m********i 发帖数: 298 | 1 如果有一个系统,LONG时75%WINNING TRADES, SHORT时,60% WINNING TRADES
,有哪些可能会造成这样的结果? |
w**n 发帖数: 175 | |
m********i 发帖数: 298 | 3 for trades < 10, maybe, but what if trades more than 1000+?
【在 w**n 的大作中提到】 : luck
|
l*******r 发帖数: 3799 | 4 SEC will invite you to have some coffee.
TRADES
【在 m********i 的大作中提到】 : 如果有一个系统,LONG时75%WINNING TRADES, SHORT时,60% WINNING TRADES : ,有哪些可能会造成这样的结果?
|
m********i 发帖数: 298 | 5 Why a?
【在 l*******r 的大作中提到】 : SEC will invite you to have some coffee. : : TRADES
|
m********i 发帖数: 298 | 6 If long and short are using the same underly rule, but used in a reversed
way, the result should be same 75% or same 60%. why there're differences.
【在 l*******r 的大作中提到】 : SEC will invite you to have some coffee. : : TRADES
|
l*******r 发帖数: 3799 | 7 I don't think it is too hard to achieve such performance in certain perriod.
For example, if simply you long RSI(2) < 50 and short RSI(2) > 50 since
2000, you will have a winning % = 67%. But if you do this during 1950 -
1980, you will be killed.
Really depending on your time frame. Adaptivity is the most important thing
.
【在 m********i 的大作中提到】 : If long and short are using the same underly rule, but used in a reversed : way, the result should be same 75% or same 60%. why there're differences.
|
b********y 发帖数: 5829 | 8 两个independent的系统合起来,必须两个同时发出favorable的指令时才操作,这个问
题基本上就
可以解决了吧?
perriod.
since
thing
【在 l*******r 的大作中提到】 : I don't think it is too hard to achieve such performance in certain perriod. : For example, if simply you long RSI(2) < 50 and short RSI(2) > 50 since : 2000, you will have a winning % = 67%. But if you do this during 1950 - : 1980, you will be killed. : Really depending on your time frame. Adaptivity is the most important thing : .
|
o****r 发帖数: 132 | 9 spread, time value decay, volatile etc...
TRADES
【在 m********i 的大作中提到】 : 如果有一个系统,LONG时75%WINNING TRADES, SHORT时,60% WINNING TRADES : ,有哪些可能会造成这样的结果?
|
l*******r 发帖数: 3799 | 10 short term contrarian and short term trending are two independent strategies
, but you cannot simply combine them using "AND"...
【在 b********y 的大作中提到】 : 两个independent的系统合起来,必须两个同时发出favorable的指令时才操作,这个问 : 题基本上就 : 可以解决了吧? : : perriod. : since : thing
|
|
|
b********y 发帖数: 5829 | 11 基于情绪指标和基于技术上超买/超卖的两个系统就可以,
或者比如说RSI加上均线系统,一个是超买/超卖,一个是trending。。。
strategies
【在 l*******r 的大作中提到】 : short term contrarian and short term trending are two independent strategies : , but you cannot simply combine them using "AND"...
|
m********i 发帖数: 298 | 12 use your example, long RSI(2) < 50 and short RSI(2) > 50. the total winning
% = 67%.
what about for this period, what's the winning % for long(then close trades)
and winning % for short (then cover) trades? it should be same around 67%
right?
perriod.
thing
【在 l*******r 的大作中提到】 : I don't think it is too hard to achieve such performance in certain perriod. : For example, if simply you long RSI(2) < 50 and short RSI(2) > 50 since : 2000, you will have a winning % = 67%. But if you do this during 1950 - : 1980, you will be killed. : Really depending on your time frame. Adaptivity is the most important thing : .
