Thru-Earnings Trades with Calendars

[dbh21 62]

I need a logic/sanity check here...

 

We all know that IV collapses after earnings. I've been trying to figure out of a smart way to consistently take advantage of this by choosing stocks that have consistently show little movement during earnings, and holding calendars.

 

Lets use Home Depot as an example.

 

My earnings analysis charts (posted earlier) show that straddles in HD consistently collapse in value (i.e. all vega and little gamma) and calendars consistently increase in value (although with smaller sample size). 

 

Additionally, I have analyze the term structure of HD's options. This is the term structure for the last earnings cycle showing the Before Earnings structure (BE) and the After Earnings structure (AE) one day later.

 

 

This chart shows that 

1. The nearest weekly options actually increased in IV
2. The biggest drop in IV was around 10-15 calendar days to expiration
3. The difference in the drop tapers off as the term extends out

If I focus on just the difference (i.e. AE-BE), I can compare the last 4 cycles (note there are no weeklies in previous cycles)

 

 

This chart confirms the observations #2 and #3. 

 

Today, TOS tells me the currently market is pricing in a $2.6 move (3.3%). OptionSlam tells me that in the last 2 years, HD's 1-day post earnings moves were on average 1.5% and at max 5.7%.

 

Given the following term structure:

 

I expect Mar to collapse about 3-5% and Apr should collapse 1-2% for a net of 1-3% percent.

 

This leaves a rough P/L as follows:

 

This indicates that there might be a fair amount if cushion, about -5% to +4%, when placing this trade. Even the stock moves 3.3%, it could be still in a profitable position. But if it moves on average of 1.5%, it looks like a 50% gain if IV collapses 3%.

 

I think in this case the sweet spot would really be a Mar1/Apr calendar - but this would require a diagonal as the strikes don't quite align.

 

The downsides I see are

1. This is commission intensive
2. if the stock really moves, this could be a full loss quickly
3. IV could go the wrong way based on special situations or future guidance, dividends changes, etc.

 

 

I'd appreciate any feedback on my analysis or this approach.

 

-D

[Vancouver 54]

So glad you started a discussion on that topic! I was just about to introduce it myself...

 

I'm actually in the position you suggested for tomorrow's earnings announcement... I bought two spreads C80 and P75 (long March 22, and short Feb 28). 

I did this trade a few times already for firms that usually don't move much (WMT, TGT, HD, COST etc.), and have a pretty good success rate so far.

 

I also tried it a few times for firms that move a lot (SCTY, GMCR etc.) and protected it with an out of the money straddle. For example, say SCTY is now at $78, so I would buy two spreads C80 and P75 (or C82.5 and P75), and buy in addition C85 (C87.5) and P70 (P67.5). It is very tricky though, to figure out how far out of the money to go with the protection, and what is the ratio of spreads to straddles. (I usually did 1 straddle for every 10 spreads, but it really depends on the IM of the stock).

PCLN was a good recent example for the profit potential in this kind of trade. The stock didn't move much after earnings, and the value of the spreads went up by 300% or so... Of course, it is extremely risky simply to wish that the stock doesn't move, and protection is needed for high volatility stocks. 

 

In any case, it is a very interesting strategy to explore, and I think we have better odds with low volatility stocks that hardly ever move after earnings.

We'll see what happens with HD tomorrow morning. 

[Kim 7,943]

Incidentally, I was going to post a post about this topic as well, and I intend to start sharing some ideas about future trades.

 

I believe your analysis is fairly accurate. Lets analyze what happened last cycle.

 

HD reported on Tuesday Nov. 19 BO. On Monday close, the Nov. w4 80 straddle was trading at 2.43, implying 3.0% move, with IV at 36%. The Dec. 80 straddle traded at 3.80 with IV at 18% (calls) and 22% (puts). So the DC would cost 1.37.

 

Next day at the close, the stock moved less than 1%, and the DC was worth $2.00, 45% gain. IV of weekly options collapsed 16%, and IV of long options decreased by 3%. At no point during the day was the trade down despite the fact that HD moved 3.3% at some point.

 

However, here are some questions/comments:

 

1. 45% is probably not typical - you would probably sell at some point during the day when the profit reached 20-30%.

2. Generally speaking, you need a strong stomach to do those trades - if the stock opens strong, do you take a small (or not so small) loss, or continue holding, risking higher loss?

3. Like you mentioned, for those low priced stocks commissions will be a big issue.

 

But I definitely will follow up if I see good opportunities, and those who decide to do some trades, please share them so we can discuss.  

