s******t 发帖数: 926 | 1 股市中,OTM option有非常可观的time value,但是retail investor一般吃不到,即
使吃到了,也是冒了极大的风险的,单边吃time decay(裸卖option)一般是
unlimited risk,风险相对较低的大家常用的covered call也有股票暴跌甚至归零的风
险。于是,如何降低风险赚Time value就成了一个课题。
这里我举个无法实际操作的例子,以防蝌蚪们盲从。例子只是给大家提供一种操作思路
,是完完全全的理论。
众所周知,FAS,FAZ是3x Financial ETF,一般不支持长期hold,因为终究是要归零的
,所以如果操作,一定要快进快出,坚决止损,切不可buy and hope。原则上不建议新
手参与任何有关FAS,FAZ的操作。
那么,既然这两个ETF是decay的,并且运动方向相反,如果相等balance并同时烧不就
可以赚到钱了么?对于相对稳定的市场,这是正确的。但是,一旦单向movement超过20
%,这个strategy就会出现问题,因为两个short的balance已经严重不平衡,不能互相
做hedge了。必须通过rebalance来寻求平衡,这可能需要更多的capital甚至margin,
而且必须时刻能借到FAS或者FAZ。一般broker很难长期借FAS,FAZ给你,一旦被要求平
仓,就会很麻烦。
现在,我们应该如何利用它们的Decay获利,又可以不用担心无法借到的问题?我能想
到的办法是同时建立相等数量的中期time neutral bearish put spread,举个例子,
就是现在建立Oct FAS 24/22 bearish put spread, 假如你用1块买到了(写到这里的时
候bid/ask 0.94/1.04),现在FAS是23, 这个就是没有time value的spread。因为这个
spread是23以下赢利,越低赢利越多直到22以下可以利润最大化,2X。 23以上开始亏
损,越高亏损越多,直到24以上见外婆。
假如我们现在做10月的或者明年1月的spread,both FAS and FAZ,由于time neutral
,加上理论上的delta zero,理想状态下,无论市场怎么走,我们都不会赔钱,如果市
场出现巨大波动,我们也赚不到钱(由于delta zero,最终结局是一个外婆,一个2X,
除非在刚才那个例子里,你可以用低于1刀的价格买到)。但是如果市场波动不大,我
们就可以完完全全吃到decay,而且几乎没有丧失本金的风险. 当然,市场没有理想情
况,远期option巨大的spread让这种strategy实际上必须要冒损失本金的风险,所以请
大家不要轻易尝试!
重申一下,这只是完完全全的理论,没有任何操作价值,只是为了给大家提供一种操作
思路,要时刻铭记市场没有任何套利可能,即使有,也不是为你准备的。
大家Happy Trading! |
G*******m 发帖数: 16326 | |
u********e 发帖数: 4950 | |
G*******m 发帖数: 16326 | 4 建议,要吃organic的Time Decay。 |
u********e 发帖数: 4950 | 5 赞分析
一般的delta hedge 需要不停的reblance, 估计一般股友没办法做到的.
sillycat 这招确实应该不错的. 说到根本, 吃time decay 的收益包括interest 和
short volatility 两个, 如果能一直握到oe 同时股价变化不大的话基本上两个都能吃
到. |
n******n 发帖数: 12088 | 6 no free lunch. what you think about may be priced in already.
【在 s******t 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 股市中,OTM option有非常可观的time value,但是retail investor一般吃不到,即 : 使吃到了,也是冒了极大的风险的,单边吃time decay(裸卖option)一般是 : unlimited risk,风险相对较低的大家常用的covered call也有股票暴跌甚至归零的风 : 险。于是,如何降低风险赚Time value就成了一个课题。 : 这里我举个无法实际操作的例子,以防蝌蚪们盲从。例子只是给大家提供一种操作思路 : ,是完完全全的理论。 : 众所周知,FAS,FAZ是3x Financial ETF,一般不支持长期hold,因为终究是要归零的 : ,所以如果操作,一定要快进快出,坚决止损,切不可buy and hope。原则上不建议新 : 手参与任何有关FAS,FAZ的操作。 : 那么,既然这两个ETF是decay的,并且运动方向相反,如果相等balance并同时烧不就
|
s******t 发帖数: 926 | 7 It is hard to say if it is "priced in" but huge bid/ask spread kills the
deal. I raised an example in FAS, but if you look at FAZ, there is no good
entry for a bearish put spread. With this spread, there is no need to "price
in" anything.
I understand there is no free lunch, if there is indeed no risk, then my
expected profit will equal to the yield of treasury notes.
【在 n******n 的大作中提到】![](/moin_static193/solenoid/img/up.png) : no free lunch. what you think about may be priced in already.
|
g****u 发帖数: 695 | 8 难为你写了这么多字。可惜从头就错了。
FAZ 这种东西压根也没有 time decay 的问题。Nx ETF overtime 不能
跟上指数的原因是它们 daily settlement reset 的机制。这种机制在市场
单调运动的时候实际上是 work for you 而不是 against you 的。
【在 s******t 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 股市中,OTM option有非常可观的time value,但是retail investor一般吃不到,即 : 使吃到了,也是冒了极大的风险的,单边吃time decay(裸卖option)一般是 : unlimited risk,风险相对较低的大家常用的covered call也有股票暴跌甚至归零的风 : 险。于是,如何降低风险赚Time value就成了一个课题。 : 这里我举个无法实际操作的例子,以防蝌蚪们盲从。例子只是给大家提供一种操作思路 : ,是完完全全的理论。 : 众所周知,FAS,FAZ是3x Financial ETF,一般不支持长期hold,因为终究是要归零的 : ,所以如果操作,一定要快进快出,坚决止损,切不可buy and hope。原则上不建议新 : 手参与任何有关FAS,FAZ的操作。 : 那么,既然这两个ETF是decay的,并且运动方向相反,如果相等balance并同时烧不就
|
u********e 发帖数: 4950 | 9 股大帅要不再详细解释一下"daily settlement reset" 和 time decay for "FAZ" 这
种东西?