|
l*******r 发帖数: 3799 | 13 I will give you a pass in this trading 101 class, hehe
【在 b********y 的大作中提到】 : 基于情绪指标和基于技术上超买/超卖的两个系统就可以, : 或者比如说RSI加上均线系统,一个是超买/超卖,一个是trending。。。 : : strategies
|
f******e 发帖数: 206 | 14 考虑下上涨和下跌的速度问题. 下跌的速度比上涨快的多,所以即使你多空策略相同,但
实际上,做多的进点一般比做空的进点好, 因为上涨慢,信号出现时,你的进点不错.但下
跌时,等信号出现,已经下来一截,进点就比做多要差,如果stop小的话,一个反弹就stop
了. 我的观点,供参考 |
l*******r 发帖数: 3799 | 15 I don't know for this one. I guess similar. but long/short winning % do
vary a lot for some strategies, I remember.
It is not surprising, I think.
winning
trades)
【在 m********i 的大作中提到】 : use your example, long RSI(2) < 50 and short RSI(2) > 50. the total winning : % = 67%. : what about for this period, what's the winning % for long(then close trades) : and winning % for short (then cover) trades? it should be same around 67% : right? : : perriod. : thing
|
b********y 发帖数: 5829 | 16 看来确实是在大学里的,三句话不离本行。。。
【在 l*******r 的大作中提到】 : I will give you a pass in this trading 101 class, hehe
|
m********i 发帖数: 298 | 17 It maybe a reason... but I doubt
(Agree that down markt may cause bigger slippage), e.g. you have a strategy,
that will short at
every 5 minites, then 5 minutes after that you cover then go long.... etc.
what's your expected winning % for long or short? it should be around same
right?
stop
【在 f******e 的大作中提到】 : 考虑下上涨和下跌的速度问题. 下跌的速度比上涨快的多,所以即使你多空策略相同,但 : 实际上,做多的进点一般比做空的进点好, 因为上涨慢,信号出现时,你的进点不错.但下 : 跌时,等信号出现,已经下来一截,进点就比做多要差,如果stop小的话,一个反弹就stop : 了. 我的观点,供参考
|
m********i 发帖数: 298 | 18 If system does appear this way, it means long and short has different risk
levels for a strategy that is neither long or short biased. how is that
possible?
【在 l*******r 的大作中提到】 : I don't know for this one. I guess similar. but long/short winning % do : vary a lot for some strategies, I remember. : It is not surprising, I think. : : winning : trades)
|
f******e 发帖数: 206 | 19 如果是这样的话,那就取决于市场的走势了,比如今天市场就很难搞,折返点多,而且还是
喇叭口走势,很多策略就傻了. 你说的5M 买卖策略,只能说看市场,不过这种策略,无论
盘整还是趋势走势,都是见外婆的感觉
strategy,
【在 m********i 的大作中提到】 : It maybe a reason... but I doubt : (Agree that down markt may cause bigger slippage), e.g. you have a strategy, : that will short at : every 5 minites, then 5 minutes after that you cover then go long.... etc. : what's your expected winning % for long or short? it should be around same : right? : : stop
|
l*******r 发帖数: 3799 | 20 I don't know the reason. But your assumption that long/short should have
the same risk looks not reasonable. But anyway, winning % is not the most
important results to look at, in my view.
【在 m********i 的大作中提到】 : If system does appear this way, it means long and short has different risk : levels for a strategy that is neither long or short biased. how is that : possible?
|
|
|
l*******r 发帖数: 3799 | 21 Why don't you take a look at the return distribution.
【在 m********i 的大作中提到】 : If system does appear this way, it means long and short has different risk : levels for a strategy that is neither long or short biased. how is that : possible?
|
m********i 发帖数: 298 | 22 This is just an example..to say that winning % for long or short should be
same right?
no matter it's waipo policy or not, if it's 30%, wild guess is: long should
be 30%, short should be 30%.
