Weekly Spy Straddle

[Dawajmeister 43]

Looking over the SPY charts, there is pretty substantial movement throughout the week. There are several weeks in a row where SPY has moved $3 or more through the week.

 

Entering a SPY straddle on Friday is generally running about $2, when IV is high it is around $2,13 or so.

 

The weeks that it hasn't moved $3 there has still been up and down movement through the week. 

 

Here it what I'm going to test

 

Enter weekly straddle by end of day Friday

 

To mitigate risk, if the straddle moves $1 up or down, I will roll

 

Always maintain a sell order for $3 or $3 minus credits received from rolling.  I am aiming to see how often this can generate a 50% gain or close to it.

 

Exit no later than the following Friday morning.

 

I will post my weekly results on here, along with any adjustments made. 

 

I plan on entering my first straddle, tomorrow.

 

 

Week 1 - Entered 2.35, rolled for .49 credit, out for 2.20.. 14.5% gain (could have sold on entry day for 2.85) 

Edited November 14, 2013 by Dawajmeister

[Kim 7,499]

Thanks for sharing. I would be interested to see the results. It could be a very interesting trade, especially to hedge our calendar trades. 

 

When you say "if the straddle moves $1 up or down, I assume you meant the underlying?

 

A quick question: why hold till Friday? The last two days have the most theta, I would probably exit no later than Wednesday.

[Dawajmeister 43]

Yes I meant if the underlying moves $1 or more. 

 

Exiting by end of day Wednesday makes sense.  I wasn't thinking about the last 2 days having the most theta.

[QZW 6]

Looking over the SPY charts, there is pretty substantial movement throughout the week. There are several weeks in a row where SPY has moved $3 or more through the week.

 

Entering a SPY straddle on Friday is generally running about $2, when IV is high it is around $2,13 or so.

 

The weeks that it hasn't moved $3 there has still been up and down movement through the week. 

 

Here it what I'm going to test

 

Enter weekly straddle by end of day Friday

 

To mitigate risk, if the straddle moves $1 up or down, I will roll

 

Always maintain a sell order for $3 or $3 minus credits received from rolling.  I am aiming to see how often this can generate a 50% gain or close to it.

 

Exit no later than the following Friday morning.

 

I will post my weekly results on here, along with any adjustments made. 

 

I plan on entering my first straddle, tomorrow.

Sounds interesting. Did you mean sell straddle for credit around the SPY closing price for the next week on Fridays? It would be 2.55-2.67 credit right now.

Edited November 7, 2013 by QZW

[Dawajmeister 43]

Yes exit the straddle no later than close on Friday or now probably Wednesday so as to avoid theta loss for thur or fri. Entry this past Friday was 2.02 so it would've been a nice gain this week

[QZW 6]

Yes exit the straddle no later than close on Friday or now probably Wednesday so as to avoid theta loss for thur or fri. Entry this past Friday was 2.02 so it would've been a nice gain this week

Cool! I will probably do it tomorrow, too. You mentioned you had looked on these trades back... Does the straddle decay in 4 days at all, or you calculate that the stock would stay within the 2X credit window?

[Dawajmeister 43]

No there is definitely decay

[Dawajmeister 43]

If the stock didn't move at all for the week it'd be worth around .7 to .8 at the beginning of Friday due to remaining time... So worst case scensrio if you actually held to Friday and there was no movement at all throughout the week

[QZW 6]

If the stock didn't move at all for the week it'd be worth around .7 to .8 at the beginning of Friday due to remaining time... So worst case scensrio if you actually held to Friday and there was no movement at all throughout the week

Sounds like a good trade! Do you enter different prices for the call/put legs or both always ATM, same price?

[Dawajmeister 43]

I'm planning to do it as an ATM straddle. That's the other nice thing with spy is that it's in .5 increments

[Dawajmeister 43]

Also I'll probably enter first thing fri morning and if there is a big movement Friday I will use one of my day trades to adjust at the end of the day

[QZW 6]

Also I'll probably enter first thing fri morning and if there is a big movement Friday I will use one of my day trades to adjust at the end of the day

Adjust how? If there's a big movement, why not buy the straddle to close outright. This would be a nice daytrade. 

[Dawajmeister 43]

If it's a big enough of a jump I'll close. I'm talking if the underlying moves by a $1 or more by end of day I'll possibly roll to the new strike if the credit is high enough

[QZW 6]

If it's a big enough of a jump I'll close. I'm talking if the underlying moves by a $1 or more by end of day I'll possibly roll to the new strike if the credit is high enough

Brilliant! That is, if you had +$1 move by SPY, you'd roll the entire straddle to the +$1 strike for the same week vertically? 

[Dawajmeister 43]

Correct

[Dawajmeister 43]

It was higher than I expected this morning, but I'm in at Nov  2.35 underlying at 175

[QZW 6]

I did one straddle 175 at 2.36 as well, only for exploring. Should have pulled the trigger yesterday, but that's always the case. We'll see what comes next. Sold PCLN, MA, NFLX shares as well bought on the dip, if you are into shares at all. Yes, also sold XBI on my wife's account. Life is good.