【在 g****u 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 难为你写了这么多字。可惜从头就错了。 : FAZ 这种东西压根也没有 time decay 的问题。Nx ETF overtime 不能 : 跟上指数的原因是它们 daily settlement reset 的机制。这种机制在市场 : 单调运动的时候实际上是 work for you 而不是 against you 的。
|
j***b 发帖数: 5901 | 10 Do you know why there is a good entry for fas but not for faz?
Because FAS won't go to zero as you said.
A common misconception about 3x etf is that people think they will all go to
zero. In fact, only the reverse 3x etf goes to zero.
price
【在 s******t 的大作中提到】![](/moin_static193/solenoid/img/up.png) : It is hard to say if it is "priced in" but huge bid/ask spread kills the : deal. I raised an example in FAS, but if you look at FAZ, there is no good : entry for a bearish put spread. With this spread, there is no need to "price : in" anything. : I understand there is no free lunch, if there is indeed no risk, then my : expected profit will equal to the yield of treasury notes.
|
|
|
w******s 发帖数: 16209 | 11 这个例子里能再具体些么?
FAS 是24/22
同时建立的FAZ的应该是啥?
20
neutral
,
【在 s******t 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 股市中,OTM option有非常可观的time value,但是retail investor一般吃不到,即 : 使吃到了,也是冒了极大的风险的,单边吃time decay(裸卖option)一般是 : unlimited risk,风险相对较低的大家常用的covered call也有股票暴跌甚至归零的风 : 险。于是,如何降低风险赚Time value就成了一个课题。 : 这里我举个无法实际操作的例子,以防蝌蚪们盲从。例子只是给大家提供一种操作思路 : ,是完完全全的理论。 : 众所周知,FAS,FAZ是3x Financial ETF,一般不支持长期hold,因为终究是要归零的 : ,所以如果操作,一定要快进快出,坚决止损,切不可buy and hope。原则上不建议新 : 手参与任何有关FAS,FAZ的操作。 : 那么,既然这两个ETF是decay的,并且运动方向相反,如果相等balance并同时烧不就
|
u********e 发帖数: 4950 | 12 Totally agree.
The way those fund to get "3X" is to use leverages and it will always cost
to get the "3X" leverage. So I believe "FAZ" does have to pay time premium
to get the "3X". But I might be wrong.
to
【在 j***b 的大作中提到】![](/moin_static193/solenoid/img/up.png) : Do you know why there is a good entry for fas but not for faz? : Because FAS won't go to zero as you said. : A common misconception about 3x etf is that people think they will all go to : zero. In fact, only the reverse 3x etf goes to zero. : : price
|
j***b 发帖数: 5901 | 13 The only reason you can buy a bear spread of FAS at such cheap price is
because people don't expect FAS to drop too much. Everyone knows that FAZ
will drop like shit, and that's why you can't buy similar spread of FAZ at
similar price.
This is what we call "priced in". |
g****u 发帖数: 695 | 14 这没啥深奥的。
以 FAS 为例。Suppose the index went down by 10% for one day then
went up by 11% for the next. Then 1x ETF doesn't lose any thing.
Now FAS is 3 times the DAILY gain/loss because it's daily reset.
So in the above scenario, FAS ends up like 0.7*1.33 = 0.931.
Note that while the index barely moved, FAS lost 7% already.
That's why you see FAS and FAZ both lose value over time, but
it has nothing to do with time!
However, consider another scenario -- If the index went up by
10% in 3 days in a row, 1x ETF ends up with 1.1^3 = 1.331, while
3x ETF ends up with 1.3^3 = 2.197. The 120% gain of the 3x ETF
is way more than 3 times of the 33% gain from the 1x ETF. In
this case the daily reset is doing a favor.
【在 u********e 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 股大帅要不再详细解释一下"daily settlement reset" 和 time decay for "FAZ" 这 : 种东西?
|
s******t 发帖数: 926 | 15 I admit I used the wrong word. I used "time decay" just for easy
understanding. Except for the little ER fee, there is no time decay. I
actually don't know why such ETF could not live up to its goal until you
mentioned the reason. Thanks a lot. But it does not matter for the rest of
my strategy. Our goal is to utilize the under-performance of such ETFs, no
matter where it comes from. Right?
The bottom line is, ideally if I can get both FAS 24/22p spread and FAZ 51/
49p spread at $1 now, there is no way for me to lose money. And due to
underperform of both FAS and FAZ, I have a chance to make some money if
market moves sideways. I understand the chance is actually very small as
most of the time, I end up with $2 payout.
Thanks for your discussion.
【在 g****u 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 难为你写了这么多字。可惜从头就错了。 : FAZ 这种东西压根也没有 time decay 的问题。Nx ETF overtime 不能 : 跟上指数的原因是它们 daily settlement reset 的机制。这种机制在市场 : 单调运动的时候实际上是 work for you 而不是 against you 的。
|
j***b 发帖数: 5901 | 16 Because the market tends to go up over time, so the positive 3x ETF usually
won't go to zero even there is "time decay". But it will suffer very badly
during bear market, but still it can recover during bull market. During bull
market, positive 3x etf out performs market by a huge margin.
【在 g****u 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 这没啥深奥的。 : 以 FAS 为例。Suppose the index went down by 10% for one day then : went up by 11% for the next. Then 1x ETF doesn't lose any thing. : Now FAS is 3 times the DAILY gain/loss because it's daily reset. : So in the above scenario, FAS ends up like 0.7*1.33 = 0.931. : Note that while the index barely moved, FAS lost 7% already. : That's why you see FAS and FAZ both lose value over time, but : it has nothing to do with time! : However, consider another scenario -- If the index went up by : 10% in 3 days in a row, 1x ETF ends up with 1.1^3 = 1.331, while
|
j***b 发帖数: 5901 | 17 That is exactly why you can get that price only in your dream.
