ |
|
y**********0
发帖数: 521 | 20 Think about a stochastic vol model which can generate this skew and has no
arbitrage.
this |
|
s*u
发帖数: 73 | 21 各位前辈
当有人说the strike skew is cheap是什么意思啊？ |
|
s*u
发帖数: 73 | 22 谢谢 是rates，在这个情况下指的是什么？
btw我看到的研报这么说：
since high strike skew appears cheap. we prefer outright payers to payer
spreads(selling the high strikes) |
|
s*****i
发帖数: 62 | 23 I have a buch of data points that are right skewed, I have tried lognormal,
gamm and weilbull, none seems fit, what is other choices?
Thanks. |
|
m*********n
发帖数: 413 | 24 what about mixture of right skewed distributions,
kaka |
|
p******r
发帖数: 1279 | 25 请教大家，我碰到一个问题，dependent variable是一个0 1 2 3...9共10个的
categorical variable，现在我要对其做 ordinal regression。
但问题是这个dependent variable 严重skewed，1200多个obs大多集中在0，1，2， 在
9那里只有可怜的1个obs，请问这种情况还能做ordinal regression不？如果不能做，
那要怎么办呢？ |
|
l***a
发帖数: 12410 | 26 for very skewed data, you can probably try over sampling when building the
model
0 |
|
p******r
发帖数: 1279 | 27 那请问一般如果x variable严重righ skewed，除了用log来搞，还可以怎样搞呢？来个
x+x^2 ？
the |
|
A*******s
发帖数: 3942 | 28 my understanding is transforming X for introducing nonlinearity. does it
matter that X variables are skewed? |
|
p******r
发帖数: 1279 | 29 well 我觉得x variable没有必要normal dist的，其实只要residue满足normal dist就
好。
不过我的y和其中一个x都很skewed，我把它们用log变来变去好像最后residue还是不满
足normal dist，所以people may question我的significance of coefficient
estimates，because t-test is not valid。。。 |
|
f********x
发帖数: 99 | 30 The world beyond batch: Streaming 101: A high-level tour of modern data-
processing concept
http://radar.oreilly.com/2015/08/the-world-beyond-batch-streami
by Tyler Akidau August 5, 2015
Editor’s note: This is the first post in a two-part series about the
evolution of data processing, with a focus on streaming systems, unbounded
data sets, and the future of big data.
Streaming data processing is a big deal in big data these days, and for good
reasons. Amongst them:
Businesses crave ever more tim... 阅读全帖 |
|
a*****e
发帖数: 1717 | 31 saw one article about this
Options Highest Since 2007 Shows Gains Intact Amid Hedging (1)
(Updates with today’s S&P 500 level in seventh paragraph.)
By Jeff Kearns and Whitney Kisling
April 25 (Bloomberg) -- The end of the Federal Reserve’s
Treasury repurchase program is prompting options traders to pay
the most in four years for protection against stock declines, a
signal that proved bullish in the past.
The cost of three-month put options to sell the Standard &
Poor’s 500 Index ... 阅读全帖 |
|
l*******1
发帖数: 113 | 32 1. what is skew delta? how to calculate?
2. if you enter into a 90-110 risk reversal, how does the skew delta compare
with the black scholes delta?
3.If you hedge your risk reversal with the black scholes delta instead of
the skew delta, do you lose/make money and why?
4.you think the skew is cheap and you buy skew by entering into a risk
reversal. when the spot goes up, do you restrike? Why or why not?
Lucas |
|
g********2
发帖数: 6571 | 33 Voters Don’t Trust Media Fact-Checking
Friday, September 30, 2016
Most voters believe news organizations play favorites when it comes to fact-
checking candidates’ statements, but this skepticism is much stronger among
voters who support Donald Trump than those who back his rival Hillary
Clinton.
A new Rasmussen Reports national telephone and online survey finds that just
29% of all Likely U.S. Voters trust media fact-checking of candidates’
comments. Sixty-two percent (62%) believe instead that... 阅读全帖 |
|
g********2
发帖数: 6571 | 34 ACTUARIAL REVIEW: Analysis of Recent Polls Shows Trump Win and Possible
Landslide
Jim Hoft Oct 18th, 2016 10:02 am Leave a Comment
Guest post by Joe Hoft
trump clinton speeches
Ann Coulter warned in her book Slander in 2003 that the far left main stream
media always uses polls to push their agenda. Polls can be skewed by
selecting an unreasonable sample size, by asking lead up questions or by
selecting more of a sample population of one side of an issue to achieve a
desired result.
Main stream m... 阅读全帖 |
|
z***a
发帖数: 69 | 35 是quantitative equity group. 看样子不是纯quant而是quant developer. 所以要求
的最低学历是本科. 面试官是2个ABC, 很nice. 别的公司都是在我们学校的career
center面试, 他们却是在我们学校的hilton包了4间大客房来面试, 看样子是真想招人.