 

P.S. Already got out of the SPY straddle for about $10  loss, LOL

Edited November 8, 2013 by QZW

[Dawajmeister 43]

The ATM straddle was at 2.50 EOD yesterday, you would've started the day down .15.

[Dawajmeister 43]

I am probably going to roll to the Nov 176.50 soon for about a .50 credit, I don't want a pullback the last couple of hours and lose the gains from today.

 

Rolled for .49, net effective of 1.86 now.

Edited November 8, 2013 by Dawajmeister

[bam1960 53]

Volley

 

Are you rolling to strangles or selling ur long straddles and buying long straddles to differnt strike/

 

Thanks

 

Bret

[bam1960 53]

never mind I read the previous post

[Stephen 4]

Volley,

 

Are you always using the front week weeklies? (IE: Nov 15 this week)  Thanks.

[Dawajmeister 43]

Yes I am always going to be using the next week weeklies, the Nov 15 for this week.  

[Dawajmeister 43]

FYI you could've closed for 2.80-2.85 on Friday

[fieldydwb 31]

i've ran some backtesting and can't get this strategy to produce a meaningful return between Jun and Aug.

 

The first backtest I ran was:

 

1. Entering on Friday Morn and closing by Wednesday close

 

Second backtest I ran was:

 

1. Entering the following week upon closing the first week (Closing Wednesday and opening a new one on Wednesday for following week)

 

I was particularly interested in this as a partial hedge to the Steady Condor trades for a large 2008/2011 type movement. If a trade can maintain itself profit wise but also provide a complimentary delta/vega hedge then I am interested.

[Dawajmeister 43]

Did you look at rolling at the end of everyday if the underlying had moved sufficiently for a decent credit or just the price movement of the entry position between Friday and Wednesday ?

[bam1960 53]

I traded the 176 straddle in and out friday for a small gain, but it looks promising.

[fieldydwb 31]

I took profits at 15%+ OR continuously rolled (which cost a lot of commissions). So if SPY straddle was at 165 and SPY hit 166.05 I'd roll the straddle to 166.  Some weeks I rolled several times which may have killed the trade.

 

Edit: I will run Aug to Nov and see how it does

 

If you have specific parameters you'd wish me to include, let me know.

 

PS I rolled the straddle when SPY was +1/-1 not at EOD.

Edited November 11, 2013 by fieldydwb

[Dawajmeister 43]

I was looking at it also and I think I might have to readjust rolling unless the credit is significant enough. Rolling for an .11 cent credit seems 1) commissions intensive 2) could possibly kill the trade

[QZW 6]

I took profits at 15%+ OR continuously rolled (which cost a lot of commissions). So if SPY straddle was at 165 and SPY hit 166.05 I'd roll the straddle to 166.  Some weeks I rolled several times which may have killed the trade.

 

Edit: I will run Aug to Nov and see how it does

 

If you have specific parameters you'd wish me to include, let me know.

 

PS I rolled the straddle when SPY was +1/-1 not at EOD.

Fieldy and Volley, it would be good to see if you can analyze a straddle with different put and call prices setup. I found a straddle for SPY with calls price setup at current price+1% and put setup at current -1% prices to be profitable in the last few weeks. It is sold on Friday and one of the legs bought back next Friday.

[Dawajmeister 43]

Out of the first week trade at 2.20. 14% gain

[Dawajmeister 43]

I might enter at the end of the day if there is a big move.

[bam1960 53]

Volley

 

 

Are you going to enter in the morning if it doesnt move or is it dependent or is it  on what spx does?

 

Thanks

 

Bret

[bam1960 53]

sorry I scrolled up and know the answer

[Dawajmeister 43]

I'm leaning towards entering at the end of the day if it's close to a strike

[indiana*josh 59]

Anyone planning to pull the trigger on the straddle today?  Looks like the Nov22 179.5 can be had for about 2.07.

Edited November 14, 2013 by indiana*josh

[Dawajmeister 43]

In the nov4 179.5 straddle for 2.08

[indiana*josh 59]

I'm in at 2.07, but just a small position as I test the waters on this.

[bam1960 53]

I missed it Gonna try tommorrow,

[indiana*josh 59]

I missed it Gonna try tommorrow,

 

Lucky for you that you missed it.  Price is down around 1.90 this morning.  I may add to my position to average my cost down.

[bam1960 53]

I have a small position @1.95

[dbh21 62]

I thought it might be fun to look at some stats on this strategy as it looks interesting... Just looking at EOD stock data on the SPY from yahoo.

 

First, I wanted to see how often the SPY moves to adjustment points. You say you roll every $1, so I use a threshold of at least one dollar. I decided for simplicity, that you could roll both up and down once per day. That is if the SPY's hi was great than $1 from the open/roll price and the low of the day was less than $1 from the open/roll price, that could be 2 adjustments in that day (technically it could be more, but I only have OHLC).

