Automated trading with stocks and 2 options

Discussion in 'Automated Trading' started by tommaso, Jul 27, 2009.

  1. tommaso



    <b>Automated trading with stocks and 2 options</b>

    Here is another automated strategy from a user of mine (to be played by a Bot). (I have permission to discuss it here for possible improvements).
    This is a fusion of 2 strategies which work together.

    [I am currently developing the calculator for him to tune the strategy parameters.]

    This strategy seems to have potential of sure profits and I would like to hear your comments before submitting it to
    extensive test.

    This can be played on either 2 accounts (straddle or out-of-the-money options) or only 1 (on-the-money options),
    depending on the options one buys.

    - IDEA -

    Assume you use 2 accounts.

    Call them: Long account and Short account.

    On the first one you always buy shares, on the second one you always sell shares.

    Assume current price for a given stock is 35.85

    On the LONG account, you buy 100 PUT options strike 36
    On the SHORT account, you buy 100 CALL options strike 35

    On the LONG account always let the bot buy shares under the PUT
    strike. If it's not an open-position order, additionally, require also the price
    has declined of a fixed amount (called Delta) since the last buy.

    On the SHORT account always let the bot buy shares above the CALL
    strike. If it's not an open-position order, additionally require the price
    has raised of a fixed amount (called Delta) since the last sell.

    Now let's consider the LONG account only. The SHORT will be symmetrical.

    While and if the price declines continue to buy, averaging down until
    you reach 10.000 shares (covered by the PUT option). Buy no more shares beyond that
    (however the Bot can close position if the prices jumps up to a profit on shares,
    and restart the process).

    After you are with the 10.000 shares, and deep in the money, may exercise the PUT option.

    While you are within of the descent, let the Bot close your position every time it sees a profit greater than a given target T (due to price reversal).
    So you will also be capturing price oscillations while waiting to reach the limit number of shares.

    The crucial point is how you size the buy orders and the Delta.
    If Bot increases the order size *much* more than linearly while descending, and if the step
    Delta is large enough, there seems to be a minimum *sure* profit (computable in advance), which is also capable to largely cover the cost of the option on the other side too (which may become worthless, or almost).

    [ Clearly if Delta is too small and the sizing is not done appropriately you will lose
    a lot, but that can be computed in advance, given Delta and the sizing factor. ]

    If the price goes up, of course you do the same, on the other account, by placing strongly size-increasing (sell) orders above the call strike price, until you reach -10.000 shares, covered by the call (deep on the money).

    (This powerful game clearly requires capitals)

  2. byteme


    This is a seriously genius idea. I can't believe you're discussing it on a public forum!

    Trading an underlying against a long option position....who would have thunk it?

    I especially like the idea of using 2 accounts - it wouldn't really work otherwise. Genius.

    I think you should dub this strategy "gamma scalping" - don't ask me why, it's a name I came up with and didn't have a strategy to apply it to...until now.

    Definitely has the potential of "sure profits" as you suspected.
  3. tommaso


    Hi byteme, I like you. I know you are smart.

    Thanks for the undeserved "genius", but as you know well, and I stated on the very first lines, it can be applied on 1 account. Actually I suggest it. It's probably better to use slightly out-of the money options. And also for margin reasons.

    I know you are being slightly sarcastic because don't like sharing the good things, but remember that every good things we achieve, we usually build up on something and someone else!

    So don't fear sharing, there is a place for everyone in the markets! :))

    From the name suggestion you propose I suspect you did not see the crucial part of this strategy. You only scratched the surface. But it's natural because I just explained the general idea.

    There are many people searching for "holy grails". But do you think that it comes in a "1+1=2" form ? And I wonder how many of us would be ready to "recognize" it, if someone offered it to us ?

  4. Aren't you supposed to only short in the short account :confused:
  5. tommaso


    Yes. You are right.

    It's a typo (I copy pasted the sentence above and forgot to change the verb).

    Anyway, I have stated previously that on the short account (and in any case, above the CALL strike, if you use a single account) you only either SELL or close the position. You never buy (which you do, instead, below the put strike).

    You are therefore right. The verb is "sell"

  6. heech


    Sarcasm aside, this really *is* gamma-scalping, a potentially profitable strategy.

    You'll profit if the underlying moves enough, and you'll lose (from theta burn) if the underly doesn't.
  7. Yeah, you are in fact long vol and you are paying for it (spread on the options, comish and time decay).
  8. tommaso


    Actually with an appropriate scheme of order sizing I am able to ensure a "worst scenario minimum profit".

    On top of it ("the minimum guaranteed profit") you have the robot which will capture all the oscillations. This is "extra profit", not included in the "worst scenario minimum profit".

    [Also sometimes the other side option is still worth something. And you can even leave it in place, in case it may regain something.]

    I also include a picture to give an idea what I mean about the extra daily profit due to oscillations within the strike and the "extreme" price (when max amount of shares is reached). It's not negligible.
  9. tommaso


    Did I forget it?
  10. So you are basically looking at gamma scalping in a very short time frame then?
    #10     Jul 27, 2009