Backtesting Help

[cwelsh 1,549]

In times of lower volatility (e.g. now), I'm sure everyone has noticed that the IV earnings trades, straddles in general, and even ICs and RICs don't necessarily perform as well. I know my performance is down, even having a down September (first down month in nine months). So, I start too look for possible other trades, and have come across a strategy that, at FIRST LOOK, seems very promising. However, if anyone is interested, I need help back testing it across different underlying instruments over a full calendar year (preferably three years, but lets start with one).

It's a variation of a double diagonal using either LEAPs, or at least longer dated options. Here are the premises:

1. Have to find a stock that trades LEAPs, and weeklies, with sufficient volume (e.g. AAPL, SPY, T, MSFT, MCD, GOOG, XOM, GS -- those are the 8 I'd like to start testing).

2. You'll eventually enter a double diagonal (I use the DD TOS platform function). First, identify a period out (for testing purposes I'm using at least six weeks, three months, and one year). Go in the money, ideally to delta's around .80 and BUY that.

3. Then SELL the weekly, slightly OTM, option against it, targeting a five percent (5%) weekly return. Roll forward each week. For initial testing assume NO inweek adjustments (we want worst case, if we can improve on that, great).

Here's an example using AAPL.

Assume I were to enter the trade last month on 9/6. AAPL was trading at 676.

I would look at the December options (three months) at .80 deltas and would identify:

Dec 595 Call cost $94.25 and has a delta of .80

Dec 775 Put cost 109.60 and has a delta of -.80

To target the 5% weekly return, I would then SELL the

Sep 14 690 Call for $5.00 (.90)

Sep 14 665 Put for $5.60 (.18)

Then next week (9/13 I buy those back for the amounts in () above and purchase:

Sep 21

695 Call 4.30 (5.00)

675 Put 5.60 (0.11)

Then:

Sep 28

705 Call 5.75 (0.07)

690 Put 5.00 (9.40)

Then:

Oct 3

695 Call 4.30 (0.22)

670 Put (6.20)

So over a month you gross $41.20 and pay to buy back $22.08 for a net gain of $19.12 on a $203.85 investment, or a 9.4% monthly return. And that's WITHOUT any adjustments (you would adjust before hitting the 9.40 loss point).

I have finished running AAPL using options three months forward over the past year, and it worked spectacularly. For MSFT over the same period, not so much (prices too low).

I need assistance back testing AAPL, SPY, T, MSFT, MCD, GOOG, XOM, and GS against options that are six weeks (so Nov), three months (Dec/Jan), and one year (LEAP) and test the performance. That's 24 separate backtesting models. If eight people would just take one stock, we could whip this out by next week and maybe have a good low volatility strategy to complement the IV earnings trades.

I'll take AAPL, because I've already started it. If you want to take one, or are just willing to, please just post which one and PM (or post) the results. Again, assume no adjustments during the week -- just roll on Thursday's.

If you have any questions about the strategy, or don't understand it, please just ask. If you want to pick a different stock (or index) to back test, by all means dive right in.

I will also be opening paper trades tomorrow on all eight of those instruments to monitor going forward as well. I'll be opening 24 paper trades using the NOV, JAN, and as close to one year forward as I can get, always with around a .80 delta.

[Rusty 0]

Sounds like a great strategy to test. I can take GS. I have access to OptionVue which has historical data every 30 minutes. In looking at your aapl numbers, it looks like you rolled EoD on Thursday. I can use that for consistency. What starting date for the test should I use?

[cwelsh 1,549]

1

Edited October 3, 2012 by cwelsh

[cwelsh 1,549]

I've been going back a full year -- to the first Thursday last September.

And yes, I rolled EOD on Thursday in testing. We probably can improve results by sometimes rolling earlier in the day, or on Fridays, or adjusting during the week, but for proof of concept, I start with the simple rolls, adjustments can just improve on that.

[Rusty 0]

I ran 2 tests so far: a 3 month test and a 6 month test. Some initial observations w/ GS:

It is very difficult to find options to sell that will result in a 5% return, especially if GS is on or close to a strike. It's been ~2-3% and 22 or less delta, on average.

Also, the largest loss has been from the long position. When you ran your test did you run it up to the long expiration or did you sell the longs after a few cycles to reduce the theta decay?

[csensen13 1]

I like the strategy in principle and look forward to feedback from others - but just a basic question: what are the margin implications of such a trade - you're long a DITM long-expiration option and short a OTM weekly - is it essentially like writing a covered call or does it have different margin requirements?

