CML TradeMachine Trade Ideas

[Guest]

Took the GOOGL trade they sent out earlier in the week.  Almost hit -40%, but didn't.  Back to even now.  Any one else take?

[Guest]

On ‎12‎/‎19‎/‎2018 at 2:37 AM, Manish71 said:

Another trade idea from CML. The problem I have found with these trades is that they are theta negative and the short duration of the trade (7 days in this case) makes it very difficult to get a profit unless the stock moves a lot.

They actually have some pretty good scanners which can used to find a variety of trading setups including theta positive trades

 

[Guest]

On 1/25/2019 at 9:13 PM, lrfsdad said:

Took the GOOGL trade they sent out earlier in the week.  Almost hit -40%, but didn't.  Back to even now.  Any one else take?

All their back testing is during low volatility times during last few years. So we have to see how the trade ideas work out during high vol regime.

[Guest]

On 1/25/2019 at 9:43 AM, lrfsdad said:

Took the GOOGL trade they sent out earlier in the week.  Almost hit -40%, but didn't.  Back to even now.  Any one else take?

out +40%.  No doubt aapl post earnings performance probably helped. 

[Guest]

AAPL out with + %60

[Guest]

AAPL still holding with 300%+.....

Rather than wait for the end of the day as they suggest, I jumped the gun and got in at $161.50 in the morning.

I might take 20% off the table, but will hold the rest longer until I get a clear sell signal.

I was already long leading up to earnings but sold at the end of the day ( right before earnings) as I never take the earnings risk.

Then got right back in.

Edited January 31, 2019 by cuegis

[Guest]

To members who use the CMLViz on regularly, are the limitations of this service the following? 1) reliance on EOD prices, not picking up intra-day prices, 2) not being able to constrain the backtests to specified vol regimes (e.g. see how XYZ strategy fared when VIX was above 20).

How often do you trade their weekly recommendations? How often do you set up trades based on your own CML backtesting?

From what I have gathered so far, the service is meant to be a backtesting tool. In this thread in the past; some members said they had left CMLViz, and some others said they like it a lot. I am trying to get a few opinions. I may end up testing the service at least for a while but still would welcome if there are pointers to be aware of.

[Guest]

 

Wow, really depressed about CMLviz after signing up and looking at some basics ... found this:

They have dumped VXX completely ... can't back test one of the most important and interesting symbols.

They can't even get a 6-month back test of SPY stock right ... prices and dividends are wrong.

They don't exercise calls for ex-dividend, even if it's a 100% lock that they would be exercised ... any back tests involving these calls will be badly wrong (show a large loss on the calls that would not be experienced in real trading, unless the long was dumb enough to not exercise).

It's impossible to back test a combo/conversion/synthetic (no capability for enforcing put strike = call strike or call strike = put strike) so can't back test a big range of interesting strategies ... can't even back test boxes!

They don't have dynamic hedging capability ... so 90%+ of interesting back tests are impossible.

There is no filter to specify trades around ex-dividend dates ... lots of interesting stuff there, but impossible with CMLviz.

This thing is a toy, not a real options back test engine ...
 

 

[Humble]

29 minutes ago, trader212 said:

 

Wow, really depressed about CMLviz after signing up and looking at some basics ... found this:

They have dumped VXX completely ... can't back test one of the most important and interesting symbols.

They can't even get a 6-month back test of SPY stock right ... prices and dividends are wrong.

They don't exercise calls for ex-dividend, even if it's a 100% lock that they would be exercised ... any back tests involving these calls will be badly wrong (show a large loss on the calls that would not be experienced in real trading, unless the long was dumb enough to not exercise).

It's impossible to back test a combo/conversion/synthetic (no capability for enforcing put strike = call strike or call strike = put strike) so can't back test a big range of interesting strategies ... can't even back test boxes!

They don't have dynamic hedging capability ... so 90%+ of interesting back tests are impossible.

There is no filter to specify trades around ex-dividend dates ... lots of interesting stuff there, but impossible with CMLviz.

This thing is a toy, not a real options back test engine ...
 

 

CMLviz has thousands of happy customers. This topic alone has over 1,700 posts, where members shared dozens winning trades based on the backtesting.

I'm pretty sure it's more than just a toy.

P.S. VXX has been replaced with VXXB, you can do full backtesting of VXXB. 

