Leaderboard

I am still holding ADI but that's it. I think there have been more losers than winners lately for me on the P.E. momentum trades -- need to update my spreadsheet this weekend. But it's no surprise that these plays really are creatures of the overall trend - with some exceptions.

This answer most certainly depends on how much confidence you desire, or "alpha level". Getting into this is beyond the scope of the thread and my expertise for sure. Just to throw out a broad brush answer, in the research studies I've done and if the sampling plan is unbiased (a huge assumption in this analysis since it's only the last 8 events, not 8 randomly sampled from all earnings events in the entire population), I've found to that in order to test significance of a single parameter (i.e., calculate a "p-value" in the vernacular), I've needed at least 20 samples for estimating each parameter in the model. And this is just to check the significance of the main effect of each parameter (not interactions between those parameters which add another level of complexity - say if entry date is set to one, does the significance and sensitivity of the optimum short delta change, etc). Again, my main point was to not read so much into the precise settings of these studies and slap that particular trade on blindly, but consider what's going on in IV, RV, and possibly technical analysis that's fundamentally leading to these undoubtedly strong trends that appear to have good edge. Tim

@cuegis You raise a good point... and illustrate why you should always look in the trade details to see what exactly went into the backtest. In the TWTR example, the 4 trades in the 1 year backtest had: collect 0.51 credit on short put spread witdh of 2.5 collect 0.75 credit on short put spread width of 3 collect 0.32 credit on short put spread width of 2 collect 0.12 credit on short put spread width of 2.5 While the first 2 trades make some sense, would anybody really want to trade the last two where the credit received is such a low percentage of the spread width??? One losing trade will take multiple winners to make up for - and why if you extend the backtest to 2 years it comes up a loser. BTW, @Ophir Gottlieb when I extended the test to 2 years I think I found a bug. Look at these trade details: Date Desc Size Symbol Price PNL Stock 30-Jul-15 Open -1 TWTR Aug28`15 $1.05 $31.47 DaysAfterEarnings 31 Put Short Puts 30-Jul-15 Open 1 TWTR Aug28`15 $0.21 $31.47 DaysAfterEarnings 27.5 Put Long Puts 26-Aug-15 Close 1 TWTR Aug28`15 $6.62 ‑$559 $25.03 DaysAfterEarnings 31 Put Short Puts 26-Aug-15 Close -1 TWTR Aug28`15 $2.46 $225 $25.03 DaysAfterEarnings 27.5 Put Long Puts The opening trade collected 0.84 credit on width of 3.5. The closing trade paid 4.16 to close the spread of width 3.5 - this is obviously wrong. Deep ITM options have wider bid/ask spreads but in practice you should never have to pay more than the width of the vertical spread to close it.

My 2 cents on this: I'm an engineer and armature user of statistics based models in my day to day work for using past data to predict future data (and yes, it is probabilistic as Kim mentioned). To me, this is most certainly "curve fitting" here, or if you don't like that term, he's come up with a model for prediction of profitably playing post-earnings events in TWTR by selling put spreads. The obvious parameters of the model are option expiration, short option delta, long option delta, entry date, and exit date- so 5 parameters. If he's training that 5-parameter model based on 8 data points (2 years of earnings events), that's n-1 degrees of freedom, or 7 (n-1 since it doesn't include the entire population of earnings events). In analysis of variance (stats 101) for regression modeling, and assuming your errors are normally distributed, you use up one degree of freedom for each parameter, and the rest are left over for calculating error (i.e., confidence intervals for each parameter and error in the overall model). So, with only 3 degrees of freedom left over for computing confidence intervals, it is very unlikely the interval would be very tight around any one parameter. Hence, that's why I'm very skeptical the confidence interval around entry date is less than +/- 1 day. So basically, in my opinion, it's wise not to focus too much on the specifics of any one setting CML is predicting based on such a small sample size (as a previous poster mentioned), and just focus on what's actually going on here: implied volatility in the options is in reversion to the mean after earnings announcement to a level that still provides edge over the realized volatility in this time frame, and there's a slight bullish tilt to price (again, for a very small sample size).

To me, this, most certainly is curve fitting, as well as, some of the other posts. Kim, you have endorsed this platform, can you give us some of your thoughts on these trade ideas?

On the TWTR study, why pick entering 2 days after earnings? Sounds suspiciously like a curve fit parameter with limited robustness, especially if entering say 1 day or 3 days after earnings don't give similar results. But, the tool does look powerful for certain purposes as long as robustness of settings is verified.

I would be very cautious about this study. You need a number of occurences of at least 30-40 to claim any significant result. With an n of 4 your evidence to support a claim of investing right after earnings in TWTR is fair, at best

Lots of good questions coming up, as I've noticed some nuances as well. I've played around with the tool quite a bit (and am looking forward to support for calendars and custom trades as a large portion of my trades have the calendar and diagonal structures). It's powerful and easy to use, but don't take the stats in the summary window at face value. The summary is a great starting point but download the trade details into a spreadsheet for a deeper dive - you may have to play with the data to get the backtest data to fit your needs. A few things to lookout for: Unequal dollar allocations for each trade iteration. Your only tuning knob is to specify how many contracts in the settings. As stock prices move and if you include earnings periods (when options prices are higher), for some stocks the dollar amount for each trade iteration can differ by factors of 4x or more - which means those trades count a lot more to the overall averages. I like to see equal weight for each iteration so I wind up using the trade details to get a percentage gain/loss for each trade iteration and then average them all out. What option series are selected for each trade (I see some questions regarding this). I have to play with that Rollover value to fully understand how it works in relation to what option expirations are selected for the trades. For credit spreads, look at the credits received. For some of the lower IV stocks when you exclude earnings and get farther OTM with your strikes, sometimes the amounts collected can be so low that you wouldn't trade them in a real trade.