CMLviz Trade Machine

[Alan]

1 hour ago, Kim said:

This doesn't make sense, especially considering that the long options expire after earnings. The gains might be reduced if the stock rallies too much, but I don't think loss is possible.

Maybe he his referring to the fact that there will be an inversion of the delta difference (DD) between the long and short. When the trade starts out, the DD is +30 (50 Delta - 30 Delta). Since the short is closer to expiration and is on the steeper part of its gamma curve (since it is OTM its gamma increases faster until it gets to ATM whereas the ATM long which will show a slower gamma increase as it is moving ITM), the short option will see its Delta increasing faster than that of the long option. At some point the short's Delta will exceed that of the long option, and there will be a negative DD. At that point further up movement of the stock causes the trade to start losing. It is conceivable on a very big move prior to the expiration of the short, the trade will lose just due to this DD inversion.

Edited December 31, 2020 by Alan

[Yowster]

13 minutes ago, Alan said:

Maybe he his referring to the fact that there will be an inversion of the delta difference (DD) between the long and short. When the trade starts out, the DD is +30 (50 Delta - 30 Delta). Since the short is closer to expiration and is on the steeper part of its gamma curve (since it is OTM its gamma increases faster until it gets to ATM whereas the ATM long which will show a slower gamma increase as it is moving ITM), the short option will see its Delta increasing faster than that of the long option. At some point the short's Delta will exceed that of the long option, and there will be a negative DD. At that point further up movement of the stock causes the trade to start losing. It is conceivable on a very big move prior to the expiration of the short, the trade will lose just due to this DD inversion.

Just as with our NEHS trades (assuming a 1:1 short to long ratio) if the entry price is less than the distance between the strikes then you cannot have a loss on the upside.   However, if the entry price is greater than the distance between the strikes then you could have a loss if the stock price rises too much.

[Fractalfriend]

I had one today with MU. 50 delta long at 71 exp. Jan 8th, 30 delta short at 71.5 exp. today Dec. 31st. Stock went from slightly over 70 to 75 in a few days. The initial cost was $202 but today it was worth only $150 because the short position went up so much. Rare, but it happens when your initial deltas are slightly off from the 50/30 backtest, in this case closer to 48/32 when I opened the position. 

[Kim]

16 hours ago, Fractalfriend said:

I had one today with MU. 50 delta long at 71 exp. Jan 8th, 30 delta short at 71.5 exp. today Dec. 31st. Stock went from slightly over 70 to 75 in a few days. The initial cost was $202 but today it was worth only $150 because the short position went up so much. Rare, but it happens when your initial deltas are slightly off from the 50/30 backtest, in this case closer to 48/32 when I opened the position. 

Yes, you are correct. As @Yowster mentioned, you need to make sure that the initial debit is less than the distance between the strikes and the P/L chart looks good.

Your initial P/L chart probably looked something like this:

 

Instead you could do something like this:



So your longs would be deeper ITM and the short calls further OTM. I'm not sure if this setup gets enough credit, so maybe in case of MU it would be better just to skip the trade.

In case of IBM for example the setup looks better:



But you also have to enter 2 weeks before shorts expiration to get a decent credit. We can watch it and see how does it look a week from now.

[ykotowitz]

Re the recent discussions of trading results of TM strategies.

I have been trading pre earnings momentum calls for the past 6 months. Mostly 14 days pre earnings (with some 7 and 3 days). The strategy involved buying a 30 or 40 delta call for 14 or 30 day expiry and exiting just pre earnings.

I have restricted transactions to cases showing better than 6/2 winning records for the past 2 years with better than 20% average returns. As well, I set exit criteria acording to best results in backtesting, although not always following through-usually to my regret.

There were 90 trades during this period, of which 58 were profitable and 32 loses. The average return was about 35% with a large variance . Several trades generated over 200% in profit and almost 100% losses.  Because of the risk, positions were relatively small, particularly as market movements lead to positively correlated results between trades  over any short period.

In total ,over the past 6 months the profitable trades generated $21,400 profit and the losing trades generated $6500 in losses on a total of $43,200 at risk.

I am very happy with the performance of the strategy. HOWEVER, the period covered was clearly a bull market, so may not be very repreresentative  of the longer term results and risks.