技术题就2个. 第一个是给4个porfolio的mean, variance, skew, korxxxxx, 然后问哪
个风险最大, 哪个值得投资. skew和那个korxxxxx不知道是什么, 我叫他们解释, 然后
我选了skew是负数的那个风险最大. 然后korxxxxx最大的那个, 正好是skew为负数的那
个, 所以我感觉就是那个porfolio风险最大了. 也不知道对不对. 第二个是写程序. 给
n个骰子, 输出所有骰出来的数字组合. 这个很简单, 可是只给了我3分钟写, 非常紧.
不知道有没有戏, 过了没有....................
祝大家也都找工作顺利了. |
|
b***k
发帖数: 2673 | 36 ☆─────────────────────────────────────☆
zehua (zehua) 于 (Fri Sep 26 08:37:13 2008) 提到:
是quantitative equity group. 看样子不是纯quant而是quant developer. 所以要求
的最低学历是本科. 面试官是2个ABC, 很nice. 别的公司都是在我们学校的career
center面试, 他们却是在我们学校的hilton包了4间大客房来面试, 看样子是真想招人.
技术题就2个. 第一个是给4个porfolio的mean, variance, skew, korxxxxx, 然后问哪
个风险最大, 哪个值得投资. skew和那个korxxxxx不知道是什么, 我叫他们解释, 然后
我选了skew是负数的那个风险最大. 然后korxxxxx最大的那个, 正好是skew为负数的那
个, 所以我感觉就是那个porfolio风险最大了. 也不知道对不对. 第二个是写程序. 给
n个骰子, 输出所有骰出来的数字组合. 这个很简单, 可是只给了我3分钟写, 非常紧. |
|
z****i
发帖数: 406 | 37 转自wilmott
mghiggins
Senior Member
Posts: 313
Joined: Nov 2001
Sat Jan 04, 03 01:54 PM
User is offline
Qualitative explanation for why there is a vol smile in stochastic
volatility models from a market-making perspective:
First, let's step back and look at the standard Black-Scholes world. Why do
options have value over intrinsic value? Because they have positive Gamma.
As a market-making trader, you buy an option and Delta-hedge it - now, the
value of your portfolio as a function of the ... 阅读全帖 |
|
i**********k
发帖数: 5274 | 38 A new Reuters poll has many Republicans worried about the outcome of the
2016 presidential race--and many on the Democratic side ready to pop the
champagne corks.
The poll shows presumptive Democratic nominee Hillary Clinton leading by a
shocking 14 percent over her Republican rival Donald Trump: 47 percent to 33
percent.
Unfortunately there’s one problem with the poll. It is heavily skewed.
The Gateway Pundit reports that of Reuters’ 1,201 respondents, 626 were
Democrats and only 423 were Repub... 阅读全帖 |
|
u********e
发帖数: 4950 | 39 ☆─────────────────────────────────────☆
aripple (aripple) 于 (Wed Apr 20 19:48:48 2011, 美东) 提到:
很多人觉得这个是历史新低，又不会破产降到0等等，觉得是个很好的机会。
这个问题没那么简单，不清楚算法应该去看看CBOE的VIX WHITE BOOK。
里面的算法并不是那么straight forward，有好多种情形可以降低VIX，
远远不止是SPX近两期的月期权的IV降低那么简单。
另外VIX的计算公式中用到了bid/ask的middle price，有点经验的人知道
这就存在了manipulate的可能。对于非常illiquid的strike，可能会通过
各种方式来改变这个参数，而并没有实际交易的意愿。
再就是自从CBOE有了WEEKLY OPTIONS以后，量增加得非常快，品种也越来越，
我记得去年的时候每周才30个左右的symbol。现在已经增加到60多个了，非
常惊人。SPX什么时候列在WEEKLY里面，我不太确定，但至少去年就列入了。
而且经常CHECK weekly lis... 阅读全帖 |
|
d****n
发帖数: 12461 | 40 成本根本不是问题，像网络带宽或者磁盘读写速度这些都是可以提前测试和估计的。
reshuffle的本质问题是需要知道skewness的样子和规模，但是streaming过程中你也许
只知道skewness(可能已经)发生了，但是对于skewness的样子和规模只能对照历史按图
索骥，而这“按图索骥”可能就是个上层内容。其实2.0开始DEA解决了一个很大的问题。
当然spark还有个命门就是DAG。当然这个和streaming下的问题无关了。 |
|
d*******a
发帖数: 518 | 41 t-test 的assumption里面，按照重要性递减是
independent >> equal variance > normality
as long as normality is not very much violated, e.g. heavily skewed, then t-
test is well approximated.
More skewed a sample, more sample size is needed.