[Guest]

Noticed this thread. Criticism appearing that the back tests are based on success in bull markets so recommendations are biased long.

They have quite a bit of content where they test strategies in the 2007-2009 period, the grandaddy of recessions and bear markets of US history. Then there's the tech bubble and bear of 2000, and the covid sell off of 2020. A significant drawdown of 2018. The period 2000 - 2021 has provided a lab for a broad range of investing environments. So there's plenty of periods to test strategies that run into the power saw of bear markets.

It hasn't been all roses.

CML doesn't lend itself to complicated Rube Goldberg contraptions. It leans toward the simple side. But in my investing experience, some of the very best, most endurable strategies are quite simple (not SMA simple, that's not what I'm talking about.) They can be expressed with a few lines of code.

I'm running a strategy (pure equity, no options) that combines two simple metrics and has doubled the S&P for 25 years and it's showing no signs of fatigue. Simplicity can be a good thing, indeed. The caveat of course is it has to be the right kind of simple. Finding the right simple is the hard part.

 

Edited November 16, 2021 by tod

[Guest]

I'm running a strategy (pure equity, no options) that combines two simple metrics and has doubled the S&P for 25 years and it's showing no signs of fatigue

 

Good points about CML @tod Interested in the above comment.  That strategy sounds exceptional. Well done. Can you share the basics. Many Thanks in advance.

Edited November 17, 2021 by mccoyb53
clarification

[Guest]

 

I read some very interesting concepts about this in a book called Skip the Dip. The idea is to have a number of complementary simple auto-trading strategies. For instance, you can have:

• long only momentum trend following strategy with a basket of stocks that tend to move a lot, and exits when it loses momentum. This tends to have low win rate but very good expectancy because they capture big moves. It loses on choppy markets and skips downturns. When a couple of names rip, it covers the tiny losses of the others that are chopping. Works well with shares.
• mean reversion strategy with very selective criteria on a basket of stocks. This one has a high win rate but makes small moves. Pads the equity curve. This tends to work best with credit put spreads but can be done with stocks.
• long volatility strategy, which loses most of the time in a tiny drift and when VIX gets crazy, it makes outsized gains when the other two above are not trading as the entry conditions are not being met. This can be done with VXX or UVXY shares.

You could add here some broken wing butterflies for income when certain technical criteria are met, to pad the equity curve a bit more.

 

After very promising backtesting results, I am now setting up the algos in paper trading, and then move into trading 1 share per ticker on signal. I am curious as to any insights from @tod

 

 

 

 

 

Edited November 17, 2021 by Bullfighter

[Guest]

10 hours ago, Bullfighter said:

 

I read some very interesting concepts about this in a book called Skip the Dip. The idea is to have a number of complementary simple auto-trading strategies. For instance, you can have:

• long only momentum trend following strategy with a basket of stocks that tend to move a lot, and exits when it loses momentum. This tends to have low win rate but very good expectancy because they capture big moves. It loses on choppy markets and skips downturns. When a couple of names rip, it covers the tiny losses of the others that are chopping. Works well with shares.
• mean reversion strategy with very selective criteria on a basket of stocks. This one has a high win rate but makes small moves. Pads the equity curve. This tends to work best with credit put spreads but can be done with stocks.
• long volatility strategy, which loses most of the time in a tiny drift and when VIX gets crazy, it makes outsized gains when the other two above are not trading as the entry conditions are not being met. This can be done with VXX or UVXY shares.

You could add here some broken wing butterflies for income when certain technical criteria are met, to pad the equity curve a bit more.

 

After very promising backtesting results, I am now setting up the algos in paper trading, and then move into trading 1 share per ticker on signal. I am curious as to any insights from @tod

 

 

 

 

 

@Bullfighter if you are looking for auto-trading, you should check out the auto trading platform of Option Alpha.

[Guest]

11 hours ago, Manish71 said:

@Bullfighter if you are looking for auto-trading, you should check out the auto trading platform of Option Alpha.

Yes, I am on their waitlist for a year now. I followed up with them and it's going to take a few months before general availability. In any case, I am doing autotrading for free with Tradestation. It is very economical to rent a Windows machine in Amazon Web Services and run the autotrading fromt there. I am also looking into Quantconnect. They have APIs with Tradier and Interactive brokers and their servers are hosted in Equinix and have very reasonable prices.