Edited January 2, 2021 by ykotowitz
typo for loses count

[Fractalfriend]

6 hours ago, Kim said:

Yes, you are correct. As @Yowster mentioned, you need to make sure that the initial debit is less than the distance between the strikes and the P/L chart looks good.

Your initial P/L chart probably looked something like this:

 

Instead you could do something like this:



So your longs would be deeper ITM and the short calls further OTM. I'm not sure if this setup gets enough credit, so maybe in case of MU it would be better just to skip the trade.

In case of IBM for example the setup looks better:



But you also have to enter 2 weeks before shorts expiration to get a decent credit. We can watch it and see how does it look a week from now.

Thanks Kim and Yowster, that’s very helpful and will part of my calculations for this type of trade going forward. 

Edited January 1, 2021 by Fractalfriend

[FrankTheTank]

20 hours ago, ykotowitz said:

Re the recent discussions of trading results of TM strategies.

I have been trading pre earnings momentum calls for the past 6 months. Mostly 14 days pre earnings (with some 7 and 3 days). The strategy involved buying a 30 or 40 delta call for 14 or 30 day expiry and exiting just pre earnings.

I have restricted transactions to cases showing better than 6/2 winning records for the past 2 years with better than 20% average returns. As well, I set exit criteria acording to best results in backtesting, although not always following through-usually to my regret.

There were 90 trades during this period, of which 58 were profitable and 38 loses. The average return was about 35% with a large variance . Several trades generated over 200% in profit and almost 100% losses.  Because of the risk, positions were relatively small, particularly as market movements lead to positively correlated results between trades  over any short period.

In total ,over the past 6 months the profitable trades generated $21,400 profit and the losing trades generated $6500 in losses on a total of $43,200 at risk.

I am very happy with the performance of the strategy. HOWEVER, the period covered was clearly a bull market, so may not be very repreresentative  of the longer term results and risks.

Thank you for sharing this.   I was hoping to find the time to backtest some of these "naked" long calls over the winter break but got side tracked with some research on calendar trades.

Have you done any of the call ratio trades?   My hypothesis is that when you normalize for risk just dong the long naked call will be more profitable then messing around with the call ratios. 

 

[FrankTheTank]

On 12/31/2020 at 12:39 PM, Fractalfriend said:

Frank, here are my TM trades just for 14-day Pre-Earnings Diagonals since March 2019. TradeMachine shows these if the stock is over its 50-day moving average. The backtester sells a 30-delta 7-day call and buys a 50-delta 14-day call. In some cases where weeklies don't exist it does this with monthly positions instead. I didn't take every trade suggested by TM, only where the stock hadn't started to rocket up yet before earnings and options liquidity looked good. So the chart had to look close to flat for the prior trading days. Since this was a test, positions are on the small side, that is a few hundred dollars per diagonal.

 

Total trades were 88 of which 64 were winners, or roughly 3 out of 4. Total invested for the 88 trades was $28,505 with a total profit of $1,846. Average position cost was $324 with an average return per trade of 6.5%, or $21, holding the trade for around a week before the position is closed. Biggest loss was $-403 and biggest winner was $327.

 

My conclusion is that, so far, this type of trade is consistently profitable. It's tempting to hold the long 14-day call past the expiration of the 7-day short call if the latter expired worthless and the stock has gone down a bit, but this doesn't tend to help you. If the position isn't profitable after a week, just close the whole thing.

 

Can you share what your largest equity drawdown was?   For example, what was the worst string of losers and how much did those pull down your account? 

[ykotowitz]

2 hours ago, FrankTheTank said:

Thank you for sharing this.   I was hoping to find the time to backtest some of these "naked" long calls over the winter break but got side tracked with some research on calendar trades.

Have you done any of the call ratio trades?   My hypothesis is that when you normalize for risk just dong the long naked call will be more profitable then messing around with the call ratios. 

  

I have not done many call ratios. Backtesting suggests that naked calls  generally generate superior average (not risk adjusted)  profit rates. I am not too concerned about the individual trade risks due to the small investment in each trade and the large number of temporally independent trades.  I did a few call ratios where the costs of a naked call were too large. These were not included in the report. I shall try to dig into these trades and report the results.

[Fractalfriend]

On 1/2/2021 at 12:59 PM, FrankTheTank said:

Can you share what your largest equity drawdown was?   For example, what was the worst string of losers and how much did those pull down your account? 