【在 f******e 的大作中提到】 : 如果是这样的话,那就取决于市场的走势了,比如今天市场就很难搞,折返点多,而且还是 : 喇叭口走势,很多策略就傻了. 你说的5M 买卖策略,只能说看市场,不过这种策略,无论 : 盘整还是趋势走势,都是见外婆的感觉 : : strategy,
|
m********0 发帖数: 2717 | 23 lili大牛,
你的这个67% based on what?
which symbol or indice?
perriod.
thing
【在 l*******r 的大作中提到】 : I don't think it is too hard to achieve such performance in certain perriod. : For example, if simply you long RSI(2) < 50 and short RSI(2) > 50 since : 2000, you will have a winning % = 67%. But if you do this during 1950 - : 1980, you will be killed. : Really depending on your time frame. Adaptivity is the most important thing : .
|
f******e 发帖数: 206 | 24 我觉得不能这样认为多空的胜率应该一样,因为多空市场的特性是不一样的. 牛市是缓
涨缓跌,熊市是急跌急涨,从市场性质来说,熊市stop的几率会比牛市要大一些,所以熊市
胜率低一些我觉得是正常的. 楼主应该统计下,在一段牛市中,熊市中和盘整市场准哦昂
,他们多空胜率分别是多少
should
【在 m********i 的大作中提到】 : This is just an example..to say that winning % for long or short should be : same right? : no matter it's waipo policy or not, if it's 30%, wild guess is: long should : be 30%, short should be 30%.
|
m********i 发帖数: 298 | 25 so, for very long long time frame, since this will include all 牛市, 熊市,
盘整市, the winning % should be expected around same? :-)
【在 f******e 的大作中提到】 : 我觉得不能这样认为多空的胜率应该一样,因为多空市场的特性是不一样的. 牛市是缓 : 涨缓跌,熊市是急跌急涨,从市场性质来说,熊市stop的几率会比牛市要大一些,所以熊市 : 胜率低一些我觉得是正常的. 楼主应该统计下,在一段牛市中,熊市中和盘整市场准哦昂 : ,他们多空胜率分别是多少 : : should
|
f******e 发帖数: 206 | 26 再换个角度考虑下, 市场上总是牛长熊短, 牛市中做空总是胜率低的事情,而牛市做多
胜率比做空要高.
所以说, 从长期来说, 胜率还是不偏向做空的,因为牛市是大部分时间. 所以这样看,做
空还是比做多胜率低一点.
【在 m********i 的大作中提到】 : so, for very long long time frame, since this will include all 牛市, 熊市, : 盘整市, the winning % should be expected around same? :-)
|
m********i 发帖数: 298 | 27 they're not same.. of coz not, because if they're same, then they will have
same winning %...
我想了很久很久,不得其解
【在 l*******r 的大作中提到】 : Why don't you take a look at the return distribution.
|
m********i 发帖数: 298 | 28 en.... This maybe the reason, if you open a stock market's chart for very
long time frame, line always points from bottom left corner to the right top
corner. :-) Nice.
【在 f******e 的大作中提到】 : 再换个角度考虑下, 市场上总是牛长熊短, 牛市中做空总是胜率低的事情,而牛市做多 : 胜率比做空要高. : 所以说, 从长期来说, 胜率还是不偏向做空的,因为牛市是大部分时间. 所以这样看,做 : 空还是比做多胜率低一点.
|
M********i 发帖数: 4082 | 29 说明是牛市
TRADES
【在 m********i 的大作中提到】 : 如果有一个系统,LONG时75%WINNING TRADES, SHORT时,60% WINNING TRADES : ,有哪些可能会造成这样的结果?
|
l*******r 发帖数: 3799 | 30 SP500 index
RSI(2) 非常非常的有名呀,是所有乱七八糟的常用指标里面最有效的一个。
【在 m********0 的大作中提到】 : lili大牛, : 你的这个67% based on what? : which symbol or indice? : : perriod. : thing
|
|
|
m********0 发帖数: 2717 | 31 一般不都是RSI(26)吗?