【在 s******t 的大作中提到】![](/moin_static193/solenoid/img/up.png) : I admit I used the wrong word. I used "time decay" just for easy : understanding. Except for the little ER fee, there is no time decay. I : actually don't know why such ETF could not live up to its goal until you : mentioned the reason. Thanks a lot. But it does not matter for the rest of : my strategy. Our goal is to utilize the under-performance of such ETFs, no : matter where it comes from. Right? : The bottom line is, ideally if I can get both FAS 24/22p spread and FAZ 51/ : 49p spread at $1 now, there is no way for me to lose money. And due to : underperform of both FAS and FAZ, I have a chance to make some money if : market moves sideways. I understand the chance is actually very small as
|
g****u 发帖数: 695 | 18 The cost to get the leverage is very low under current interest
rate environment. The so-called authorized participants are
institutions whose short-term borrowing rates are close to zero
nowadays. Even there is a cost, it will be at most 1%/YEAR. I
don't think anybody on this board will ever care or try to make
any money out of that.
premium
【在 u********e 的大作中提到】![](/moin_static193/solenoid/img/up.png) : Totally agree. : The way those fund to get "3X" is to use leverages and it will always cost : to get the "3X" leverage. So I believe "FAZ" does have to pay time premium : to get the "3X". But I might be wrong. : : to
|
s******t 发帖数: 926 | 19 I actually don't know. I was just proposing a possible strategy to lead some
discussion... I mentioned 51/49p in an earlier post but it does not mean I
would do it that way. It's just an example.
【在 w******s 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 这个例子里能再具体些么? : FAS 是24/22 : 同时建立的FAZ的应该是啥? : : 20 : neutral : ,
|
e***z 发帖数: 7126 | 20 买淫子(SLV) 10-15 strike的长期普特
应该是稳赚得。 呵呵
【在 s******t 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 股市中,OTM option有非常可观的time value,但是retail investor一般吃不到,即 : 使吃到了,也是冒了极大的风险的,单边吃time decay(裸卖option)一般是 : unlimited risk,风险相对较低的大家常用的covered call也有股票暴跌甚至归零的风 : 险。于是,如何降低风险赚Time value就成了一个课题。 : 这里我举个无法实际操作的例子,以防蝌蚪们盲从。例子只是给大家提供一种操作思路 : ,是完完全全的理论。 : 众所周知,FAS,FAZ是3x Financial ETF,一般不支持长期hold,因为终究是要归零的 : ,所以如果操作,一定要快进快出,坚决止损,切不可buy and hope。原则上不建议新 : 手参与任何有关FAS,FAZ的操作。 : 那么,既然这两个ETF是decay的,并且运动方向相反,如果相等balance并同时烧不就
|
|
|
u********e 发帖数: 4950 | 21 I understand the daily settlement. The 3X ETF fund management has to do the
rebalancing to make sure the etf will have a promised 3X performance on a
daily basis compared to the underlines.
I guess my real questions is about "How does the 3X ETF get the 3X leverage
without paying the time value"?
Or in other words, it looks to good to be true that you can buy 3X etf get
3X up side profit room when the down side risk is still -100% compared to
buying the 1X etf. So there must be something the "3x etf" needs to pay.
My guess was that those "3X etf" was using a complicated leveraging tools (
options, futures , etc) to accomplish the "3X" effect. But they do suffer
the "time value decay" as LZ pointed out.
【在 g****u 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 这没啥深奥的。 : 以 FAS 为例。Suppose the index went down by 10% for one day then : went up by 11% for the next. Then 1x ETF doesn't lose any thing. : Now FAS is 3 times the DAILY gain/loss because it's daily reset. : So in the above scenario, FAS ends up like 0.7*1.33 = 0.931. : Note that while the index barely moved, FAS lost 7% already. : That's why you see FAS and FAZ both lose value over time, but : it has nothing to do with time! : However, consider another scenario -- If the index went up by : 10% in 3 days in a row, 1x ETF ends up with 1.1^3 = 1.331, while
|
s******t 发帖数: 926 | 22 Thanks for your input.
But here is my question:
If there is indeed a substantial difference between FAS and FAZ, who would
buy FAS? They would rush to short FAZ instead. Who would buy FAZ? They would
rush to short FAS instead. This will balance the FAS/FAZ performance.
You may argue that they are hard to borrow, but I would also expect the
option volume of FAZ is way more than FAS too, but it is not the case. Any
idea why? Because FAS is safer than FAZ?
【在 j***b 的大作中提到】![](/moin_static193/solenoid/img/up.png) : The only reason you can buy a bear spread of FAS at such cheap price is : because people don't expect FAS to drop too much. Everyone knows that FAZ : will drop like shit, and that's why you can't buy similar spread of FAZ at : similar price. : This is what we call "priced in".
|
g****u 发帖数: 695 | 23 No, they don't use anything complicated. They have to keep their
cost low. It's a simple leverage just like your margin account.
There is no point to use options.
