today some of you asked me to follow up on this, so here you
go:
Thursday marked 11 consecutive bars above the 3 day SMA. Here's
a scan I ran from 9 bars to 15 from 1995 to present. Note
that 11 bars over only occurred 6 times since 1995, however only
50% of those trades were winners by the time price closed back
below a 4 SMA, and had a max draw down of -1.7%.
However if you wait for 12 consecutive days you have 100%
winning trades and only 4 examples sine 1995, to present so quite
rare. The max draw down was 0.9% so quite low. If the
market closes above the 3 SMA on Friday that would be day 12.
However notice on the stats scan that if price stays above the 3
SMA for a 13th day the condition only occurred 3 times since
1995, and your max draw down moves down from -0.9% to -0.16%, which
is extremely low . Over history the longest price stayed
above the 3 SMA was 14 bars, as you can see on 15 bars there were
no trades
anyway guys that's an idea of things I look at for the SPY and
ES systems. People write in to the website all the time who are not
members and sometimes send scathing comments/questions about the
systems claiming that the systems are worthless because I backtest.
LOL well I hope you see my point here, in order to find rare
statistical conditions like this, you ONLY find them via
backtesting! Backtesting has gotten a bad name over the
years, but if you do a targeting backesting like this, it's very
useful and necessary. I hope this is a good lesson on that
our power went out last night about 2 am, just came back up, so
trying to get situated and it's cold in the house until the heat
catches up. Computers all hard shut off so trying to get them back
up. Yes I have batter backup but that only lasts an hr or so, not 7
hrs.
anyway those 3 SMA stats obviously playing out as Friday was day
12 and 12 days above the 3 SMA had 100% winning trades with a
maximum of 13 days above occurring only once in 40 years
Posted by disciple33 on 30th of Dec 2019 at 11:53 am
Matt, battery backups sometimes have automated shutdown routines
that shut computer down gracefully before a hard shut off. I
imagine you have to set them up so that they don't shut down until
the power has been off for awhile, so that they won't shut things
down during short outages.
no it doesn't work like that. Also with my SPY and ES systems
that I have up on the computers, you can't have to manually close
those programs, no way you can automate something to close those
programs
anyway I ahve to get situated here, luckily I had some nice
shorts from Friday, SPY PUT's, SPXU, LADB, so despite missing he
open by about an hr, I'm having a good day with those
Posted by winter39 on 27th of Dec 2019 at 07:37 am
Thanks very much Matt for all your due diligence. Can't
understand why there are anti backtesters. When you go to a
physician's office they always take a history (kind of backtest).
Reviews are always helpful. Best to you and Steve for the
New Year.
well there are anti backtesters because backesting has been
generally abused and many or most people dont' use it right.
I've seen so many systems out there are simply a collection
of a bunch of indicators and the person runs a huge ramge backtest
on all the variables and while a backtest report might spit back
settings that would have been profitable in the past, if you looked
at the setups, it's really just a bunch of nonsense you would never
do in real life. I think backtest should be targeted to a
logical set of conditions and ranges that you are trying to verify,
that you see visually.
Per request my statistics on the current exhaustion condition that is building
Posted by matt on 26th of Dec 2019 at 11:49 pm
today some of you asked me to follow up on this, so here you go:
Thursday marked 11 consecutive bars above the 3 day SMA. Here's a scan I ran from 9 bars to 15 from 1995 to present. Note that 11 bars over only occurred 6 times since 1995, however only 50% of those trades were winners by the time price closed back below a 4 SMA, and had a max draw down of -1.7%.
However if you wait for 12 consecutive days you have 100% winning trades and only 4 examples sine 1995, to present so quite rare. The max draw down was 0.9% so quite low. If the market closes above the 3 SMA on Friday that would be day 12.
However notice on the stats scan that if price stays above the 3 SMA for a 13th day the condition only occurred 3 times since 1995, and your max draw down moves down from -0.9% to -0.16%, which is extremely low . Over history the longest price stayed above the 3 SMA was 14 bars, as you can see on 15 bars there were no trades
anyway guys that's an idea of things I look at for the SPY and ES systems. People write in to the website all the time who are not members and sometimes send scathing comments/questions about the systems claiming that the systems are worthless because I backtest. LOL well I hope you see my point here, in order to find rare statistical conditions like this, you ONLY find them via backtesting! Backtesting has gotten a bad name over the years, but if you do a targeting backesting like this, it's very useful and necessary. I hope this is a good lesson on that
our power went out last
Posted by matt on 30th of Dec 2019 at 10:25 am
our power went out last night about 2 am, just came back up, so trying to get situated and it's cold in the house until the heat catches up. Computers all hard shut off so trying to get them back up. Yes I have batter backup but that only lasts an hr or so, not 7 hrs.
anyway those 3 SMA stats obviously playing out as Friday was day 12 and 12 days above the 3 SMA had 100% winning trades with a maximum of 13 days above occurring only once in 40 years
Matt, battery backups sometimes have
Posted by disciple33 on 30th of Dec 2019 at 11:53 am
Matt, battery backups sometimes have automated shutdown routines that shut computer down gracefully before a hard shut off. I imagine you have to set them up so that they don't shut down until the power has been off for awhile, so that they won't shut things down during short outages.
no it doesn't work like
Posted by matt on 30th of Dec 2019 at 12:25 pm
no it doesn't work like that. Also with my SPY and ES systems that I have up on the computers, you can't have to manually close those programs, no way you can automate something to close those programs
anyway I ahve to get
Posted by matt on 30th of Dec 2019 at 10:26 am
anyway I ahve to get situated here, luckily I had some nice shorts from Friday, SPY PUT's, SPXU, LADB, so despite missing he open by about an hr, I'm having a good day with those
so today is number 12
Posted by matt on 27th of Dec 2019 at 03:56 pm
so today is number 12
Bought Puts on IWM and
Posted by ssaffer on 27th of Dec 2019 at 04:11 pm
Bought Puts on IWM and QQQ midday. Selling picked up into the close (going off of the 15 min).
fascinating. thanks for your work
Posted by morton7 on 27th of Dec 2019 at 09:14 am
fascinating. thanks for your work on this.
Thanks very much Matt for
Posted by winter39 on 27th of Dec 2019 at 07:37 am
Thanks very much Matt for all your due diligence. Can't understand why there are anti backtesters. When you go to a physician's office they always take a history (kind of backtest). Reviews are always helpful. Best to you and Steve for the New Year.
well there are anti backtesters
Posted by matt on 27th of Dec 2019 at 12:47 pm
well there are anti backtesters because backesting has been generally abused and many or most people dont' use it right. I've seen so many systems out there are simply a collection of a bunch of indicators and the person runs a huge ramge backtest on all the variables and while a backtest report might spit back settings that would have been profitable in the past, if you looked at the setups, it's really just a bunch of nonsense you would never do in real life. I think backtest should be targeted to a logical set of conditions and ranges that you are trying to verify, that you see visually.
agreed.
Posted by ssaffer on 27th of Dec 2019 at 12:53 pm
agreed.
thanks for this
Posted by shecar on 27th of Dec 2019 at 01:59 pm
thanks for this