[dbh21 62]

So glad you started a discussion on that topic! I was just about to introduce it myself...

 

I'm actually in the position you suggested for tomorrow's earnings announcement... I bought two spreads C80 and P75 (long March 22, and short Feb 28). 

I did this trade a few times already for firms that usually don't move much (WMT, TGT, HD, COST etc.), and have a pretty good success rate so far.

 

I also tried it a few times for firms that move a lot (SCTY, GMCR etc.) and protected it with an out of the money straddle. For example, say SCTY is now at $78, so I would buy two spreads C80 and P75 (or C82.5 and P75), and buy in addition C85 (C87.5) and P70 (P67.5). It is very tricky though, to figure out how far out of the money to go with the protection, and what is the ratio of spreads to straddles. (I usually did 1 straddle for every 10 spreads, but it really depends on the IM of the stock).

PCLN was a good recent example for the profit potential in this kind of trade. The stock didn't move much after earnings, and the value of the spreads went up by 300% or so... Of course, it is extremely risky simply to wish that the stock doesn't move, and protection is needed for high volatility stocks. 

 

In any case, it is a very interesting strategy to explore, and I think we have better odds with low volatility stocks that hardly ever move after earnings.

We'll see what happens with HD tomorrow morning. 

 

I entered the Mar/Apr 77.5 spread for 0.37 earlier to day. Looks like earnings movement was modest, so lets see what the collapse is like. I thought the weekly was too close for my comfort, but I am tracking about 12 different permutations. I'm curious to see how the Mar1/Apr 77.5/78 diagonal does as well.

[dbh21 62]

1. 45% is probably not typical - you would probably sell at some point during the day when the profit reached 20-30%.

2. Generally speaking, you need a strong stomach to do those trades - if the stock opens strong, do you take a small (or not so small) loss, or continue holding, risking higher loss?

3. Like you mentioned, for those low priced stocks commissions will be a big issue.

 

But I definitely will follow up if I see good opportunities, and those who decide to do some trades, please share them so we can discuss.  

 

I followed a few calendars last week, and I found that they were fairly resilient (i.e. no losses), but their prices did fluxuate. That being said, I would likely exit quickly targeting 30%.

 

2. Generally speaking, you need a strong stomach to do those trades 

I like going into expecting a loss as great as the max move in recent history, and keeping sizes smallish. 

 

But... three weeks ago I was going to add a post about a Tastytrade idea I came upon (I don't watch them, but it somehow hit my queue). They called it Sunnyside up, which involved buying ATM bull call spreads and financing them by selling naked calls that had an 84% probability of expiring out of the money (of course on weeklies). If the stock fell you might make a marginal profit, but if the stock rose a bit (like HD right now), it would be tidy profit. Of course, if the stock skyrocket, you were doomed. But they claimed it was consistently profitable. I watched several dozen stocks with this strategy and it was decent, until GMCR... and their agreement with KO caused a 50% gain overnight. I can't remember the exact numbers, but my max gain was like 1k, and it would have ended up with a $25k loss.

 

That was the end of my backtesting and paper-trading the sunnyside up strategy

[Marco 223]

yep I know this strategy as Call-Ladder or this website calls it simply ratio call spread. It's very cheap premium (or you can design it to even give you a small credit) however as you have potentially unlimited loss if the stock sky rockets your broker will charge you margin accordingly. So you should look at your return on margin to evaluate performance and certainly be aware of the risk of the naked call like the GMCR/KO example shows.

 

I like the calendar though earnings trades better. What would be good entry criteria?

 

- implied move > average historical move (which should be the case most of the time actually - so see point 2)

- maybe look at a ratio IM 4% vs. historical 2% (ratio 2) should be better than IM 8% vs. historical 6% (ratio 1.33) , the higher the better (even though a very high ratio would make me suspicious - what is the market pricing in that I don't know of) 

- low max historical move (maybe again expressed as a multiple of current IM, so IM 2% vs. max historical move of 5% = 2.5 multiple, the lower the better)

 

earnings in average are overpriced (IM > realised move) the trick is to survive the massive moves that come once in a while so that excludes naked Straddles/Strangles. condors are very commission consuming so calendars could be the middle ground. Max loss is the premium you pay with losses below 50% if the move isn't way above what the market priced in so if you can average about 30% after comm you can deal with the odd 50% loser and a rare 80%+ loser  

 

So I hope the forum looks into this closer and comes up with a viable strategy/plan for this.