The index won't drop 33% in a single day, at least that's their
assumption. If it ever did, everybody is busted anyway.
the
leverage
(
【在 u********e 的大作中提到】![](/moin_static193/solenoid/img/up.png) : I understand the daily settlement. The 3X ETF fund management has to do the : rebalancing to make sure the etf will have a promised 3X performance on a : daily basis compared to the underlines. : I guess my real questions is about "How does the 3X ETF get the 3X leverage : without paying the time value"? : Or in other words, it looks to good to be true that you can buy 3X etf get : 3X up side profit room when the down side risk is still -100% compared to : buying the 1X etf. So there must be something the "3x etf" needs to pay. : My guess was that those "3X etf" was using a complicated leveraging tools ( : options, futures , etc) to accomplish the "3X" effect. But they do suffer
|
j***b 发帖数: 5901 | 24 The performance of FAZ, FAS is not determined by demand/supply, they are
determined by performance of components.
would
【在 s******t 的大作中提到】![](/moin_static193/solenoid/img/up.png) : Thanks for your input. : But here is my question: : If there is indeed a substantial difference between FAS and FAZ, who would : buy FAS? They would rush to short FAZ instead. Who would buy FAZ? They would : rush to short FAS instead. This will balance the FAS/FAZ performance. : You may argue that they are hard to borrow, but I would also expect the : option volume of FAZ is way more than FAS too, but it is not the case. Any : idea why? Because FAS is safer than FAZ?
|
s******t 发帖数: 926 | 25 Thanks. Although I think there can be premium/discount.
But their options are determined by demand/supply, right? Who would buy FAS
calls if there is a better option: FAZ puts. Do we see a larger volume/
higher price(due to high demand) of FAZ Puts than FAS calls? I did not....
Thanks for the explanation.
【在 j***b 的大作中提到】![](/moin_static193/solenoid/img/up.png) : The performance of FAZ, FAS is not determined by demand/supply, they are : determined by performance of components. : : would
|
j***b 发帖数: 5901 | 26 Yes there is discount and premium. But very little.
Puts are usually more expensive than calls, even though stocks tend to go
higher.
FAZ puts are like stock calls, because it's reverse index. Thus, it's not
surprising that FAZ calls are more expensive than puts.
The fact is that anything that can hedge downward risk will be expensive.
Stock puts or FAZ calls can all serve this purpose, thus they are expensive.
FAS
【在 s******t 的大作中提到】![](/moin_static193/solenoid/img/up.png) : Thanks. Although I think there can be premium/discount. : But their options are determined by demand/supply, right? Who would buy FAS : calls if there is a better option: FAZ puts. Do we see a larger volume/ : higher price(due to high demand) of FAZ Puts than FAS calls? I did not.... : Thanks for the explanation.
|
r********e 发帖数: 1879 | 27 学习了。 Bid/ask Spread 恐怕会给执行造成困难。 另外Index ETF 都有Volatility
Skew,综合下来会稍微增加PUT SPREAD COST。
【在 s******t 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 股市中,OTM option有非常可观的time value,但是retail investor一般吃不到,即 : 使吃到了,也是冒了极大的风险的,单边吃time decay(裸卖option)一般是 : unlimited risk,风险相对较低的大家常用的covered call也有股票暴跌甚至归零的风 : 险。于是,如何降低风险赚Time value就成了一个课题。 : 这里我举个无法实际操作的例子,以防蝌蚪们盲从。例子只是给大家提供一种操作思路 : ,是完完全全的理论。 : 众所周知,FAS,FAZ是3x Financial ETF,一般不支持长期hold,因为终究是要归零的 : ,所以如果操作,一定要快进快出,坚决止损,切不可buy and hope。原则上不建议新 : 手参与任何有关FAS,FAZ的操作。 : 那么,既然这两个ETF是decay的,并且运动方向相反,如果相等balance并同时烧不就
|
s******t 发帖数: 926 | 28 exactly, 这是纯理论讨论,没有实际操作价值,而且我很多理解都错了,看看后面回
帖还能涨更多知识 :)
Volatility
【在 r********e 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 学习了。 Bid/ask Spread 恐怕会给执行造成困难。 另外Index ETF 都有Volatility : Skew,综合下来会稍微增加PUT SPREAD COST。
|
s******t 发帖数: 926 | 29 我理解你说的,不过你好像误解了我的问题
我是说,为什么大家还要买FAS的call?换成买FAZ的put不是更划算?因为FAZ的
performance不如FAS。
如果是这样,那么FAZ的put势必成交量或者价格(由于需求增多)要高于FAS call。但
实际上我没看到这个现象,FAZ put的交易量不如FAS call,而且IV也不比FAS call高。
expensive.
【在 j***b 的大作中提到】![](/moin_static193/solenoid/img/up.png) : Yes there is discount and premium. But very little. : Puts are usually more expensive than calls, even though stocks tend to go : higher. : FAZ puts are like stock calls, because it's reverse index. Thus, it's not : surprising that FAZ calls are more expensive than puts. : The fact is that anything that can hedge downward risk will be expensive. : Stock puts or FAZ calls can all serve this purpose, thus they are expensive. : : FAS
|
S*******s 发帖数: 13043 | 30 if u shorted soxs call@75 like what i did this morning, you will see it
could bite u really hard |
|
|
G*******m 发帖数: 16326 | |
j***b 发帖数: 5901 | 32 Then there must be something wrong with reality, since your thoughts must be
true.
高。
【在 s******t 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 我理解你说的,不过你好像误解了我的问题 : 我是说,为什么大家还要买FAS的call?换成买FAZ的put不是更划算?因为FAZ的 : performance不如FAS。 : 如果是这样,那么FAZ的put势必成交量或者价格(由于需求增多)要高于FAS call。但 : 实际上我没看到这个现象,FAZ put的交易量不如FAS call,而且IV也不比FAS call高。 : : expensive.
|
t****g 发帖数: 3434 | |
u********e 发帖数: 4950 | 34 Thanks for your insight views.
So what is the explanation why it will lose so much on the dividend adjusted
basis if one investor buy the "FAS,FAZ" at the same time and hold for 1
year?
As you can see, if an investor bought 1 share FAS and 1 share FAZ on 2010-6-
16 at the same time and held them untill yesterday, he would have lost 22.9%
of his investment ( Costs: 95.28 Proceeds: 73.47 Return: -22.9%) .