The largest drawdown for me with this type of trade, 14-day earnings diagonals was in early 2019: 4 trades lost in a row for a total of $835. I think this might have been before TM put in the technical condition of the stock being over its 50-day MA to open this trade. Just eye balling my data, typically I've had a few winners in a row followed by one loser. And again, I'm not taking this type of trade where the stock has already made any sort of significant move upwards in the previous few days or where the price of the whole position is above $600.

[Duesseldorf]

Where do I find the definitions? What does Bull Squeeze, Bear Squeeze, Bull Mammoth and Bear Mammoth mean? I haven't find any information for that... 

[rdr]

the videos are very helpful ! would suggest going through them. You can find the details of the trades by looking at the trade history.

[Duesseldorf]

Sorry, stupid questions. I subscribed TradeMachine® Pro. Now I want to cancel the subsription...

 where can I do the cancellation? I am quite confised:-(

[zxcv64]

@options4winners, after searching, the best I could find was this  (from https://pro.trademachine.com/index.php#discover) :

 

Edited February 24, 2021 by zxcv64

[FrankTheTank]

@Ophir Gottlieb   I was having trouble testing pre-earnings straddles and diving into the trade logs it looks like if the stock reports before market CML is still closing that trade on the earnings date (i.e. after earnings).  See for example this KO test where I selected to close before earnings and the program was closing the trade after the earnings event.  Earnings for KO were on February 10, 2021 before market.  

 

Am I looking at this the correct way?  

 

 

I suspect a 'work around' from the user perspective is to change this to 1 day before earnings for all stocks that report before market open and use 0 days before for stocks that report after close.  However, I was trying to backtest these in bunches so this may not work for me unless I organize my portfolios between groups that report before market and groups that report after market.

Thanks

 

Edited March 16, 2021 by FrankTheTank

[FrankTheTank]

Now that is a nice looking backtest

3 years of pre-pre call calendars (short leg is before earnings - long leg after earnings).  50% profit target.

 

 

 

[ixero@20]

5 hours ago, FrankTheTank said:

Now that is a nice looking backtest

 3 years of pre-pre call calendars (short leg is before earnings - long leg after earnings).  50% profit target.



@FrankTheTank, can you please share the link to this backtest? I am very curious to see how you set it up.

Also, why are you calling them pre-pre calendars? I was thanking that pre-pre calendars are when both short and long legs expire before earnings.  

[FrankTheTank]

2 minutes ago, ixero@20 said:

@FrankTheTank, can you please share the link to this backtest? I am very curious to see how you set it up.

Also, why are you calling them pre-pre calendars? I was thanking that pre-pre calendars are when both short and long legs expire before earnings.  

Yes the naming is weird and I never liked it but at SO we call them "pre pre" when the short leg expires before earnings and just "pre calendars" if both legs expire after earnings but the trade is placed before earnings.

Here is the link. Note that some of these stocks have low liquidity so I don't really trust this: 

https://pro.trademachine.com/index.php?share_key=20210321020801_QuftKcBrrOgYf9hs

 

 

[FrankTheTank]

Here is one for a call ratio backtest (5 years)

https://pro.trademachine.com/index.php?share_key=20210321022035_vpYDuqGNtV9q3wjn

 

 

 

[ixero@20]

Thank you @FrankTheTank! This is very interesting setup indeed.

You enter the trade very early on 21st day before earnings with 14 days roll for the short and 28 days for the long. Which means you start the trade 3 weeks before the earnings date with the short leg expiring 1 week before and the long expiring 1 week after the earnings. In this way the short leg is not protected by the earnings and will decay quicker. And on top of that you close the trade very quickly, just in 1 week to avoid any gamma risk. Based on other TM ER strategies it seems that pre-earning stock price movement begins to materialize after T-10 (14 calendar days) and you get out of a calendar trade just right before that. Smart!

I am very excited to see how these trades will play during the next ER cycle. We just need to hope that these companies will announce the ER at least 3 week before the date. 

And a question if you don't mind - why did you setup this strategy with Calls and not Puts calendars. I re-run this strategy for Puts calendars and return was much worse. Do you know why?