[dbh21 62]

yep I know this strategy as Call-Ladder or this website calls it simply ratio call spread. It's very cheap premium (or you can design it to even give you a small credit) however as you have potentially unlimited loss if the stock sky rockets your broker will charge you margin accordingly. So you should look at your return on margin to evaluate performance and certainly be aware of the risk of the naked call like the GMCR/KO example shows.

 

Actually, it had a different profile. It used a debit spread instead of a naked long call resulting in a much larger cushion for profit:

 

Anyways - way outside my comfort zone.

 

I like the calendar though earnings trades better. What would be good entry criteria?

 

- implied move > average historical move (which should be the case most of the time actually - so see point 2)

- maybe look at a ratio IM 4% vs. historical 2% (ratio 2) should be better than IM 8% vs. historical 6% (ratio 1.33) , the higher the better (even though a very high ratio would make me suspicious - what is the market pricing in that I don't know of) 

- low max historical move (maybe again expressed as a multiple of current IM, so IM 2% vs. max historical move of 5% = 2.5 multiple, the lower the better)

 

 

I'm not sure about comparing the IM or not. A high IM would mean a higher IV, but might also indicate there is a higher likelihood of a big move, no? You are probably right though.

 

My preference is basing it on backtesting, but these are pretty impossible to backtest well with EOD data that I have. The IV drop is easy, but the spreads seem to fluxuate wildly after earnings.

 

I'm also interested in looking at the momentum after earnings. That is do investors pile in and drive the stock higher (like Goog) or does it typically spike and retract. The good thing with further out calendars is there is time if the stock spikes.

 

earnings in average are overpriced (IM > realised move) the trick is to survive the massive moves that come once in a while so that excludes naked Straddles/Strangles. condors are very commission consuming so calendars could be the middle ground. Max loss is the premium you pay with losses below 50% if the move isn't way above what the market priced in so if you can average about 30% after comm you can deal with the odd 50% loser and a rare 80%+ loser  

 

I would actually prefer to use flys, but commissions kill any advantage.

 

FWIW, I close the HD trade for a 18% profit after comm. 

[Kim 7,943]

Stock with higher IM might actually have an advantage.

 

Look at it this way:

 

If IM is 3%, and average historical move is 2%, your margin for error is fairly small - the actual move cannot be less than zero, and it will be probably around 1-2% at least, so the potential gain is usually not too high. In the HD example that I gave - the gain was 40-45% even when the stock moved less than 1% which is very rare. If the stock surprises and moves 4-5%, the loss can be substantial. Taking the same HD, it moved almost 6% a year ago, which was much higher than the IM, and the loss was 50-60%.

 

In comparison, when a stock with high IM moves less than expected, you can easily look at 70-100% gains. And if it moves more, the loss is still limited to 40-50% most of the time. So the risk/reward might be actually better for high flying stocks. 

[Yowster 9,174]

This may be over-simplifying, but wouldn't a good initial screen for this type of trade be to look at OptionSlam data and look for stocks who's straddles typically produce a loss after earnings.  Even if a stock with a higher IV before earnings typically loses on the straddle then it may be a decent candidate since the post-earnings moves tend to stay within the bounds of the implied move.   For example, look at SINA today, there was a pretty big move from 76 to 70 after earnings today but the 72.5 put and 75 call calendars both have higher prices than yesterday at the end of the day.

[Kim 7,943]

Yes, this is actually my main criteria. 

[Vancouver 54]

HD yesterday, and TGT today were great examples of how things can go wrong when you hold the calendar through earnings.

TGT is up almost 7% at this point today. Much much more than the anticipated IM. Luckily, I sold the both spreads at 9:40 or so (when the stock was up less than 3% or so), for a tiny profit, but if I have sticked to it, at this point I would have a 50% loss or more.

 

I wasn't as successful with HD yesterday as I held it for most of the day and had to exit with a 20% loss.

 

Indeed, maybe the risk/return is better with high flying stocks. It will be interesting to watch BIDU after earnings tomorrow. 

[dbh21 62]

I don't think BIDU is a good choice. Look at the straddle chart I made for BIDU... all the straddles finish up on average, implying big moves.

[Vancouver 54]

Yes, I agree, I exited the position yesterday and don't want to buy the spread now at a high price. But it'll be a good sample point to see how wide the tent is.

[dbh21 62]

AZO is a very good candidate. The straddles have all sunk in the last 4cycles and the calendars have all gained value.