This is much higher than the regular interest charge.
2010-6-16
FAS: 24.03
FAZ: 71.25 |
j***b 发帖数: 5901 | 35 Ok, Let me explain why FAS near the money calls are not very cheaper then FAZ puts. It's because the infinite upward potential. In a bull market, 3x positive etf goes up very quick, you just need to look at TNA before the Japanese earthquake. This is exactly why LZ's strategy in the original post is very risky. 3x positive etf won't go to zero, actually it has a fair chance of beating the market in the long run. |
S*********g 发帖数: 5298 | 36 this is so-called volatility drag effect
adjusted
6-
9%
【在 u********e 的大作中提到】![](/moin_static193/solenoid/img/up.png) : Thanks for your insight views. : So what is the explanation why it will lose so much on the dividend adjusted : basis if one investor buy the "FAS,FAZ" at the same time and hold for 1 : year? : As you can see, if an investor bought 1 share FAS and 1 share FAZ on 2010-6- : 16 at the same time and held them untill yesterday, he would have lost 22.9% : of his investment ( Costs: 95.28 Proceeds: 73.47 Return: -22.9%) . : This is much higher than the regular interest charge. : 2010-6-16 : FAS: 24.03
|
c***a 发帖数: 260 | 37 飘过
【在 s******t 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 股市中,OTM option有非常可观的time value,但是retail investor一般吃不到,即 : 使吃到了,也是冒了极大的风险的,单边吃time decay(裸卖option)一般是 : unlimited risk,风险相对较低的大家常用的covered call也有股票暴跌甚至归零的风 : 险。于是,如何降低风险赚Time value就成了一个课题。 : 这里我举个无法实际操作的例子,以防蝌蚪们盲从。例子只是给大家提供一种操作思路 : ,是完完全全的理论。 : 众所周知,FAS,FAZ是3x Financial ETF,一般不支持长期hold,因为终究是要归零的 : ,所以如果操作,一定要快进快出,坚决止损,切不可buy and hope。原则上不建议新 : 手参与任何有关FAS,FAZ的操作。 : 那么,既然这两个ETF是decay的,并且运动方向相反,如果相等balance并同时烧不就
|
E**O 发帖数: 1980 | 38 根本没有time decay,从fasfaz出来做到现在了,只有volatility造成的decay。在猪
市基本flat,市场上窜下跳越厉害越decay。 |
b********0 发帖数: 339 | 39 不错,是波动衰减,不是时间衰减。计算很容易。
【在 E**O 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 根本没有time decay,从fasfaz出来做到现在了,只有volatility造成的decay。在猪 : 市基本flat,市场上窜下跳越厉害越decay。
|
u********e 发帖数: 4950 | 40 解释得有道理,见教了.
看来这些etf也一样要吃"类似于"margin call" 的亏啊(输多了翻不回来), 用margin来
获取leverage 在波动大时吃的亏不比time decay的亏少啊.
【在 b********0 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 不错,是波动衰减,不是时间衰减。计算很容易。
|
|
|
s******t 发帖数: 926 | 41 其实这也是一种time decay, 因为市场永远是波动的,之前有人举例子说单边市场FAS
涨的更快,是没错,但是只要波动性存在,长期来看这些3x ETF都存在极其严重的
decay,我打算有空做个simulation,一目了然
继续之前的简单例子,当时assume市场连续三天上涨10%,市场涨幅为33.1%,FAS的涨
幅接近120%,现在假设调整两天,每天-10%,指数变化为 1.1^3*0.9^2=1.07811
3x ETF变化为 1.3^3*0.7^2=1.07653
可见,即使市场最终向上,由于调整出现,很容易逆转FAS带来的优势
【在 u********e 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 解释得有道理,见教了. : 看来这些etf也一样要吃"类似于"margin call" 的亏啊(输多了翻不回来), 用margin来 : 获取leverage 在波动大时吃的亏不比time decay的亏少啊.
|
b********0 发帖数: 339 | 42 明天若有时间我会写出相关的公式推导。
【在 u********e 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 解释得有道理,见教了. : 看来这些etf也一样要吃"类似于"margin call" 的亏啊(输多了翻不回来), 用margin来 : 获取leverage 在波动大时吃的亏不比time decay的亏少啊.
|
u********e 发帖数: 4950 | 43 对,这确实也是一种"变相的time decay". 和普通的derivative time decay 的区别是
这种decay 在市场走完全一根直线的时候比较小一些(只须支付利息,没有volatility
premium loss).
这里面的关键在于"每天调整保证日收益3X". 其实就象每天满仓用3x的margin, 一旦跌
了就会有margin call要求割掉股票来保证仍然是满仓用3x的margin. 所以就是越涨投
资额越大,越跌投资额越小,最后总体上是跌得比涨得多, 在市场是波动持平的状态下会
损失越来越大.