[K. Miller 3]

I have tested XOM for 6 months so far. Most months where profitable, but their was a pretty heavy large loss on a major stock surge during a week. The put became worthless but the call lost around 400%. However, this is something that can probably be adjusted a little.

In order to generate decent returns, the strikes where always just out of the money, or at most 1 strike away.

[cwelsh 1,549]

I like the strategy in principle and look forward to feedback from others - but just a basic question: what are the margin implications of such a trade - you're long a DITM long-expiration option and short a OTM weekly - is it essentially like writing a covered call or does it have different margin requirements?

It's like a covered call -- that's why you own the long option positions

[cwelsh 1,549]

I ran 2 tests so far: a 3 month test and a 6 month test. Some initial observations w/ GS:

It is very difficult to find options to sell that will result in a 5% return, especially if GS is on or close to a strike. It's been ~2-3% and 22 or less delta, on average.

Also, the largest loss has been from the long position. When you ran your test did you run it up to the long expiration or did you sell the longs after a few cycles to reduce the theta decay?

There shouldn't be much theta decay -- that's why we are going in on deltas of .80 or more. It's also very easy to get 5% (on avg some weeks its 4% some 6%) -- I must have explained it poorly. Going back one year, here's my first example with GS:

On September 1, 2011 (a Thursday):, GS was at 112.16

The October Options (50 days -- closest available to six weeks out),

Id have bought the Oct 11 95 C for 19.95 (.82 delta)

Id have bought the Oct 11 130 P for $19.00 (-.83 delta)

That's a total of $28.95, so a five percent weekly target is $1.44 (or about .70 each side).

Then I'd have sold the Sept2 Weekly (the numbers in () are what they cost to buy back on the roll the next week):

Sell Sept2 120 C for .46 (.05)

Sell Sept2 100P for .52 (.30)

Sep

115 C .40 (.02)

95P .90 (.02)

Sep4

115 C .44 (.03)

100P .65 (3.20)

Sep5

105C .44 (.06)

85P .70 (.02)

Oct1

105C 1.00 (.06)

90P .77 (.08)

Oct 2

105C .77 (.01)

85P .62 (.02)

At that point, only 8 days to expiration, on the Long, so I would look to roll. The Oct 95 C has lost a ton of value and is at 4.10 but the 130 put has gained even more and is at 34.00 (total of 38.10). So the six week option looks great. Here's the profit break down:

Long position we bought for $28.95 then sold for $38.10 (gain of $9.15).

Over the six weeks we gained $7.67 in revenue at a rolling cost of $3.87 for a net of $3.80. That's a 13.12% return on BEFORE including the gain on the long position. If that's included you get a gain of 44.73.%.

Now granted this is ONE six week window on a year, but right now GS is looking like a great candidate for the six week rolling -- even with that one miserable week of -3.20. (without adjustments).

Edited October 4, 2012 by cwelsh

[srf335 14]

I'll take a run at MSFT if no one else is doing it yet. Will have some time this weekend.

[Kim 7,943]

I wouldn't do low priced stocks, the commissions and slippage on weekly options will kill you. I would concentrate on higher priced stocks, probably in the $80-200 range.

[trhanson 590]

How would we have adjusted the Sep 4 put so that it wouldn't have been so painful? I'm just trying to get a good handle on how to limit those major risks.

[cwelsh 1,549]

Kim's right, I've penciled through low priced stocks -- those just don't work. We need to stick to ones where we can get credits of .25 or more, a couple strikes out.

I'm also noticing some volaitlity swings around earnings that can be accounted for. Basically the ideal time to enter the trade is sometime after earnings (either immediately after or 1-2 weeks after) on the long positions, and don't pick an expiration that expires right after earnings (that also lowers the long position).

We'll also always want to roll with at least two weeks left, otherwise theta CAN make an impact, depending on the instrument.

Also, if there's an earnings period in the middle, just don't sell short that week -- that would eliminate well over half of the "bad" weeks I'm seeing.

As for adjusting rules, not there yet, still validating the basic strategy. As for "how" to adjust, that's easy. If the underlying moves near the short strikes, you just roll. For instance, if stock XYZ was at 100 and you sold the 110 and 90, 3 days later XYZ was at 109. Well you would buy back the original spread (110/90) and sell another. This will STILL be a loss, but it pares it. (so if you got a $1.00 credit originally, you might have to buy it back at $2.00 and then sell the next at 0.50 -- that's a net loss, but still less of a loss).

[hllous 0]

Great topic.

Is it worth trying this strategy on indexes like RUT?

[Rusty 0]

I see... I was trying to get 5% on each side (so, effectively 10%) so I will change that for the next tests.

In your GS example, 19.95+19.00=3,895 (not 2,895) so 5% ~ 1.94 so I will target ~.95 per side.