 

Here is some quick analysis of the terms structure:

 

The blue points are the IV pre-earnings for the last 4 cycles measured against the number of days to expiration. The blue line is a trend line.

 

The red points are the IV post-earnings. The red line is the trend line.

 

The green points are the current IV.

 

Its interesting that the current front month is above the trend and the back month is below. Assuming earnings are not exciting, this means the front month should have much father to fall, good for the calendar (~9% vs 2%).

[Kim 7,943]

Good analysis. Still, a word of caution: all this is true if the stock doesn't move much beyond the historical averages. If it moves 30 points or more, the trades will still be a loser, although I believe the chances of a big loss are very small (but possible).

[Yowster 9,174]

Looking at PETM as a potential hold thru earnings calendar for tonight.  Straddle usually loses money, and the IV setup seems similar to the successful AZO scenario today.  The only negative is that the spreads are relatively inexpensive (ATM March/April is around 0.45) so the affect of commissions could be a concern.

[dbh21 62]

PETM does look good. Its vol is a little higher given there are 17 days to exp. 

 

 

This chart shows the drop in IV across the last 4 cycles. The last column is the net drop between the front and back month. At 17 days, I would expect the net IV drop to be somewhere between 3.5 and 14 points. But I quickly manually backtested two of the cycles and they didn't look good. The mid prices were higher than they are now, so that might be a factor.

 

I was going to write something up on PLCE, but I didn't like some of what I saw. Straddles collapse, but calendars didn't do all that well (mostly broke even)

[dbh21 62]

FWIW, I just closed my AZO calendar that I opened shortly before the close yesterday for 42% gain. Unfortunately, it was a partial fill opening it.

[luxmon 233]

I put a small position on PETM late yesterday - bought the Mar/Apr 67.5 call spreads for $.42, sold for $.57 at the open (35% gain).  I realize this was a gamble that worked out as the stock didn't move and it was commissions consuming (about 14% of my profits).  

 

Thanks for the analysis and the strategy being discussed.

 

Tim

[Yowster 9,174]

PETM worked out well for me as well with the 67.5 Mar/Apr call calendar - in at 0.43 out at 0.65 for a nice 51% gain (43% after commissions).

 

PLCE looks like another good candidate tonight, but dbh21 mentioned something not looking too good?   The setup looks good in that the straddles typically lose money and the IV differential between March and April looks good as well.  I think I'll either play it with a very small allocation or just track it.

[dbh21 62]

Here is a slightly different analysis on JOY.

 

 

I wrote a program to backtest the change in ATM IV for the different term structures (this is different than the numbers I pull from TOS as they include options that are not ATM (I believe)). Then I also added the performance of buying a double calendar EOD on earnings day and selling EOD the following day. The RV (relative value) is the same as Implied move (the price of the DC divided by the spot).

 

The double calendar I simulated was selling the given front option (for the row) and buying the first option that was at least 30 days out. For example, the first row would be selling the weekly (3 days from expiration) and buying the monthy with 39 days. This is equivalent to selling Mar1 and buying April now.

 

Then I decided to look how each on performed in OptionNet intraday. The interesting thing is in the days that it closed poorly, there was still opportunity during the day (T+1) to exit at a good return.

 

The worst earnings was -10.77% (not in the table above). This would have resulted in a loss of 35-55% on the same type of DC.

 

I might put a small position on.

[Kim 7,943]

Thanks for sharing.

 

The ATM March week1 straddle is trading around 6% IM which is higher than the average move, and JOY rarely moved more than 6% even intraday. So leaving small position for earnings seems reasonable.

[dbh21 62]

PLCE (disregard what I said earlier)

 

ok... so using my new(er) analysis, I get the following data:

 

Nov-14 was a big loss, but the stock moved 14%. I measure the loss anywhere from -40 to -90%. But all the other data points look good. The gaps in the data is because PLCE does not offer weeklies and I couldn't setup a calendar with the front being less than 30 days. But I tested those manually and added notes. 

 

In general, this seems like a pretty good play. I'm going to try to enter a small position

[dbh21 62]

Well, I took a 40% gain on JOY and a 36% loss on PLCE, but it looks like PLCE is at breakeven now. Should have held a little longer

[Yowster 9,174]

I actually eked out a small gain from PLCE.  I played a DC with the 52.5 puts and 57.5 calls (was expecting a little move),  I paid 0.90 for both, and sold both for 1.00 when the stock with within 50 cents of the strike.  I might have made a little more had I waited until the stock was right at the strike, but after such a large move down at the open I was happy to get out with a small gain.