I finally got a good rhythm going on my testing for the basket
of ETF's to the SPY and ES systems.
here's a table that shows the ETF basket with current SPY and ES
and the sub systems. I'm placing X in the box where I'm finished.
Obviously the SPY and ES are done.
my goal this week is to finish the QE 3 short. Then I have 20
more to go LOL. I'm happy with how the trades are looking for now.
But clearly will take quite a while to do all these
in the end we could have a basket of 10 ETF's - I would get rid
of ES I think since we won't really need it. The whole point of
adding ES was that it would catch trades sometimes that SPY missed.
However having QQQ, IWM, DIA and the other ETF's to the mix, I
don't think that would be needed as those others would catch plenty
of trades
I've decided for now that I'm keeping these end of day instead
of next day open. I won't be able to have all these open during the
market hrs because of Tradestation's propensity to crash. My
thought at this time would be to open these systems after the close
and issue trade notification for the after hrs market. All of these
ETF's are highly liquid in after hrs market
I would offer up a bit of caution around the basket of systems:
Dont over-optimize or curve fit. What works for SPY/ES should in
theory work for other instruments as well without
over-optimization.. While attempting to “diversify”, you may find
that your signals are highly correlated. Product non-correlation
does not necessarily ensure signal non-correlation. May be
necessary to test correlation risk using TS portfolio maestro
(skills beyond my realm of expertise-lol). Having said what
I’m sure you already know, I look forward to seeing the results of
what I am sure is an extremely time-consuming endeavor!! Good
luck!
(Also- EOD systems on GLD tend to be difficult as the product
rarely closes at it’s highs and lows like the equity markets do. At
least that’s been my experience. I’ve never been able to make a
good GLD/SLV “system” so looking forward to that one
especially)
Posted by DigiNomad on 11th of Jul 2022 at 03:31 pm
System on Matt's Systems? I'm thinking to build a VAR
based model that allows for an allocation to these models. Anyone
already have one? It would be a bit complex and would probably have
to backtest the number of signals generated by the various systems
over the years. For example, if you allocated 30% of a portfolio to
system trades, including option trades, how would you maintain that
average on a VAR basis over time? Maybe 7 or 8% of 30% per system
in order to allow for times when multiple trades are in play?
Just off the top of my head...
Posted by rmoore100 on 11th of Jul 2022 at 12:21 pm
Matt, all I can say to this post is "WOW!!!!!!!" If you
end up with your Systems applied to 10 ETFs, that would be
FANTASTIC !!!!!!!!!!!!!!!!!!!! You and
Steve just continue to do a great job for us subscribers.
I've been with y'all for something like 6 or 7 years.
And i've subscribed to a bunch of investing newsletters, etc. , but
y'all are truly #1 !!
Posted by mastermind on 11th of Jul 2022 at 12:12 pm
I concur that I would rather have signals for other indexes
rather than ES. I think ES/SPY just muddies the waters when they
are out of sync, because you would think that they should be
correlated. SPY vs QQQ are not always so well correlated, so it
might make more sense for them to be biased in opposite
directions.
Posted by mastermind on 11th of Jul 2022 at 01:32 pm
I don't suppose that you are taking votes, but I would
personally prefer XLY to XLP. And I'm fairly surprised that
XLF didn't make the cut considering how big a part of the market
financials are. Maybe because they are already strongly represented
in the Dow (as XLK is in QQQ)?
it will be. but don't expect it next week, gonna take time to
build out - this week my goal is the complete the QE 3 for the 10
ETF's and maybe get a start on another sub system
I like that ES trades can be out of sync with SPY - that's the
whole point, while they do most of the same or similar trades, SPY
will catch trades that ES misses and ES will catch trades that SPY
misses. If they both did the same trades I wouldn't bother
with the ES signals. If you don't have the other ETF's then you
want ES with SPY because you fill in more trade dates that way -
get more opportunities, that's the whole reason I even added it.
SPY might do a trade for 1 or 2 months at times but ES will
generally catch a trade or two during those air pockets when SPY is
stubborn.
with the other ETF's in the mix the ES will probably no longer
be needed for that as they would provide the additional trades to
catch trade opportunities that SPY would miss.
SPY and ES are 99% correlated - but they are way different
prices and ES trades 23 hrs a day unlike SPY and SPX cash. The
systems get their signals from price levels that various indicators
are are at, and when you have different prices the RSI is not going
to be in the same spot on SPY as it is on ES, same for other
indicators, just common sense
Reversion to mean systems basket
Posted by matt on 11th of Jul 2022 at 11:28 am
I finally got a good rhythm going on my testing for the basket of ETF's to the SPY and ES systems.
here's a table that shows the ETF basket with current SPY and ES and the sub systems. I'm placing X in the box where I'm finished. Obviously the SPY and ES are done.
my goal this week is to finish the QE 3 short. Then I have 20 more to go LOL. I'm happy with how the trades are looking for now. But clearly will take quite a while to do all these
in the end we could have a basket of 10 ETF's - I would get rid of ES I think since we won't really need it. The whole point of adding ES was that it would catch trades sometimes that SPY missed. However having QQQ, IWM, DIA and the other ETF's to the mix, I don't think that would be needed as those others would catch plenty of trades
I've decided for now that I'm keeping these end of day instead of next day open. I won't be able to have all these open during the market hrs because of Tradestation's propensity to crash. My thought at this time would be to open these systems after the close and issue trade notification for the after hrs market. All of these ETF's are highly liquid in after hrs market
I would offer up a
Posted by pep8261 on 15th of Jul 2022 at 04:17 am
I would offer up a bit of caution around the basket of systems: Dont over-optimize or curve fit. What works for SPY/ES should in theory work for other instruments as well without over-optimization.. While attempting to “diversify”, you may find that your signals are highly correlated. Product non-correlation does not necessarily ensure signal non-correlation. May be necessary to test correlation risk using TS portfolio maestro (skills beyond my realm of expertise-lol). Having said what I’m sure you already know, I look forward to seeing the results of what I am sure is an extremely time-consuming endeavor!! Good luck!