Also, for your Sep4 buyback, I believe that price should have been 6.10 to buy-back the short put, not 3.10. On 9/22 GS was 93.98, so the 100 Put had to be at least 6.00.

[cwelsh 1,549]

Great topic.

Is it worth trying this strategy on indexes like RUT?

It could be -- haven't backtested RUT yet.

AAPL worked stunningly well, while SPY, over a two year period, did not work near as well.

[tjlocke99 18]

Chris,

I know we had a similar discussion on a separate thread, but I am not sure why you are throwing out MSFT? I have started doing covered calls on it with the weeklies, and the first two weeks I have gotten around .24 credit for a slightly OTM call on Thurs for the weekly. This is a stock that doesn't move too much and is currently trading around $30.

So let's say you can get around a $.45 credit on MSFT for the weekly strangle you would short. Considering the stock is almost 1/4 the price of GS in your example that equates to getting around $1.80 if you multiply the $.45 by 4! Yes you have the higher commissions but the credit is larger AND MSFT maybe less volatile.

[cwelsh 1,549]

Chris,

I know we had a similar discussion on a separate thread, but I am not sure why you are throwing out MSFT? I have started doing covered calls on it with the weeklies, and the first two weeks I have gotten around .24 credit for a slightly OTM call on Thurs for the weekly. This is a stock that doesn't move too much and is currently trading around $30.

So let's say you can get around a $.45 credit on MSFT for the weekly strangle you would short. Considering the stock is almost 1/4 the price of GS in your example that equates to getting around $1.80 if you multiply the $.45 by 4! Yes you have the higher commissions but the credit is larger AND MSFT maybe less volatile.

Well I hadn't backtested it, but some of the other initial testing was showing lower prices wouldn't work -- I could be wrong -- I'll put it back on the list to test.

[fieldydwb 31]

Hi Chris,

Did you make any more headway with this? I've been running some backtesting with optionvue. I'll post some stuff ASAP.

[cwelsh 1,549]

I have made headway, AAPL was a spectacular success story, but with the ups and downs the last week, I'm going to wait till after earnings and backtest this past month too. If it works, then I'll do a live trade on AAPL. MSFT also proved positive, I just have to verify someone's results, but that shouldn't take long.

Other helpful members tested GS, and over the period it was tested, it worked extraordinarily well. I'm half done testing VXX, and it might be the best of all candidates if you stick to the rules. I'll have time this weekend to post a whole set of trading rules around this strategy for different instruments, and I'll start live trading them next week.

[Xfanman 9]

This sounds encouraging Chris, I know we all appreciate yours and the other testers efforts!

[Kim 7,943]

Chris,

I was thinking combining this idea with the earnings trades. For example:

CRM reports on November 20. You buy the December straddle, and sell weekly strangles against it. The earnings should keep the straddle price stable (or at least not lose much) and the weeklies will add some income and mitigate the theta effect.

[cwelsh 1,549]

Chris,

I was thinking combining this idea with the earnings trades. For example:

CRM reports on November 20. You buy the December straddle, and sell weekly strangles against it. The earnings should keep the straddle price stable (or at least not lose much) and the weeklies will add some income and mitigate the theta effect.

That could work very well -- just has to be backtested. I'm going to finish testing this one out first, but some general rules are coming very clear.

1. Ideally target .70-.80 deltas for the long strangle position. It also appears as if 45-120 days out works. Longer requires too much adjustment to the long position, shorter you don't have long enough to build a credit pool if you take a hit on one or two of the short positions. Different instruments have different ideal periods.

2. Limiting losses is key. The rule that keeps reappearing for me is if your short strike is hit, you get out. This will, typically, limit losses to nor more than 2x (NORMALLY not always) of your credit. That means it makes up for itself the next week.

3. Some instruments a 30 delta short position is ideal, while in others you need a 15. For instance, VXX does much better at the 30 delta level, while AAPL was better closer to 15.

4. Always exit the longs when (a) there are two weeks left to expiration, saves a ton on accelerating theta or ( you cannot sell a short against the long anymore. For instance, on the VXX, if you were in the 14/9 (put/call) strangle when VXX was at 12, and the VXX declined to 8 -- well to sell a 30 delta would require a sell price of 8.5 -- or less than your long call, meaning you face a large potential loss. Exit at that point.

5. If your strangle crosses an earnings period DONT sell a short position that week. This single rule makes a larger impact than any other. An earnings price swing can just crush your profits. However, if you haven't sold short, a large, unanticpated, post earnings move can sometimes pay off in spades on the long position. This happened on AAPL numerous times. If you sold short during earnings, you had a losing trade over a six month period. If you didn't, you averaged above a 10% monthly return -- and in one cycle close to 40% when the long positions value blew up in your favor.