【在 l*******r 的大作中提到】 : SP500 index : RSI(2) 非常非常的有名呀,是所有乱七八糟的常用指标里面最有效的一个。
|
l*******r 发帖数: 3799 | 32 不是RSI(14)吗?:)
【在 m********0 的大作中提到】 : 一般不都是RSI(26)吗?
|
m********0 发帖数: 2717 | 33 恩恩,不是RSI(14)吗?
【在 l*******r 的大作中提到】 : 不是RSI(14)吗?:)
|
l*******r 发帖数: 3799 | 34 那个都是经典TA啦,呵呵
【在 m********0 的大作中提到】 : 恩恩,不是RSI(14)吗?
|
m********0 发帖数: 2717 | 35 GIVE ME A LESSON ON RSI(2) PLEASE
【在 l*******r 的大作中提到】 : 那个都是经典TA啦,呵呵
|
G*******m 发帖数: 16326 | 36 Win 和 Lose 的定义是什么?
赚一元是Win,输5元是Lose。
50%?
应该是Dollar加权才有意义。 |
G*******m 发帖数: 16326 | 37 你胆小的时候容易Win,不过押得小,胆大的时候可能容易输,但押得大。
如果是这样,怎么算? |
m********i 发帖数: 298 | 38 en. yes. good point, but system does not have bias on long or short.
Short can 胆大. Long can 胆大 too.
Long can 胆小, Short can 胆小 too.
【在 G*******m 的大作中提到】 : 你胆小的时候容易Win,不过押得小,胆大的时候可能容易输,但押得大。 : 如果是这样,怎么算?
|
m********0 发帖数: 2717 | 39 statistics on win trade and loss trade is convention
also
max loss/max gain.
【在 G*******m 的大作中提到】 : Win 和 Lose 的定义是什么? : 赚一元是Win,输5元是Lose。 : 50%? : 应该是Dollar加权才有意义。
|
T*******x 发帖数: 72 | 40 典型的马后炮系统,优化了stop limit所导致。
还有就是 25% 或 60% 错的时候得损失远远超过赢得利润。
这种系统的最大特点就是一上真钱try try,立刻见光死。
TRADES
【在 m********i 的大作中提到】 : 如果有一个系统,LONG时75%WINNING TRADES, SHORT时,60% WINNING TRADES : ,有哪些可能会造成这样的结果?
|
|
|
t*******b 发帖数: 43 | 41 what system is not 马后炮系统 in the world?
【在 T*******x 的大作中提到】 : 典型的马后炮系统,优化了stop limit所导致。 : 还有就是 25% 或 60% 错的时候得损失远远超过赢得利润。 : 这种系统的最大特点就是一上真钱try try,立刻见光死。 : : TRADES
|
F***8 发帖数: 24 | 42 Jim Simons 的系统呗。
【在 t*******b 的大作中提到】 : what system is not 马后炮系统 in the world?
|
m********i 发帖数: 298 | 43 你都没见过这样的系统,你就知道这不存在?LOL。。
Too simple, sometimes too naive.
【在 T*******x 的大作中提到】 : 典型的马后炮系统,优化了stop limit所导致。 : 还有就是 25% 或 60% 错的时候得损失远远超过赢得利润。 : 这种系统的最大特点就是一上真钱try try,立刻见光死。 : : TRADES
|
l*******r 发帖数: 3799 | 44 LOL. I thought you were a high-hand.
A lot of very simple strategies can achieve this kind of performance without
optimizing anything, yet only in
certain period.
【在 T*******x 的大作中提到】 : 典型的马后炮系统,优化了stop limit所导致。 : 还有就是 25% 或 60% 错的时候得损失远远超过赢得利润。 : 这种系统的最大特点就是一上真钱try try,立刻见光死。 : : TRADES
|
M*****g 发帖数: 3145 | 45 把系统模型留给数学家(尤其是从左下角到右上角),咱们就拍脑袋吧。 |