会不会认同？
时候，t test逼近z test，对normality假设轻度或者中度的violation可以容忍，是指
该test在大样本的时候还能给出相对可靠的结论，不等于说t
还怎么t-test？ 对于right skewed data还可以做对数变换，我怎么动不动建议用u-
test？不是你说要用不带假设的test么？那只好non-parametric 了，parametric都带
假设。
tendency, Variance/STD--measure of dispersion, 这都对normality分布的数据才有
意义。举个 |
|
s***e
发帖数: 267 | 42 1.如果没有其他条件，只是要求population correlation是0.6, 可以用normal和
cholesky分解：
X 是2维standard normal, A=[1, 1/3; 1/3, 1]，则Y=A*X的两个元素满足条件。
2.skewness是measure是否对称的parameter。normal的变量的skewness都是0。没有其
他限制可以用一个简单的不对
称分布构造，比如bernoulli(p). 给定skewness可以计算p的值，就是要解一个二次方
程（可能需要rescale）。kurtosis
是measure分布的tail是否light/heavy的parameter。可以用laplace distribution或
者bernoulli来构造。 |
|
l***u
发帖数: 91 | 43 assuming a correlation structure then price a option which pays
max(A/B,0) then implied vol from price. You get your vol skew
Otherwise, if you can model your skew in a functional form (local vol),
do an ito on A/B
prices, constant rate, and correlation, is there a practical method to imply
the vol skew of the price ratio of commodity A / B? |
|
w********r
发帖数: 253 | 44 比如federal fund，我把１９８８年６月开始到２０１０年６月，每个月的interest
rate都拿到手，kurtosis大约有１１，所以是fat tail，然后skewness是－１左右，所
以negative skewed。头们誓死不肯用normal distribution近似，所以小的我必须找出
合适的distribution。
大牛门帮忙啊！！！！
目的是为了predict下个月interest rate变动的９９。９５％confidence interval。
我本来是用time series建了个类似AR（２）的model，predict出来下个月的mean和std
，本来想说就可以算９９。９５％confidence interval了，可是头说下个月那个点也不能说是
normal，所以也不可以简单的mean＋３。xx std就得到confidence interval。头说
interest rate现在很低，所以下一点的预测应该是上走的probability大于下降的，所
以每点的propability也应该是skewness的。。。 |
|
w******o
发帖数: 3259 | 45 两个variable,
一个skewness 是-.836 , kurtosis是.235,
一个skewness是-.224, kurtosis 是-.653.
sample size 是27.
请问这kurtosis代表的是什么？从这些distribution的数据来看能不能证明两个
variable是independent的呢？skewness -.224和-.653是significantly different
from each other的吗？怎么证明呢？-.836和.235呢？
万分感谢！ |
|
b********i
发帖数: 1252 | 46 问题是这样的：
一组数据大约几百个
highly left skewed
all above 0
我尝试用各种分布去拟合
没有合适的
log transform之后也还是left skewed
用各种分布去拟合
没有一个合适因为p都很小
我想请问大家还有什么transform的方法
可以让我这组highly left skewed的数据
能够用normal或者gammar或者其他的什么分布来描述
我用的是sas的proc univariate histgram/normal weibull gamma...
真是多谢了！ |
|
c***z
发帖数: 6348 | 47 Hi all,
Does anyone have experience in bias correction? I am working on a project
and the initial result was bad. I would really really appreciate any help!
Here is the details:
Business objective: Correct for panel bias in terms of demographic breakdown
, to obtain an accurate multiplier, in order to calculate site traffic,
sales, etc
Input data:
1. Site visits by people (own data, company Q);
2. Third party (C) 2000 site gender decomposition;
3. Third party personal gender labels (two... 阅读全帖 |
|
l**********8
发帖数: 305 | 48 同事发我的email， 请教大家的见解？
The p-value tests for Normality (kolmogrov-smirnov and shapiro-wilks) are
good indicators of whether data are truly Normal, but the are NOT good
indicators of whether you can use the data in ANOVA or Regression. These
tests will always be significant for large sample sizes, even if the data
are only slightly skewed, and the ANOVA is robust to non-normality for large
sample-sizes from the central limit theorem (even only n=11), meaning your
final conclusions and p-values are ... 阅读全帖 |
|
c***z
发帖数: 6348 | 49 Just did a second round and it was way off.
Here is what I did. Any input is extremely welcome!
Business objective: Correct for panel bias in terms of demographic breakdown
, to obtain an accurate multiplier, in order to calculate site traffic,
sales, etc
Input data:
1. Site visits by people (own data, company Q);
2. Third party (C) 2000 site gender decomposition;
3. Third party personal gender labels (two companies, E and L);
Technical logic:
1. Assign a score to sites based on thei... 阅读全帖 |
|
c****e
发帖数: 3522 | 50 以前在一个老美老太太家中住过，人说了
生了儿子就可以keep family name，
那叫做family honer， 在美国，上次调查是
６０％多的人要男孩，２０％多的人无所谓
剩下的要女孩
look at this article
http://www.theatlanticwire.com/national/2011/06/twice-many-amer
Twice as Many Americans Want Sons Over Daughters
A new book, "Unnatural Selection," by Science writer Mara Hvistendahl
looked at the tremendous shortage of women, particularly in Asia, due to
sex selection. Hvistendahl focused on the West's role in exacerbating the
disparity by exporting abortion technologies, and ... 阅读全帖 |
|