FAS
【在 s******t 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 其实这也是一种time decay, 因为市场永远是波动的,之前有人举例子说单边市场FAS : 涨的更快,是没错,但是只要波动性存在,长期来看这些3x ETF都存在极其严重的 : decay,我打算有空做个simulation,一目了然 : 继续之前的简单例子,当时assume市场连续三天上涨10%,市场涨幅为33.1%,FAS的涨 : 幅接近120%,现在假设调整两天,每天-10%,指数变化为 1.1^3*0.9^2=1.07811 : 3x ETF变化为 1.3^3*0.7^2=1.07653 : 可见,即使市场最终向上,由于调整出现,很容易逆转FAS带来的优势
|
j***b 发帖数: 5901 | 44 类似的simulation我早就做过了。我用dji的数据从1928年模拟到现在。结果是dji的hypothetical 3x etf只是略微under perform dji本身。这是在经历过大萧条那样小屁孩们连想都想象不到的大熊市。如果去掉大萧条,dji三倍etf大大outperform dji本身。
其实即使股价上下波动,最后回到原点,那也同样是三倍反向etf比正向的decay大。原因很简单,股价跌了10%,需要涨11%多一点才能回到原点。
三倍正向etf的特点就是熊市的时候跌的非常快,牛市的时候涨的非常快。但是总体上基本是上涨的,甚至可以outperform市场。只是熊市的drawdown极其巨大。
FAS
【在 s******t 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 其实这也是一种time decay, 因为市场永远是波动的,之前有人举例子说单边市场FAS : 涨的更快,是没错,但是只要波动性存在,长期来看这些3x ETF都存在极其严重的 : decay,我打算有空做个simulation,一目了然 : 继续之前的简单例子,当时assume市场连续三天上涨10%,市场涨幅为33.1%,FAS的涨 : 幅接近120%,现在假设调整两天,每天-10%,指数变化为 1.1^3*0.9^2=1.07811 : 3x ETF变化为 1.3^3*0.7^2=1.07653 : 可见,即使市场最终向上,由于调整出现,很容易逆转FAS带来的优势
|
u********e 发帖数: 4950 | 45 如果像你所做的simulation 那样"从1928年模拟到现在dji的hypothetical 3x etf只是
略微under perform dji本身", 这已经说明了3x etf 的明显弱点: 风险是3x,但收益
却略低.
另外感觉这些3x etf 都适合做单边运动. 你如果将simulation 模拟到2009年3月8日,
也许就DJI 会大幅beat 你的dji hypothetical 3x etf 了.
身。
【在 j***b 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 类似的simulation我早就做过了。我用dji的数据从1928年模拟到现在。结果是dji的hypothetical 3x etf只是略微under perform dji本身。这是在经历过大萧条那样小屁孩们连想都想象不到的大熊市。如果去掉大萧条,dji三倍etf大大outperform dji本身。 : 其实即使股价上下波动,最后回到原点,那也同样是三倍反向etf比正向的decay大。原因很简单,股价跌了10%,需要涨11%多一点才能回到原点。 : 三倍正向etf的特点就是熊市的时候跌的非常快,牛市的时候涨的非常快。但是总体上基本是上涨的,甚至可以outperform市场。只是熊市的drawdown极其巨大。 : : FAS
|
j***b 发帖数: 5901 | 46 如果从1950年开始算,跳过大萧条的话,到现在dji是6.9%左右,而dji 3x etf是13.6%
。去掉1%的管理费也是12.6%。
而如果从1950年开始,到2009年3月6日的话,dji是6.1%,3x etf是11.1%。
,
【在 u********e 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 如果像你所做的simulation 那样"从1928年模拟到现在dji的hypothetical 3x etf只是 : 略微under perform dji本身", 这已经说明了3x etf 的明显弱点: 风险是3x,但收益 : 却略低. : 另外感觉这些3x etf 都适合做单边运动. 你如果将simulation 模拟到2009年3月8日, : 也许就DJI 会大幅beat 你的dji hypothetical 3x etf 了. : : 身。
|
b********0 发帖数: 339 | 47 杠杆是要用点利息,但现在利息很低。
主要的损耗来自波动。这一损耗是由于用日增长率的 m x 来逼近瞬间增长率的
m x 所造成的。所以只要一动就有损耗,只不过这一损耗有时被相应 ETF 固有的
增长(定义为瞬间 m x ETF 的增长)所掩盖了。如果可能用瞬间 m x 的ETF,则增长
只与原指数的首尾有关,
不会有波动损耗。可惜实际上做不到,只能每日调整。
固定某一天,假设 r 为这一天的指数增长率并假设 mr 〉-1。
这一天的波动衰减系数为 D =(1+mr)/(1+r)^m。
譬如 m=2 时, D=(1+2r)/(1+2r+r^2);
m=-1 时, D=1-r^2。
在这两种情形下,只要 r 不为零,D 总是小于 1。
多天的衰减系数就是这些 D 的连乘。假设指数多天的总增长系数为 A,
如果是瞬间 m x ETF,则增长系数为 A^m。再乘上那些 D 的连乘,便得到了
实际ETF增长系数的理论值。这里面忽略了杠杆利息、股息、及其他开销。当然还有实际
操作中的误差。 |
j***b 发帖数: 5901 | 48 你这些衰减系数的公式是哪里来的?
【在 b********0 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 杠杆是要用点利息,但现在利息很低。 : 主要的损耗来自波动。这一损耗是由于用日增长率的 m x 来逼近瞬间增长率的 : m x 所造成的。所以只要一动就有损耗,只不过这一损耗有时被相应 ETF 固有的 : 增长(定义为瞬间 m x ETF 的增长)所掩盖了。如果可能用瞬间 m x 的ETF,则增长 : 只与原指数的首尾有关, : 不会有波动损耗。可惜实际上做不到,只能每日调整。 : 固定某一天,假设 r 为这一天的指数增长率并假设 mr 〉-1。 : 这一天的波动衰减系数为 D =(1+mr)/(1+r)^m。 : 譬如 m=2 时, D=(1+2r)/(1+2r+r^2); : m=-1 时, D=1-r^2。
|
i****e 发帖数: 451 | 49 It really depends on the model assumption and what the true stock market
behavior is.
If we assume on day t, the return of index is X_t, which are mean zero,
independent RV, then the expected return over T days will be
E( \prod (1+X_t) ) = 1.0 === E( \prod( 1 + 3*X_t ) ) (expectation)
This shows that if we are in an ideal/theoreticaly world where stock markets
are independent from day to day, then 3X leverage is fine to hold. The main
problem of the above assumption is that X_t will not be independent. In
fact, \sum X_t will typically be a mean reverting process, rather than a
random walk.
If you make more realistic assumptions, it is not difficult to derive that the decay of leverage ETF is proportional to volatility^2.