 

Since Jan 1, 2010, these leaves an average of 3.6 adjustments per week (starting Thurs ending Wed). So thats good - it shows movement.

 

Then I wanted to see how the theta will kill you over the weekend. I wrote a script against my historical option data that bought at EOD friday and sold EOD monday (I only have EOD data so this is the best I could do). The results show an average loss of 4.7% on a straddle over that period.

 

Ideally we want to sell on opening on monday to test just the weekend effect, but since I don't have that data - I analyzed the SPY price movement on monday's to see how much delta/gamma vs theta (we'll ignore vega since these are weeklies) may have influenced that -4.7%. From close of friday to open of monday, the SPY moves on average 4 cents. And on monday's alone it moves on average 0.6 cents. So 4.6 cent movement is very very little, and the 0.6 cent confirms that most of the -4.7% loss is theta.

 

Clearly the weekends are not your friend.

 

But - I don't think this says anything about your strategy as I am not able to test rolling through out the day. We know the SPY can spike 20 points and then retract 15 in the same day. I can't test the ability to roll on that spike and then roll again back down.

 

Anyways, take it with a grain of salt. I need to get some more granular data to look at. Would love to test this at 15 minutes.

[indiana*josh 59]

I thought it might be fun to look at some stats on this strategy as it looks interesting...

 

Thanks for sharing your findings, dbh21.  Just curious, what sort of software do you use for your backtesting?

 

Is there anywhere we can get 15 minute data without having to shovel out a truckload of money?

 

And you mentioned the over-weekend theta.  Do you think there is a significant overnight theta effect as well?  For instance, the straddle lost somewhere in the neighborhood of 5-10% between yesterday's close and today's open without SPY moving much.  VIX is down a little, so some of this is vega.  I would be interested to know how much was theta.  (i.e. in general, and all else being equal, is it better to enter the position in the morning rather than the end of the previous day?)

[dbh21 62]

I write my own software or use excel for simple stuff. 15 minute data? It would be extremely expensive to buy outright. ONE has it - but AFAIK i can't write scripts against it (and I don't like manually stepping through trades). 

 

I can't answer overnight theta. Most major non-planned announcements seem to occur on weekdays after the markets are closed. My guess would be that in general overnight delta would cancel out theta assuming its not a few days from expiration. Vega is generally not a factor for weeklies. It would be interesting to test day trading straddles.

[indiana*josh 59]

How hard would it be to capture data from a live broker data feed (like IB) and warehouse it for backtesting purposes?  We are already paying for that data to be streamed to us, and IB has some sort of API although I've never looked into it.  Would it violate the terms to capture and store that data for selected contracts?

 

Not sure this would even be practical since you would have to decide which contracts to store in advance, but it might be possible to write a script that would capture data pertaining to all contracts within a certain distance of the ATM strike and with a certain expiration window. 

 

I wonder if it would be possible to get historical data this way.  That data is available to the charting tool, but I'm not sure whether the raw data is actually available or if you'd have to try to scrape it from the chart.  Even with no historical data, one could capture data for a few months and then have at least some basis for backtesting strategies.

 

Does anyone have experience with IB's API?

Edited November 15, 2013 by indiana*josh

[dbh21 62]

Well - it depends. Technically its not hard. IB's API is pretty straightforward (although I'm not using it, I've read it). There are three problems though.

 

1. Option data is huge. I estimated a while ago that my historical option data is 4000 times larger than just historical stock data. But this is all strikes. For these kind of strategies - we only need strikes near ATM.

 

2. Broker APIs throttle you heavily. I did test out Etrade's API when I was a client. They heavily throttled my connection and prevented the amount of data I could download or the timeliness of the data. 

 

3. IB requires you to physically login with your token every day - so you can't just host it somewhere to collect data and sleep-in or take vacation.

 

If you are only tracking a few options - its not a big deal. Still work, but no limitations.

 

But I have scripts that work against yahoo's data (not the best source, but its much easier and its there). Have it on my list to start collecting data every 15 mins or so around earnings so I can have better analysis - but I have a long list of TODOs.

[indiana*josh 59]

bam or volley, are you guys still in this?  This position has hit my stop loss so I am thinking to exit today.  Otherwise, the theta is really going to bleed us dry if the index doesn't make a move.

[bam1960 53]

bam or volley, are you guys still in this?  This position has hit my stop loss so I am thinking to exit today.  Otherwise, the theta is really going to bleed us dry if the index doesn't make a move.[/quot

I am at dentist so doing this on phone but I am watching it close but have a small position so kind of just experimenting with it but if it gets below 1.50 will be out

[Dawajmeister 43]

I'm doing this as a test an learn so I will be in the position till end of day tomorrow

[Dawajmeister 43]

I also will not be entering the position this Friday because I expect a slow week next week with the holiday