[ixero@20]

On 3/20/2021 at 10:22 PM, FrankTheTank said:

Here is one for a call ratio backtest (5 years)

https://pro.trademachine.com/index.php?share_key=20210321022035_vpYDuqGNtV9q3wjn

I didn't graduate to the ratios yet and personally I still have challenges understanding the benefits of this strategy. I know that downside is capped and even in case of large downside movement it can make money. Similar to straddle it will benefit from the rise in IV and sometimes @Yowster is using it as an alternative to a straddle if stock has upside tendency approaching the earnings. But if I run the backtest for the same list of stocks as you have but for a simple Delta 40 call instead of a ratio, the results look much more appealing. I am using the similar setup as your backtest - open the trade 6 days prior to earnings and close at 0 days, 40 Delta calls with 14 DTE. 

The simple calls will be much more of a rollercoaster but in the long term it should deliver much higher returns (if implemented properly). So while ER Calls and Diagonals are not SO style trades because of larger loss potential, I am not that convinced that personally I should prefer ratios instead. 

[FrankTheTank]

8 hours ago, ixero@20 said:

I didn't graduate to the ratios yet and personally I still have challenges understanding the benefits of this strategy. I know that downside is capped and even in case of large downside movement it can make money. Similar to straddle it will benefit from the rise in IV and sometimes @Yowster is using it as an alternative to a straddle if stock has upside tendency approaching the earnings. But if I run the backtest for the same list of stocks as you have but for a simple Delta 40 call instead of a ratio, the results look much more appealing. I am using the similar setup as your backtest - open the trade 6 days prior to earnings and close at 0 days, 40 Delta calls with 14 DTE. 

The simple calls will be much more of a rollercoaster but in the long term it should deliver much higher returns (if implemented properly). So while ER Calls and Diagonals are not SO style trades because of larger loss potential, I am not that convinced that personally I should prefer ratios instead. 

 

Yes - I had the same thought.   A month or two ago I ran a bunch of backtests manually (using ONE) and compared 40 detal calls to ratios and diagonals.   Sometimes the long calls did better, sometimes the ratios did better.   However, because you can very easily lose 100% of your long call, the risk-reward profile of the ratios looked more appealing.  In other words, you get more unit of profit per unit of risk for the ratios and then you can size your positions accordingly.

I had posted this in the ratio forum and can try to find it later to link it here.

 

[rdr]

pre-pre earnings calendars are the traditional way earnings are played - the problem is they are expensive because the short option is such a low IV relative to the long option so there is a bit more risk on the overall calendar.  I tend to like them and have used them in cases where SO might look at selling the earnings date and purchasing a longer date.  Especially when i'm late and the SO calendar is already higher,.   i have found pre-pre to be more directional- almost like ratios-   you are hoping for a move in the direction of the strike..

[Yowster]

The biggest difference between the pre-pre-earnings calendar (short leg expires prior to earnings) and the pre-earnings (both legs expire after earnings) is the amount of stock price movement that can be tolerated.  A break-even (or even small gain) on a pre-earnings calendar where the stock price has moved near the implied move away from the calendar strike is often a substantial loss on a pre-pre-earnings calendar using that same strike.

[FrankTheTank]

39 minutes ago, Yowster said:

The biggest difference between the pre-pre-earnings calendar (short leg expires prior to earnings) and the pre-earnings (both legs expire after earnings) is the amount of stock price movement that can be tolerated.  A break-even (or even small gain) on a pre-earnings calendar where the stock price has moved near the implied move away from the calendar strike is often a substantial loss on a pre-pre-earnings calendar using that same strike.

Agree but I think it is all relative.  You take on more risk for more reward so if you kept your position sizes smaller you could make the same risk-adjusted returns between the pre-pre calendars and the normal SO calendars.

 

The main issue I have with pre-pre is that VOLHQ does not track or report these in the scanner so there is no way to really screen for them. 

 

 

Edited March 22, 2021 by FrankTheTank

[Kim]

2 minutes ago, FrankTheTank said:

Agree but I think it is all relative.  You take on more risk for more reward so if you kept your position sizes smaller you could make the same risk-adjusted returns between the pre-pre calendars and the normal SO calendars.

 

I'm not so sure about it. In order to risk the same amount, you need to reduce the position sizing by half. Is the average return double? I highly doubt it.

[cuegis]

39 minutes ago, FrankTheTank said:

Agree but I think it is all relative.  You take on more risk for more reward so if you kept your position sizes smaller you could make the same risk-adjusted returns between the pre-pre calendars and the normal SO calendars.

