Yeah your fine trading TWM...
Automated trading with stocks and 2 options
- Thread starter tommaso
- Start date
Do you have access to an option calculator...? Many option brokers provide one.Quote from tommaso:
Hmm, perhaps making it more narrow may help (?) Here is another plan attempt.
Yes, perhaps its a good idea to close it early in case it's passing too much time.
Anyway sec like twm are quite wild and I think that most of the time we should pull it out... after all we cannot win each and every time (or someone would be soon chasing us!)
I use this sometimes:
http://www.hoadley.net/options/optiongraphs.aspx?
If you plug in volatility of 57% (which is what the current implied volatility of TWM options are trading at), and if you plug in days until expiration of 20 days (assuming we're looking at Aug09 options)...
That tool tells you the probability of the stock ending *below* $30 at expiration is about 12%. The probability of the stock going above $40 at expiration is also going to be about 12%. So, your scenario is 24% likely.
Now, if you think you're going to beat the odds... you need a good understanding of BS, and have a good theory of why the assumptions implicit in that model doesn't apply to what you're doing.
Kind of my inital impression but as i looked deeper into the model a simple formula can determine his net liquidation value at any moment.
Essentially he's wrapped his trading range and is churning share trades waiting for a break out in either direction. A $40K vegas wager of sorts betting on a 20% break out. 12% odds based on the post above but that's not the complete picture.
Seems to me if he keeps a running tally of net oscilation profits against the position's liquidation value he may be better off closing out the entire position using some type of calculation:
Net profit target
time in play limit
or whatever best takes advantage of the position.
I understand he's trying to capture the curve... churning trades and scaling up his position while it's covered. Buying on pull backs, scalping small profits... pretty effective plan.
My question is Why hold and play out the hand if you can take a substantial portion of the pot now?
On the otherhand once he's hit a breakout Why exercise the option and close out? Seems he can lock his profits with a stop and churn 10K scalps on pull backs like a mad man until the market checks up.
But I understand he's trying to determine a worst case scenario.
So the danger I see outside of execution FUP's is:
1. Not hitting the 10K share load up to liquidate and just cycling low volume scalps while the options decay. Flat Mountain Syndrome
2. Liquidating a well covered breakout too soon.
3. Trying to eliminate losses in lieu of minimizing losses and time in play.
Essentially he's wrapped his trading range and is churning share trades waiting for a break out in either direction. A $40K vegas wager of sorts betting on a 20% break out. 12% odds based on the post above but that's not the complete picture.
Seems to me if he keeps a running tally of net oscilation profits against the position's liquidation value he may be better off closing out the entire position using some type of calculation:
Net profit target
time in play limit
or whatever best takes advantage of the position.
I understand he's trying to capture the curve... churning trades and scaling up his position while it's covered. Buying on pull backs, scalping small profits... pretty effective plan.
My question is Why hold and play out the hand if you can take a substantial portion of the pot now?
On the otherhand once he's hit a breakout Why exercise the option and close out? Seems he can lock his profits with a stop and churn 10K scalps on pull backs like a mad man until the market checks up.
But I understand he's trying to determine a worst case scenario.
So the danger I see outside of execution FUP's is:
1. Not hitting the 10K share load up to liquidate and just cycling low volume scalps while the options decay. Flat Mountain Syndrome
2. Liquidating a well covered breakout too soon.
3. Trying to eliminate losses in lieu of minimizing losses and time in play.
Ok that's very good. Now the discussion become interesting and I begin to learn from you guys.
I have a question. Your sentence:
"the probability of the stock ending *below* $30 at expiration is about 12%"
I want to understand *precisely* what event is being considered here.
1) "stock ending *below* $30 at expiration"
2) stock arrives at $30 at least once before expiration.
You have mentioned event 1). But I wish to ask because in case you instead meant the second one or they are the same event, that would make me sad.
In fact I have a feeling that the probability of event 2), which is the one more interesting here, should be higher. Am I saying nonsense ?
For this reason I have been asking the code snippet for exercising. Because I don't want to miss the chance while being at the beach
) [no actually I'm kind of on the workaholic side].Very good PocketChange, you are addressing to most important problems here.
If I understand some of your suggestions, perhaps instead of scalping, we could sell position only when we get back to the PUT strike.
Perhaps that would be more effective against the time decay, instead of the scalping (?).
On the other hand we can observe that the reversal necessary to take profit gets bigger. While we are going down and taking a couple of reversal can be better than hoping to do all the way back to strike.
Actually your expression "wrapped his trading range" captures exactly the original purpose of introducing the options. That is to create a sort of "highway" where the robot could operate "protected".
Clearly we pay a price for that (time decay) and probably the is question who is expected to be faster, the time decay or the oscillations (captured by the bot).
One could even employ a hybrid strategy. That is, for a certain time take the oscillations, and later require a full pull back.
What do you think? Better to take the pull backs and close with a small profit each time (it's not just the small profit, as we are actually "realizing" the cost to arrive at that price level), or requiring to go back to the strike?
Also what do you mean by "Seems he can lock his profits with a stop and churn 10K scalps on pull backs like a mad man". Are you suggesting an action alternative to exercising? What do you suggest in particular?
[Another idea could be to overlap various of this games at different price levels]
T
Off the Cuff...
Using your Long Bot Example Averaging Down:
Cover: 35 Put.
Bought:
56 @ 35.53
106 @ 34.03
307 @ 32.53
891 @ 31.03
2,584 @ 29.53
6,056 @ 28.03 mean 28.93
+ oscillations... = X
20% Break out and your trade plan calls to exercise the options for a net of $21K.