Again, I'll develop this in a cleaner version this weekend, as well as put something up on Kim's idea to have volunteers assist with backtest work.

[Kim 7,943]

For earnings plays, 70-80 deltas might be less critical since theta should be at least partially offset by rising IV. So if you go for example with straddle, then a sharp move should hurt less since the long position will benefit more.

[cwelsh 1,549]

For earnings plays, 70-80 deltas might be less critical since theta should be at least partially offset by rising IV. So if you go for example with straddle, then a sharp move should hurt less since the long position will benefit more.

I agree -- but when you are looking to hold the strangle and sell slightly OTM calls against it, you have to have a high delta or theta will destroy you. Your suggested model is slightly different, but as I've thought about it today, and just did XYZ generic theory trades on paper, it seems like it will play out well -- I'll do a whole separate posting for it, and not on this thread, so we can research each independently.

[cwelsh 1,549]

Ok, here are the ones Im going to do:

AAPL (start after earnings)

BIDU (start after earnings)

APA (start after earnings)

XOM (start after earnings)

MSFT (start today)

SNDK (start today)

VXX (start today)

GS (start today)

I'll open an independent thread on them all.

I also have opened an overall volatility hedge, purchasing December VXX 33 calls to hedge against a sudden move that wipes all of my short positions at once. The hedge cost me 20% of the anticipated profits. It might be after close before I get each thread open.

And to clarify, I am PAPER TRADING GS, MSFT, XOM, and BIDU.

I am using real money on VXX, SNDK, APA, and AAPL, as well as the hedge. I'm only committing 50% of my possible allocation to this type of trade for the first six weeks.

Edited October 25, 2012 by cwelsh

[trhanson 590]

In your backtesting, which instruments provided the best returns (and were therefore probably the most consistent)?

[cwelsh 1,549]

The above list, and in this order:

AAPL

VXX

SNDK

APA

(which is why I'm leading real money on those)

XOM

MSFT

GS

BIDU

That above list was assuming no trade management, or only very basic adjustments -- both XOM and GS moved into the top 4 if you quickly cut your losing short positions. And all of the trades improved if you skipped earnings weeks. Basic rule, don't sell short in an earnings week.

[Hannes Kury 7]

Ok, here are the ones Im going to do:

AAPL (start after earnings)

BIDU (start after earnings)

APA (start after earnings)

XOM (start after earnings)

MSFT (start today)

SNDK (start today)

VXX (start today)

GS (start today)

I'll open an independent thread on them all.

I also have opened an overall volatility hedge, purchasing December VXX 33 calls to hedge against a sudden move that wipes all of my short positions at once. The hedge cost me 20% of the anticipated profits. It might be after close before I get each thread open.

And to clarify, I am PAPER TRADING GS, MSFT, XOM, and BIDU.

I am using real money on VXX, SNDK, APA, and AAPL, as well as the hedge. I'm only committing 50% of my possible allocation to this type of trade for the first six weeks.

Chris,

thank you very much for all the analysis and work! I would like to slowly get into this strategy with very small size. You mention the VXX hedge (different from the VXX trade with this strategy). My biggest insurance will be the small size I will be trading at the beginning. However, I would like to use/understand your VXX hedge. Can you elaborate a little more about the hedge? Is there an easy formula? How do you determine the strike of the calls, the # of contracts and the expiration. No quick response needed, thank you!

[cwelsh 1,549]

For the strikes of the calls on the hedge, I use the one closest to .80 delta when I enter. I want fairly close to a $1 for $1 move if the price goes up. I also want to be able for it to lose as little value as possible over time. Also please note that this only protects a sudden move to the downside. It is quite rare to have a sudden "spike" of 400-500 points, but if that did happen, we would be exposed. You could protect against this by having a VXX strangle, but that gets cost prohibitive.

As for "easy formula", no, you have to hedge what you're comfortable with. I vary it depending on how exposed I am to a swift downturn in the market, so its a position which gets adjusted from time to time. If I "only" have this type of trade on, I'll use about 20% of my expected profits over a period to purchase. So, if I was expecting, all things behaving, to make $1,000.00 over one month, I'd spend about $200 purchasing a hedge. This is what I've grown comfortable with -- if you want more protection, buy more.

[Hannes Kury 7]

Chris,

I was wondering if a thread for general questions (hedging, rolling, etc.) about your long dated income strategy would make sense. You could also use/rename this backtesting thread but maybe a general would be better.