I structure my trades depending on the log(implied move/realized) move. Personally I have never bot a strangle with a clear bias. You can structure them however you want, just make sure you are not over paying!
Great read! Just making sure I have this correct ... this is all about "buying" straddles to trade the volatility ramp into earnings, as it mentions on page 16, "For instance, for the strategy over [-3,0], we buy the straddle on day -3, sell the straddle on the earnings announcement day, and the holding period for this straddle is 3 days." If that is the case I can understand why they go after small volatile companies with frequent earnings surprises.
I am still reading this but there is some confusion on if they are actually holding through earnings. From the article I believe T=0 is the earnings date but they are not accounting for earnings that are released pre-market or after market closes. My guess is their results at T=0 includes a mix of trades that include and don't include an actual earnings move. As a side note, I have been trading these pre-earnings straddles for awhile now. What I failed to realize initially but is now (sadly) very clear to me is how many of these individual straddle trades rely on overall market volatility (VIX). For example, you can lose a lot of money buying one of these pre-earnings straddles 3 days before earnings and if the VIX drops a few points you will lose money on vega, even if the the stock moves a bit for you. I am more interested now in learning how to hedge out market/VIX risk and there were some citations in this paper that mentioned this which I want to dive into next. On the plus side, VIX is currently very low right now so many straddles are cheap compaired to last cycles. Some that I have my eye on right now are: ebay, hog, wmt, mcd
Hey guys, in the paper, they hold through earnings. T0 is the day of the move. T1 is the day after the move. Jon can you give some more colour here? Are these long vol plays? Thanks I REALLY like short GS vol here. But make sure you close the position first thing. It looks like most of the move is during the day!
Yes long vol plays similar to the paper where you long an ATM straddle and close before earnings. Chart of MCD ATM straddles heading into earnings below (we are at T-8) so this suggests closing around T-3 for average 9.7% profit over past 8 cycles: chart is from https://www.chartaffair.com - where I am a beta user - Not my site as I don't know how to code
Jon I am very happy you joined the community. I am learning a lot from your posts and the data vendors you are presenting. Keep posting! I do not know how Market Chameleon calculates their implied move but I am getting a bit different numbers. They most likely use a term structure model which is only good for gamma scalping into events not trading post event. Here are the actual implied move numbers over the last few quarters (using a time series model). Currently, GS this quarter is implying a 3.7% move (considers skew) which is at the top of the range. Secondly regarding gamma scalping into events, be careful with using the "in the past it has returned X% so it looks like a good opportunity". That will almost certainly kill you! I believe steadyOptions promotes a strategy like that. The above data from chartaffair is over fitting at it's finest (we can talk more in detail about this if you want). Instead you wan't to find stocks where the implied move or ambient vol (vol going into the event) is under/overpriced. Here is an example using AKAM (first stock that popped up on vague screen). The stock implied move is only 74% of what it usually is. The ambient vol is a little more expensive than the current realized vol. So maybe a calendar spread would be good here to iso the implied earnings move. Of course you would still need to double check to make sure AKAM did not already pre-release earnings etc...
Thanks. I am learning a lot here as well. Agree about possible overfitting. They have another chart they are called RV (relative value) where RV equals the daily price of the ATM straddle divided by the current price of the underlying. Using this chart helps to visualize how cheap/expensive the current straddle price is relative to prior cycles but since its recalculated each day, it does not take gamma into account. In the chart below, MCD appears to be on a "cheap" trajectory this cycle (red X) but the price still moves down as theta overcomes any benefit from vega increase. You can see a slight uptick at T-2 and T-1 which is consistent with the paper you found. For your AKAM example, are you saying the price of the straddle is only 74% that of the avg. implied move? The same RV chart for AKAM is posted below which seems to show how cheap this is (relative to prior cycles) Agree a calendar looks interesting but where do you set your short leg? I have been burned a few time with calendars in the past (lost 100%) where the stock moved too much prior to the earnings release. A more conservative style would be to place both legs after earnings where past returns look something like this:
Hey guys just want to mention the SWKS calendar Jul19/26. I am in on the 81.5 Cal and 82.5 Cal. The 81.5 cal avg .71 and the 82.5 cal .73 It's the best risk reward trade I have come across in awhile. Just make sure you take off the trade tomorrow morning no matter what. I am in size relative to my account