So it should be better called "volatility decay" rather than time decay. Time decay is also a nice way to say it as most people on this board don't have a good understanding of volatility.
【在 b********0 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 杠杆是要用点利息,但现在利息很低。 : 主要的损耗来自波动。这一损耗是由于用日增长率的 m x 来逼近瞬间增长率的 : m x 所造成的。所以只要一动就有损耗,只不过这一损耗有时被相应 ETF 固有的 : 增长(定义为瞬间 m x ETF 的增长)所掩盖了。如果可能用瞬间 m x 的ETF,则增长 : 只与原指数的首尾有关, : 不会有波动损耗。可惜实际上做不到,只能每日调整。 : 固定某一天,假设 r 为这一天的指数增长率并假设 mr 〉-1。 : 这一天的波动衰减系数为 D =(1+mr)/(1+r)^m。 : 譬如 m=2 时, D=(1+2r)/(1+2r+r^2); : m=-1 时, D=1-r^2。
|
i****e 发帖数: 451 | 50 Another conclusion from this is, do not hold them if the currently
volatility is high and you think the market will pig...
markets
main
【在 i****e 的大作中提到】![](/moin_static193/solenoid/img/up.png) : It really depends on the model assumption and what the true stock market : behavior is. : If we assume on day t, the return of index is X_t, which are mean zero, : independent RV, then the expected return over T days will be : E( \prod (1+X_t) ) = 1.0 === E( \prod( 1 + 3*X_t ) ) (expectation) : This shows that if we are in an ideal/theoreticaly world where stock markets : are independent from day to day, then 3X leverage is fine to hold. The main : problem of the above assumption is that X_t will not be independent. In : fact, \sum X_t will typically be a mean reverting process, rather than a : random walk.
|
|
|
s******t 发帖数: 926 | 51 谢谢你的数据,我刚刚自己做了另外一组数据,与你的相差甚远。
我取1998年12月至今的XLF performance作为基准,98年12月adjusted closing price
为 18.46, 周五收盘是14.89, 12年半的APY为 -1.7%,如果当时有FAS,价格100刀,
现在还值40美分,APY为-35.7%,远远超过3x
6%
【在 j***b 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 如果从1950年开始算,跳过大萧条的话,到现在dji是6.9%左右,而dji 3x etf是13.6% : 。去掉1%的管理费也是12.6%。 : 而如果从1950年开始,到2009年3月6日的话,dji是6.1%,3x etf是11.1%。 : : ,
|
b********0 发帖数: 339 | 52 是的,做空这类 ETF 等于一种变相的均归操作。
markets
main
【在 i****e 的大作中提到】![](/moin_static193/solenoid/img/up.png) : It really depends on the model assumption and what the true stock market : behavior is. : If we assume on day t, the return of index is X_t, which are mean zero, : independent RV, then the expected return over T days will be : E( \prod (1+X_t) ) = 1.0 === E( \prod( 1 + 3*X_t ) ) (expectation) : This shows that if we are in an ideal/theoreticaly world where stock markets : are independent from day to day, then 3X leverage is fine to hold. The main : problem of the above assumption is that X_t will not be independent. In : fact, \sum X_t will typically be a mean reverting process, rather than a : random walk.
|
j***b 发帖数: 5901 | 53 你看看你的etf的价格曲线就知道了。你看看它从2007到2009是怎么运动的。2007年fas还是110多,到了2009年就剩几分钱了。这个大熊市是它接近归零的原因。没有这个大熊市它是不会归零的。
所以说不是说3x 正向etf不会归零,而是要看市场。而你追求的难道不是不需要看市场的万能策略么?如果碰上大牛市3x etf翻几十倍都根本不奇怪。而反向3x etf翻个4倍就很不得了了。
一个可能翻几十倍的东西,尽管它有可能在某一时段归零,可以把short它来作为万能的策略么?
price
【在 s******t 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 谢谢你的数据,我刚刚自己做了另外一组数据,与你的相差甚远。 : 我取1998年12月至今的XLF performance作为基准,98年12月adjusted closing price : 为 18.46, 周五收盘是14.89, 12年半的APY为 -1.7%,如果当时有FAS,价格100刀, : 现在还值40美分,APY为-35.7%,远远超过3x : : 6%
|
b********0 发帖数: 339 | 54 我虽然主张称这一衰减为波动衰减,但也许这并非最恰当名称。
有人称之为杠杆重置衰减,也许更准确些。
我前面的计算不对市场做任何相关性假设。得出的结论也并非是说
m x ETF 本身是否衰减,而是设置一个理想的对数回报的因素模型
(其中 beta=m,alpha=0),然后计算 ETF 相对这一模型的衰减系数。
这一系数描述的正是杠杆重置衰减。
如果假设市场是完全单向的,反而是波动性越高则更可长期持有。
如果想知道 ETF 本身是否衰减,就要看均归性的强弱了。 |
j***b 发帖数: 5901 | 55 另外,你确定xlf是跟faz,fas对应的么?
price
【在 s******t 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 谢谢你的数据,我刚刚自己做了另外一组数据,与你的相差甚远。 : 我取1998年12月至今的XLF performance作为基准,98年12月adjusted closing price : 为 18.46, 周五收盘是14.89, 12年半的APY为 -1.7%,如果当时有FAS,价格100刀, : 现在还值40美分,APY为-35.7%,远远超过3x : : 6%
|
b********0 发帖数: 339 | 56 一般说来持有一个 alpha=0,beta=2 的股票优于持有S&P 2X ETF。
记得有本量化交易的书上说选择高 beta 股票,不如选择低 beta 股票,
然后用马金增加 beta。我看作者正好搞反了。 |
s******t 发帖数: 926 | 57 好说,咱们把2007年10月以后去掉如何?