Your position is covered and you can lock in profits closing out stops at: (if my math is right)
$21k @ 28.03
$31k @ 27.03
$11k @ 29.03
You've got some room and size to play here:
Sell 3k at 28.13 and buy it back at 28.03.
Keep Churning on .10 pull backs if the market stays flat or in your favor. $3K GP per oscillation... Ride the market down adjusting your mean. Ride it out until you are stopped out or at expiration.
As the market checks up hit it with a $6K sell averaging down as your stop/exit. If the market dips your back in the churn game. If not you should be out with more profits.
You've got a nice tight range to scalp oscillations while protecting your profits: Even A 3 steps average down : 1K, 2K 6K in .30 increments should churn profits while your cover is in place.
What premium is your put fetching? Better to exercise or liquidate?
Time is your trigger:
Break Outs you ride down and churn as long as you can.
Sideways oscillating entries without substantial share scaling you liquidate if X profit target is made before y time has elapsed.
Give it enough time to ferment but not enough to lock you in through decay.
Just ideas
I think it's a good idea to consider the time into the model and keep track also of scalping profit so in case doesn't break out we can still exit without excessive losses.
In you example:
$21k @ 28.03
$31k @ 27.03
$11k @ 29.03
I am not much clear why you indicate the $31k at 27.03. When we break out and we are "even" with the PUT, there should be no interest going further down, because the gain in intrinsic value, will be deleted by the loss of the shares. But I understand your general idea to scalp while holding a good number of shares.
In the meantime, the various insightful suggestions and comments you guys have made me imagine a variant that i expect "intuitively" to be more efficient than the original idea (then we will see in actual tests what is better).
<b>Variant with no "slippage" and increasing steps</b>
(Consider "Long" side.)
Start from the PUT strike and go down fixing
a number of levels.
These levels are not necessarily equidistant (as before), but they may grow much more than linearly while you go down.
Same for the order size: it grows enormously while you go down (as before).
Now every time a *prespecified* level L is reached or surpassed, the bot takes,
by buying, the position predefined for that level (thus averaging down).
In case the price goes back towards the strike, wait until the unrealized become positive
(say > 20$) then assume you are back to level L-k. The bot does not close, but merely sells the
difference of the quantities of Level L and L-k (scalping).
With this variant there is no (much) "slippage" down. Also the bot will not close entirely the position while it reverses.
It will "close" all only with a full reversion to the strike, but in this case it should
have scalped quite a good amount of money.
If it does not break out at least once before expiration or a given time, we may quit when
scalping inside the range has provided sufficient profits.
If it breaks down, exercise on that side and leave the opposite-side option in place (may regain).
Do you see pitfalls in this version ? Or can we reasonably expect an improvement ?
T
ok. I have recoded the robot trade planner to envision the new ideas.
Here is an example. This is realistic (either TWM or SKF).
On any 20% (19,97%) break out before option expiration we get minimum 25.000 $
If there is no break out, the price is "trapped" in the scalping range for a long time and the robot can take care of it, for all this time.
On the LONG side, we have here 3 scalp levels with respectively: 10$, 192.15$, and 2,791.13$
(the 10$ I have assumed arbitrarily. The other are the result of the sell of shares to reduce to the upper
level quantity)
Before option expiration these scalps can be performed several times.
They would add to profit if price breaks out, or else will mitigate losses.
[The probability that those ultra short won't break out at least once should be acceptably small.]
T
Here is an example. This is realistic (either TWM or SKF).
On any 20% (19,97%) break out before option expiration we get minimum 25.000 $
If there is no break out, the price is "trapped" in the scalping range for a long time and the robot can take care of it, for all this time.
On the LONG side, we have here 3 scalp levels with respectively: 10$, 192.15$, and 2,791.13$
(the 10$ I have assumed arbitrarily. The other are the result of the sell of shares to reduce to the upper
level quantity)
Before option expiration these scalps can be performed several times.
They would add to profit if price breaks out, or else will mitigate losses.
[The probability that those ultra short won't break out at least once should be acceptably small.]
T
Try running numbers using OTM Options: 33 put and 37 call.
Should knock the cost of your option hedge in 1/2. Your trade plan already requires a 20% swing to exercise, Inside this range your fighting time decay.
I'd also set a profit target of $5K in net of scalps less options value to liquidate the entire tradeset. You can always renter but if your made 5k in scalps the market is sideways burning time.
finally, your counter short bot is just sitting idle while the long is being drawn down. Can you not scalp along the same increments as the long so you have a net zero position between the two accounts?
Your playing for the market to check up so you can exit and scalp or run down so you can exercise and liquidate. Seems like any activity that can generate transactions inside these parameters would be to your benefit.
If you try applying this strategy to futures you want to be as close to net zero by the end of each trade session.
Food for thought
Should knock the cost of your option hedge in 1/2. Your trade plan already requires a 20% swing to exercise, Inside this range your fighting time decay.
I'd also set a profit target of $5K in net of scalps less options value to liquidate the entire tradeset. You can always renter but if your made 5k in scalps the market is sideways burning time.
finally, your counter short bot is just sitting idle while the long is being drawn down. Can you not scalp along the same increments as the long so you have a net zero position between the two accounts?
Your playing for the market to check up so you can exit and scalp or run down so you can exercise and liquidate. Seems like any activity that can generate transactions inside these parameters would be to your benefit.
If you try applying this strategy to futures you want to be as close to net zero by the end of each trade session.
Food for thought