(Also- EOD systems on GLD tend to be difficult as the product rarely closes at it’s highs and lows like the equity markets do. At least that’s been my experience. I’ve never been able to make a good GLD/SLV “system” so looking forward to that one especially)
System on Matt's Systems? I'm
Posted by DigiNomad on 11th of Jul 2022 at 03:31 pm
System on Matt's Systems? I'm thinking to build a VAR based model that allows for an allocation to these models. Anyone already have one? It would be a bit complex and would probably have to backtest the number of signals generated by the various systems over the years. For example, if you allocated 30% of a portfolio to system trades, including option trades, how would you maintain that average on a VAR basis over time? Maybe 7 or 8% of 30% per system in order to allow for times when multiple trades are in play? Just off the top of my head...
Matt, what a fantastic addition
Posted by tradeit007 on 11th of Jul 2022 at 02:50 pm
Matt, what a fantastic addition to your service.
Very exciting, thanks for all
Posted by greggone on 11th of Jul 2022 at 12:42 pm
Very exciting, thanks for all your hard work on this!
Matt, all I can say
Posted by rmoore100 on 11th of Jul 2022 at 12:21 pm
Matt, all I can say to this post is "WOW!!!!!!!" If you end up with your Systems applied to 10 ETFs, that would be FANTASTIC !!!!!!!!!!!!!!!!!!!! You and Steve just continue to do a great job for us subscribers. I've been with y'all for something like 6 or 7 years. And i've subscribed to a bunch of investing newsletters, etc. , but y'all are truly #1 !!
Thanks Again !!!!!
do you have to re-establish
Posted by watcdy on 11th of Jul 2022 at 01:39 pm
do you have to re-establish a track record now with these new trades?
Or are these operating under same rules as before?
Ditto on WOW
Posted by srusso1 on 11th of Jul 2022 at 12:28 pm
Ditto on WOW
I concur that I would
Posted by mastermind on 11th of Jul 2022 at 12:12 pm
I concur that I would rather have signals for other indexes rather than ES. I think ES/SPY just muddies the waters when they are out of sync, because you would think that they should be correlated. SPY vs QQQ are not always so well correlated, so it might make more sense for them to be biased in opposite directions.
I don't suppose that you
Posted by mastermind on 11th of Jul 2022 at 01:32 pm
I don't suppose that you are taking votes, but I would personally prefer XLY to XLP. And I'm fairly surprised that XLF didn't make the cut considering how big a part of the market financials are. Maybe because they are already strongly represented in the Dow (as XLK is in QQQ)?
XLY Components Please understand that XLY
Posted by steve on 11th of Jul 2022 at 01:52 pm
XLY Components
Please understand that XLY is heavily concentrated - Just 3 issues make up 50% of the Sector. AMZN TSLA and HD
OK, thanks for that info.
Posted by mastermind on 11th of Jul 2022 at 01:56 pm
OK, thanks for that info. I didn't realize it was that concentrated.
Fantastic addition to the service,
Posted by fundamentalvalues on 11th of Jul 2022 at 02:46 pm
Fantastic addition to the service, Matt!
it will be. but don't
Posted by matt on 11th of Jul 2022 at 02:57 pm
it will be. but don't expect it next week, gonna take time to build out - this week my goal is the complete the QE 3 for the 10 ETF's and maybe get a start on another sub system
I like that ES trades
Posted by matt on 11th of Jul 2022 at 12:17 pm
I like that ES trades can be out of sync with SPY - that's the whole point, while they do most of the same or similar trades, SPY will catch trades that ES misses and ES will catch trades that SPY misses. If they both did the same trades I wouldn't bother with the ES signals. If you don't have the other ETF's then you want ES with SPY because you fill in more trade dates that way - get more opportunities, that's the whole reason I even added it. SPY might do a trade for 1 or 2 months at times but ES will generally catch a trade or two during those air pockets when SPY is stubborn.
with the other ETF's in the mix the ES will probably no longer be needed for that as they would provide the additional trades to catch trade opportunities that SPY would miss.
SPY and ES are 99% correlated - but they are way different prices and ES trades 23 hrs a day unlike SPY and SPX cash. The systems get their signals from price levels that various indicators are are at, and when you have different prices the RSI is not going to be in the same spot on SPY as it is on ES, same for other indicators, just common sense
Sounds like you will no
Posted by keyhole7 on 11th of Jul 2022 at 12:40 pm
Sounds like you will no longer give /ES trade call outs. I hope that is not the case because that is my trade of choice.
I also like that ES
Posted by steve on 11th of Jul 2022 at 12:28 pm
I also like that ES can provide alternative trades - best way to think of the systems is as follows: ALL SYSTEMS ARE MUTUALLY EXCLUSIVE
We appreciate the work you
Posted by steverobin on 11th of Jul 2022 at 12:12 pm
We appreciate the work you are doing here!
Looks like a great plan
Posted by iaranger on 11th of Jul 2022 at 12:06 pm
Looks like a great plan !