[Guest]

@Bullfighter and @mccoyb53

I came upon the system I mentioned while researching sector rotation strategies. The web is full of references to how sector rotation beats the S&P. The same set of papers are dragged out to support this ad nauseum. One by Mebane Faber always gets a mention (I like Mebane Faber), but there's a dozen others; Asness (I like Asness), etc.

I worked for sometime to validate this using the Spyder sector ETFs . I could never show the sector rotation strategy exceeded S&P returns in the long term. Unless they consider occasional returns in excess of 0.1% as outperformance. I could show spurts of outperformance, followed by underperformance, to ultimately end in the same place as simply owning SPY over a 20 year period.

But with the code for testing that (I use R), I started playing around with other things. And I stumbled onto a set of parameters in unrelated "financial toolkits"  that provided a novel entry and exit trigger; and I built a variant rotation scheme that did outperform SPY in a robust manner.

I've found no reference to my happenstance discovery in system strategies, so I'm holding that secret sauce close to the vest for now.  I don't mean to be a tease about this. I brought it up to emphasize the best systems are often simple. But finding them can take years and proves devilishly hard. I spent several hundred hours on what now I run with a few dozen lines of code. Most of that for grabbing data, compiling output. The actual 'secret sauce' is less than 50 lines of code. BTW, this system is not scalable beyond a 7 figure account, which may explain why it lives under the radar.

My systems building philosophy is strongly influenced by a famous math theorem that states: for any continuous curve (any financial time series will do, drawn as a line chart), there exists a polynomial of finite degree that will match the curve to arbitrary precision.

This theorem establishes the downfall of 99% of back testing efforts. A historic timeseries is a fixed, continuous curve. So the theorem above guarantees that with enough indicators and parameter adjustments, you can build a system that matches that time series dot for dot. 

So most back testing efforts are often futile exercises of crafting the (metaphorical) polynomial that matches the data curve they're testing on. The more indicators you're using, the greater the probability you're just building the equivalent of a curve matching polynomial, which holds true only for that one, fixed time series.

As soon as I'm using more than 4 inputs (which is two to many), I get that "uh-oh" feeling.

Of course, a complicated system that works on an in-sample set and proves robust on out-of-sample, or on time series from other classes that weren't used to develop the back test, is legitimate. So simple is not a requirement. But the combination of simplicity with robustness is a beautiful thing. I find CMLViz has some properties like that, so I wanted to defend it a bit.

[Guest]

@tod I totally understand you willing to keep that alpha you found private.

Taylor's theorem is the one you are referring to, and the problem is called overfitting. This why those using complex deep-learning algorithms use strategies to dumb the neural network's learnings, so they work out of sample This is also why I am wary of optimization, Monte-Carlo and walk forward analysis, if used to tune up. What I like use optimization for is to find the range of parameters where the strategy works best and worst, not to tune it up, but to understand its limitations. I also like to break it, using it with tickers and timeframes that I know their behavior, to see how the strategy behaves in those situations, as a way to know what to expect, and how to size it. If I can break the strategy easily, I can dismiss it and move on to something else. If it only works with a very specific, single parameter or narrow range, then I dismiss it because it will not work out of sample. It it is hard to break and the parameter range is wide, then we are onto something.

For instance, I know that a ticker like GRPN is going to have a massive loss, NVDA or TSLA are going to be rockets, /BTC is going to have massive swings, GLD is going to chop a lot, then make a big move, etc. And I know when the market has flash crashes, continued bear markets, and so on. If I know the drawdown in a very bad scenario, I can size the allocation to that strategy. If I have a strategy that loses very little or even makes some on GRPN, but still makes money on the famous ones and others like 

For trend momentum strategies, there's a lot written about selecting stocks with a big move history on the leading pack within the leading sector. You will miss an early entry and therefore bragging rights, but you jump in to a confirmed trend that may continue 35% of the time, and when if does, you make enough to compensate the small loses of the other 65% of times the trend died.

[Guest]

Hello all,

I notice this thread hasn't been touched in years. I wanted to reach out and hear some current/updated thoughts on the CML TradeMachine software. Is anyone still using it? Do you like it? Hate it? Any opinions appreciated, thanks!

Edited March 14, 2024 by J10