Earnings Straddle - Why Not Buy Way Ahead of Time?

[qlai]

I was almost always long vol.

I can't say it enough, with options, if you see free money, don't get happy - check your shit

Ok, so may I ask how you were making money with options then? Were you finding mis-priced options? We're you taking other side of someone's trade when you felt they were mis-pricing? We're you simply hedging portfolio and were willing to pay fair value for it? I assume you were NOT simply making directional bets.

I vaguely remember reading in the Market Wizards book when the guy was on the floor market making options. He was famous for killing it during 87 crash. He was saying, if I understood correctly, that while he was mainly market making, he would be assembling long vol positions for his own account in the process that he was able to get close to free.

 

[Wheezooo]

"Were you finding mis-priced options?"
"We're you taking other side of someone's trade when you felt they were mis-pricing?"

98% of the time, if you were lifting my offer, or hitting my bid, by definition you were mispriced. The other 2% of the time, I was about to get run over, but didn't know it.

"We're you simply hedging portfolio and were willing to pay fair value for it?"

Hard to answer, I always tried to hedge everything, and at the same time there were levels I felt the need to take a stand. Although I don't remember ever taking that stand from the short vol side.

"I assume you were NOT simply making directional bets."

I would like to say never, but reality is always. Movement, non-linearity, and multi-variable models kinda does that to you. You always have some open or expanding position you need to decide what to do with. I sometimes let those 'residual' positions get a bit out of control.

 

[TheBigShort]

Hey guys, I thought I might add some more intuition here. What I did was strip out the event vol from the implied earnings move. Each unique coloured line is the event vol in %, 21 days before the event. The data is on Facebook. Each colour is a different earnings event.
As you can see most of the time the event vol is pretty steady. There is one time (aqua) where the event vol spiked going into the event. (Day21 is the day before the event, Day 1 is 21 days before the event).

 

[qlai]

So does this mean there was no volatility pop except for one time? If so, these straddles should suffer from theta.

 

[TheBigShort]

So does this mean there was no volatility pop except for one time? If so, these straddles should suffer from theta.

It always suffers from theta.

Only 1 time (aqua) you made considerable Vega gains.

 

[qlai]

Right, so originally I got "excited" because it sounded like vol pop was a given and I thought that even if it was just enough to counter weight theta, that would be pretty nice - lots of break evens, some small winners, and occasional home runs.
Do you think that if you do your study on large sample, you may see that the vol pop actually happens more often - the opposite of what you found with FB?

 

[TheBigShort]

Do you think that if you do your study on large sample, you may see that the vol pop actually happens more often - the opposite of what you found with FB?

Yes, there are opportunities out there. But you have to find a way to filter through the data. Taking a look at my screener FDS could be a potential log straddle. However, the options have 26 days to expiration, so event volatility is not a dominant factor.

Right, so originally I got "excited" because it sounded like vol pop was a given and I thought that even if it was just enough to counterweight theta, that would be pretty nice

Your vega does not counterweight your theta. Gamma counterweights your theta.
The ideal situation is event vol is underpriced & the ambient implied vol is fairly priced. They don't come around often but they do show up here and there. Quite often event vol is underpriced (not by a large amount) a few weeks out, however, implied ambient vol is expensive. You can isolate the event vol using a calendar spread. But that opens up some more risks.

 

[ffs1001]

Each unique coloured line is the event vol in %, 21 days before the event. The data is on Facebook. Each colour is a different earnings event.
As you can see most of the time the event vol is pretty steady. There is one time (aqua) where the event vol spiked going into the event. (Day21 is the day before the event, Day 1 is 21 days before the event).

Interesting chart.

That aqua line looks like it was the 30-Oct-18 earnings cycle, in which case the rise in the line can be explained by the general market volatility rise. VIX had gone up that month as the SPX started to wobble.

From my personal experience of trading straddles leading up to earnings, I have found that :

1) vega rarely makes up for the theta loss (this FB chart is a perfect example of an exception).

2) It's the movement in the underlying that gives the profit, and if this doesn't happen, then vol increase will just reduce your loss at best.

3) For most stocks the vol ramp happens in the last few days, and not weeks before the event. So, buying early means you are victim to the theta, but the volatility doesn't come to your rescue just yet.

4) The earlier you buy (say 4 weeks before event), the lower your delta and gamma, so the large the stock needs to move to make a profit.

5) Long straddles can be a nice little 'black swan' portfolio hedge. If you are lucky enough to have some of these trades open at the time when the SPX takes a dive (Feb-18, Dec-18 etc), then you are sitting on very decent profits as the gamma and increased vega both echo the 'Ker-ching' sound in the background.

Happy trading.

 

[qlai]

1) vega rarely makes up for the theta loss (this FB chart is a perfect example of an exception).

K, that is what I was trying to get to. Thank you for sharing your experience. And thanks to everyone for contributing! ... Turned out to be a pretty good thread I think.

 

[TheBigShort]

guys, if you are trading theta vs vega you are trading it wrong. The goal is to isolate vega exposure.

The strategy is broken down into 2 parts.
part 1) Will your daily gamma gains cover your theta bleed. This is equal to your forecast realized vol minus the implied ambient vol multiplied by your vega exposure.
part 2) Will the event vol increase or decrease. Assuming your gamma gains equal your theta losses, if event vol increases, you make money if it decreases you lose money.

So you need two forecasts, realized vol and event vol. It's simple when you break it down and you know what you are looking for.

From my personal experience of trading straddles leading up to earnings, I have found that :

You trade calendars into earnings don't you? That means your taking a view on Forward Vol leading up into the event. In my opinion this is quite odd, however I have not looked into it.

Lastly, I would say "theta bleed" is not a bad thing. Being short theta means you are long gamma. If your forecast realized vol is higher than the implied vol, I will be short theta everyday of the week.