Using fixed relations

Bben1006 · Aug 8, 2009

Talk about relationships--check this one out:

http://www.market-topology.com/index.php?option=com_impactopia&view=friend&Itemid=2

and say thanks you!

Quote from JJacksET4:

First of all, I don't think you will ever find a correlation between 2 stocks prices that always works exactly (other then Ultra EFTs vs. ETF, etc). but here are my thoughts.

For simplicity, pretend AAPL and ORCL were each $50.

The problem even if you know that ORCL will move 2X what AAPL will is that the options will be priced for that - in other words, the ORCL options would cost more then the AAPL options in this case (btw I think it's AAPL that would probably be more volatile then ORCL, but that's a different point).

First of all, you said it's intraday, right? Lets say they both open at $50 on expiration Friday - IF and that would be a big IF the options were priced the same, you could sell a 50 call on AAPL and buy a 50 call on ORCL for even money. If AAPL went up $3 and ORCL went up $6 then, you would have a nice profit. If AAPL fell $2 and ORCL fell $4, you wouldn't lose anything. If you were bearish going into the trade, you could do the same thing with puts and reverse the numbers. So, you could also sell a straddle on AAPL and buy one on ORCL (again if it was even money and was certain that ORCL would move 2x as much as AAPL).

Again however, I don't think there is ever a correlation that strong and consistant between 2 actual stocks - and with ETF, the ultras and 3X Bear/Bull do have higher options premiums then the standard ETFs, thus making it impossible to automatically profit so easily from the relationships.

JJacksET4

spindr0 · Aug 9, 2009

Quote from uexkuell:

Whenever e.g. AAPL moves up or down 1% ORCL will move 2% in the same direction.

In other words: The change in price of one symbol will always be twice the change of the other regardless of direction.

If this is true then how do you explain that in the past 100 days, ORCL has risen 35% and AAPL had appreciated almost double that?

spindr0 · Aug 9, 2009

Quote from JJacksET4:

First of all, you said it's intraday, right? Lets say they both open at $50 on expiration Friday - IF and that would be a big IF the options were priced the same, you could sell a 50 call on AAPL and buy a 50 call on ORCL for even money. If AAPL went up $3 and ORCL went up $6 then, you would have a nice profit. If AAPL fell $2 and ORCL fell $4, you wouldn't lose anything. If you were bearish going into the trade, you could do the same thing with puts and reverse the numbers. So, you could also sell a straddle on AAPL and buy one on ORCL (again if it was even money and was certain that ORCL would move 2x as much as AAPL).

If you start with a ridiculous assumption and layer on more ridiculous assumptions, what do you end up with?

Even if you had a very high correlation, selling a naked option on expiration day is no more than a crap shot in the dark. The whole point of a pairs strategy is to take advantage of a reversion to the mean and that takes time.

dagnyt · Aug 9, 2009

Quote from spindr0:

If you start with a ridiculous assumption and layer on more ridiculous assumptions, what do you end up with?

Ridiculous or not, the OP was giving a theoretical example and asking a theoretical question.

He never suggested that the correlation exists in the real world.

Mark

spindr0 · Aug 9, 2009

Quote from dagnyt:

Ridiculous or not, the OP was giving a theoretical example and asking a theoretical question. He never suggested that the correlation exists in the real world.

When you use real symbols (AAPL and ORCL), that implies real world. ABC and XYZ implies theoretical.

JJacksET4 · Aug 9, 2009

Spin,

I agree with Mark and with you - he was starting with an unrealistic assumption that was meant to be fictional, yet he used real symbols, which like you say would have been better with XYZ, ABC, etc. I assumed AAPL and ORCL to be basically fictional for the purposes of this thread.

JJacksET4

dagnyt · Aug 9, 2009

Quote from spindr0:

When you use real symbols (AAPL and ORCL), that implies real world. ABC and XYZ implies theoretical.

Yes, XYZ implies theoretical.

Good day for a swim?

Mark

uexkuell · Aug 9, 2009

Quote from dagnyt:

.. the OP was giving a theoretical example and asking a theoretical question.

He never suggested that the correlation exists in the real world.

Thank you for taking the time to truely read the posts.

It was repeated several times that it was hypothetical, that AAPL and ORCL were used as placeholders.
This was mainly to make the question not too complicated, to avoid contaminating it with other potentially distracting facts and to make it more concrete.
The relation seems to exist for some baskets of stocks under certain market conditions.
Never asked to check if such a relation exists for AAPL and ORCL.

If the question is too confusing in the current form please feel free to insert ABC and XYZ or whatever suits better.

spindr0 · Aug 9, 2009

Quote from uexkuell:

- It's only the relation between 2 symbols that seems to be fixed, never know whether both will be going up or down.

- It has to be an intraday strategy. Never know what the relation will look like at open next day (might shift anywhere overnight).

With options I am absolute newbie. Hope you don't mind if I have some questions therefore.

Sorry, can't see how pairs trading with the underlyings could use the relation.

In the hypothetical example (AAPL/ORCL) data might look like:

1h: AAPL +0.1% ORCL +0.2%
2h: AAPL +0.2% ORCL +0.4%
3h: AAPL -0.1% ORCL -0.2%
4h: AAPL -0.5% ORCL -1.0%

Any swing to either side just followed but with constant amplification.

Unfortunately I cannot spot any sure signs that a reversal to the other side (+ => - or vice versa) takes place. I think for a mean reversion strategy I'd need such signs.

OK, let me try to cobble something relevant together.

A "hypothetical" intraday 2:1 price movement relationship does not give you any edge unless you can figure out the direction and catch some trend. Otherwise, as noted in you 4 hour numbers, you haven't gotten anywhere, particularly when you're talking about something moving in 10ths of a percent.

Even before conjuring up a strategy, I'd say that because options have more slippage than equities, anything that moves up and down so little intraday is going to eat you alive. Only large intraday moves will get you anywhere and that isn't always the case.

spindr0 · Aug 9, 2009

Quote from spindr0:

Even before conjuring up a strategy, I'd say that because options have more slippage than equities, anything that moves up and down so little intraday is going to eat you alive. Only large intraday moves will get you anywhere and that isn't always the case.

Before going forward, I present to the jury that there's insufficient information to do more than speculate at some hypothetical possibilities. Be that as it may...

What's distracting is the 2:1 relationship. You can't be long ABC and short XYX or short ABC and long XYZ in equal amounts because in one direction you make 2:1 and in the other you lose 2:1

Since you stated that you have no clue which direction they'll move, that implies that you have to be directionally long in both directions which suggests long straddles on both. This will do quite nicely in trends, quite poorly in congestion and will eat you alive in intraday slippage if no trend.

To avoid the 2:1 vs 1:2 conundrum, you'd have to double up on one side in order to get to 2:2 which is no edge since regardless of where it goes, you make what you lose. So where does that leave you?

For me, it leads back to my suggestion of a possible pairs strategy. (see next reply)