 

The main issue I have with pre-pre is that VOLHQ does not track or report these in the scanner so there is no way to really screen for them. 

 

 

Volhq does have the ability to track these. I had him add it in around 3 years ago.

Just go into Advanced Options and choose "Before" under "Short leg expiry before or after earnings".

 

[FrankTheTank]

4 minutes ago, cuegis said:

Volhq does have the ability to track these. I had him add it in around 3 years ago.

Just go into Advanced Options and choose "Before" under "Short leg expiry before or after earnings".

 

Sorry I meant in the scanner section you cannot filter or screen for these as it only shows calendars with both legs after earnings

[Ringandpinion]

@FrankTheTank I've been modeling some of those "pre-pre" trades.  JPM looked pretty but the tent is is real tight, BE's 148.2/156.8.  That said, the chart looks like it could hold those boundaries but that's speculation.  The TOS probability shadow on the P&L tab puts it at 33% in the tent, not good. 

What time to expiry were you using on the back test.  I would like to model some more, maybe do a couple with single contracts.

[FrankTheTank]

3 hours ago, Ringandpinion said:

@FrankTheTank I've been modeling some of those "pre-pre" trades.  JPM looked pretty but the tent is is real tight, BE's 148.2/156.8.  That said, the chart looks like it could hold those boundaries but that's speculation.  The TOS probability shadow on the P&L tab puts it at 33% in the tent, not good. 

What time to expiry were you using on the back test.  I would like to model some more, maybe do a couple with single contracts.

I could not get the times exactly on my backtest but in general they are opened 2-3 weeks before earnings and I tried to have the short leg expire right before earnings and the long leg two weeks after that.

 

When you are in TOS see what happens to your tent and your BE when bump the IV for the legs up.  Should make it look better but I agree the normal pre-earnings SO trades look prettier but the problem is you have to wait around for theta to come in and sometimes it doesnt and your calendars can just sit there even thought TOS says you should be making $X per day in theta (this is the equivalent of the RV line going flat on you).

 

The pre-pre trades have very nice looking RV charts but come at a risk of too much stock movement.  I still think it is all relative but I am thinking about Kims comment and he may be right that the SO type trades do still offer better risk-adjusted returns even though theta is slower to come in (or sometimes does not come in at all until after earnings).

 

 

 

Edited March 22, 2021 by FrankTheTank

[Ringandpinion]

@FrankTheTank I'll keep working with the "pre-pre" trades and if I decide to trade any, they will most likely be a single.  I'm willing to take the SO approach as a successful given, since it has been so for me.  But, like you, I am always looking for another angle.  And trying new ideas is fun if you don't get stupid.  My wife has a t-shirt with a picture of John Wayne in his Marine getup, the quote is "Life is tough, tougher if you're stupid", I try to live by that sentiment, often successfully.😁

[awwenzovs]

Is there still a special rate to test out CMLViz TradeMachine Pro for forum members?

[Kim]

Whatever you see in the link in the first post of this topic is the current rate.

[awwenzovs]

Based on the website link: https://www.trademachine.com/register/steadyoptions-cml-trademachine-pro/

It's no more $49 per month but $129 per month, am I correct? Anyway to get the discounted rate after joining as a member which I did?

[zxcv64]

2 minutes ago, awwenzovs said:

It's no more $49 per month but $129 per month, am I correct?

Yes, you are correct. I subscribed earlier this year and I remember paying $129/month.

 

[Kim]

The price is set by CML. Early adopters always pay lower prices (we still have members who pay $49 or even $25 per month).

As far as I know, they intend to raise prices to ~$200 area, and it's worth every cent.

[t'pee]

Just joined the service and from what I can tell so far from playing around with it is worth it, very powerful backtesting capabilities on a massive data set, the best part its totally geared at generating actionable trade ideas with alerts to boot, I really like it.

 

Warning: possibly addictive.

Edited July 11, 2021 by t'pee

[minitacotruck]

Can anyone comment on how the results from the recent market has been -- whether backtesting or live trading? 

[Ringandpinion]

On 7/11/2021 at 1:14 PM, t'pee said:

Just joined the service and from what I can tell so far from playing around with it is worth it

Was the webinar made available as a recording?  I signed up but missed it.

[Ophir Gottlieb]

Here is the webinar

https://www.youtube.com/watch?v=RHOAlgpiPbo