2007年10月10日,XLF收盘32.56,从1998年12月算,9年APY +6.5%,同期FAS收益率 +0
.28%
当然,XLF和FAS track的index确实略有区别,但是不会差别到这么大吧
你用DJIA index有个缺陷,就是financial相对DJIA更volatile,而正是volatility
kills 3x ETF
fas还是110多,到了2009年就剩几分钱了。这个大熊市是它接近归零的原因。没有这个
大熊市它是不会归零的。
场的万能策略么?如果碰上大牛市3x etf翻几十倍都根本不奇怪。而反向3x etf翻个4
倍就很不得了了。
能的策略么?
【在 j***b 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 你看看你的etf的价格曲线就知道了。你看看它从2007到2009是怎么运动的。2007年fas还是110多,到了2009年就剩几分钱了。这个大熊市是它接近归零的原因。没有这个大熊市它是不会归零的。 : 所以说不是说3x 正向etf不会归零,而是要看市场。而你追求的难道不是不需要看市场的万能策略么?如果碰上大牛市3x etf翻几十倍都根本不奇怪。而反向3x etf翻个4倍就很不得了了。 : 一个可能翻几十倍的东西,尽管它有可能在某一时段归零,可以把short它来作为万能的策略么? : : price
|
j***b 发帖数: 5901 | 58 你就说你用多大仓位short这东西吧。这十年间它最多的时候涨到10倍。
不要跟我说调整仓位。这东西衰减的原因就是因为它相当于股市涨了它加仓,股市跌了
它减仓。你如果调整仓位那衰减系数就完全不一样了。
+0
4
【在 s******t 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 好说,咱们把2007年10月以后去掉如何? : 2007年10月10日,XLF收盘32.56,从1998年12月算,9年APY +6.5%,同期FAS收益率 +0 : .28% : 当然,XLF和FAS track的index确实略有区别,但是不会差别到这么大吧 : 你用DJIA index有个缺陷,就是financial相对DJIA更volatile,而正是volatility : kills 3x ETF : : fas还是110多,到了2009年就剩几分钱了。这个大熊市是它接近归零的原因。没有这个 : 大熊市它是不会归零的。 : 场的万能策略么?如果碰上大牛市3x etf翻几十倍都根本不奇怪。而反向3x etf翻个4
|
s******t 发帖数: 926 | 59 我什么时候说我short了?我一直强调这是纯理论分析,看看FAS到底有没有decay,具
体怎么操作就是另外一回事了。
现在看来,从98年到07年,financial sector APY 6.5%的情况下,3x ETF 收益在
after fee 之后是负的,这并不代表我要short,我从一开始就是讨论做time neutral
bearish put spread。当然,huge bid/ask difference 又是一个问题。
总之,这是纯理论讨论,不涉及实际操作。
【在 j***b 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 你就说你用多大仓位short这东西吧。这十年间它最多的时候涨到10倍。 : 不要跟我说调整仓位。这东西衰减的原因就是因为它相当于股市涨了它加仓,股市跌了 : 它减仓。你如果调整仓位那衰减系数就完全不一样了。 : : +0 : 4
|
j***b 发帖数: 5901 | 60 理论上的讨论根本就没有必要进行。理论上是不可能存在arbitrage的。
你那个put spread的问题根本就不单单是bid ask spread。而是价钱根本就不是那么好。
option理论上的价格就是要达到期望为零。就是说理论上讲,你的任何一笔option交易在价格模型下都是不赔不赚。一个option是这样,spread也还是这样。只有在实际中才有over/under priced option,或者给予对市场方向的正确猜测赚钱。
neutral
【在 s******t 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 我什么时候说我short了?我一直强调这是纯理论分析,看看FAS到底有没有decay,具 : 体怎么操作就是另外一回事了。 : 现在看来,从98年到07年,financial sector APY 6.5%的情况下,3x ETF 收益在 : after fee 之后是负的,这并不代表我要short,我从一开始就是讨论做time neutral : bearish put spread。当然,huge bid/ask difference 又是一个问题。 : 总之,这是纯理论讨论,不涉及实际操作。
|
|
|
G*******m 发帖数: 16326 | 61 在大风大浪之中,alpha,beta,都是狗屁。 |
G*******m 发帖数: 16326 | 62 首先是要有合适的对象;
其次是要有合适的trend,猪也是trend;
最后要有合适的方法。 |
b********0 发帖数: 339 | 63 以前我对这类 m x ETF 组合做过整体性的测试,主要是测试同时放空他们的正反对,
包括定期或定量仓位重调。得到的结论是:风险高(定义为MDD),回报低。或者说单
位风险上的回报低,没有实际操作价值。前面说过,放空这类ETF等于天天做均归操作
,均归操作风险高是很自然的。但即使只考虑均归型组合交易,也有很多好得多的方法
。但我没做过相关OPTION的组合交易测试。 |
b********0 发帖数: 339 | 64 顺便说几句题外话。今年的股市对动量型统计套利交易极有利,尤其是风险极低。
当动量强时,均归型交易一般效果要差些。 |
G*******m 发帖数: 16326 | |
G*******m 发帖数: 16326 | |
u********e 发帖数: 4950 | 67 是,直接放空估计肯定不行, 除非用大量的资金放很少量的股数. 但这样一来资金利用
效率又会大大下降. 不知道有没有人试过sillycat 的bear put spread
【在 b********0 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 以前我对这类 m x ETF 组合做过整体性的测试,主要是测试同时放空他们的正反对, : 包括定期或定量仓位重调。得到的结论是:风险高(定义为MDD),回报低。或者说单 : 位风险上的回报低,没有实际操作价值。前面说过,放空这类ETF等于天天做均归操作 : ,均归操作风险高是很自然的。但即使只考虑均归型组合交易,也有很多好得多的方法 : 。但我没做过相关OPTION的组合交易测试。
|