I wanted to share with everyone some things I have been
working on with the SRS Swing 2 System. I wanted to
first say that I have only back-tested these results to May 1,
2009. I can almost guarantee that my improvements will not
hold up when there is increased market volatility like there was
before May 1st. The percentages listed below would have
to be increased to account for the increase in market
volatility. The only downside would be that one would not
make as much in gains.
On to the improvements! I have implemented 2
things: Protecting losses and protecting against
whipsaws. Now these improvements are not 100% but it
does reduce each of these areas. As you can see in the
comparison charts below, I have reduced the losses of 190k to only
90k. This has increased the overall profit from 76k to
112k. The only negative I can see on my improvements is that
the Percent Profitability has dropped from 50% to 30%. This
is because of all the breakeven trades.
Settings:
---WIPSAWS------the past couple weeks we have had several large
whipsaws including today's action. V-type action is what
doesn't work well with these systems. For my settings I took
the average between the Fast and the Slow moving averages. I
then tried to find the percentage above and below the MA's
where a trade does not become profitable to take. I
back-tested this and came up 3.2%. So when the current price
of the closing bar MA cross is 3.2% or higher than the
averaged MA's price then I don't take the trade.
---Breakeven stop-loss-----I then applied a breakeven
stop-loss. As soon as the price of SRS is 2% above the entry
price I applied a breakeven stop-loss. This reduces the
amount of profitable trades but greatly increases the total dollar
amount of profit.
I welcome comments, concerns. Any additional ideas would
be greatly appreciated.
thank you for your comments rank. I like the overall idea
of setting stops to break even when you are up a certain
percentage. However, it seems all too many people know about
the parameters for the 15min srs strategy. I discontinued my
use of it today. After I shut it down I ran a simple
optimization report and it came up with 3,34 instead of 9,39,
It has a profit factor of 1.97. I temporarily turned it on to
watch it trade, and will probably give it some money once we get
into the sp 1060-1070 area. srs in general just seems to be
dead and manipulated here, its not trading as well as other stocks
and can quickly take away your profit (like it did on
friday). just my thoughts...
Nice work rank! Perhaps (?) if the % from the average band was
based on ATR (or ATR% meaning ATR as a % of price) it could move
with the market volitility.
In the meantime since Ninja code is beyond me and I don't use TS
I wanted to set something to keep an eye on using Strategy Desk.
Here is a chart with "Overbought" and "Oversold" bands above and
below the price and MAs. The chart ends 5/27 in order to show a
trade that should be eliminated (blue arrow). The crossing signals
are indicated in the bottom pane (hard to see the little
backtesting symbols and/or the MA cross). This trade was selected
because it was the worst in the time period and met the conditions
to be eliminated assuming I'm understanding what rank is doing. I'm
not worrying about the profit target or the stop even exit for
now.
Here are the formulas for the bands if anyone wants to set them
up (parameters are replaced with xx):
Posted by davetrader on 3rd of Nov 2009 at 08:34 am
I just got off the phone with StrategyDesk and there seems to be
no way to automate these settings. He said the program would
need to be able to do a if/then caculation and currently SD is not
able to do that. He said that as the ThinkorSwim trading
platform developes that this may be posible, but that will not be
for several months.
I would appreciate comments if anyone is able to set this up in
another trading platform.
its pretty easy in tradestation as long as you know how to use
Easy Language.
all the settings above I have already automated into my
strategy. If you switch to trade station and have any
questions on how to set this up I would be glad to help
Posted by mansilvcas on 3rd of Nov 2009 at 07:11 am
Thanks a lot. Could you share the partial results of each
improvement, so we can see how each affect the profit factor
and the percent profitable? Many thanks
rank10 -- very nice equity curve. Thanks for the work and
the ideas. Of course, the lower percentage profitable doesn't
mean anything -- its just the breakeven trades being added
in. winners to losers are almost 50%. Will be
interesting to do more bakctesting and optimizing!
Posted by davetrader on 3rd of Nov 2009 at 06:28 am
Nice improvements, thanks for sharing. Anyone have any
ideas how to write this formula for SrategyDesk for these
changes? I have a few ideas how to write it but need some
help, I plan on trying to call Ameritrade this morning and see if
they can offer any help.
Thanks, really nice. I tried improvements with breakeven stops
and no entering with high stochRSI but did not succeed to get
long term results much better. The MA% difference is a nice
angle.
IMO with your improvements the % winning vs % losers is
the same (about 50%) if you take out the even trades. Don't see a
negative there.
The profit factor is higher which makes it more attractive.
Until I have TS running as second system and third broker
(risk reduction) I will try to code it in my NT stategy.
Of course he cut the losses in half, don't forget that. So, a
few more losses here and there, but they're smaller than what we're
used to. SRS is still a good system as it stands though.
Posted by mitchellvisa1 on 3rd of Nov 2009 at 12:09 am
This is great IMO. One of the things I was discussing with
a fellow full time trader of these systems today is when the MA's
have large whitespace between them on the actual chart we get
losing trades. This white space I describe is exactly the
same as what you are talking about...significant spread betweem the
actual price and the MA. I believe a mean reversion
program could do well on SRS in this current environment.
Thanks for the insight, will try to post other ideas tomorrow.
Newsletter
Subscribe to our email list for regular free market updates
as well as a chance to get coupons!
SRS Swing 2 mechanical system possible improvements
Posted by rank10 on 2nd of Nov 2009 at 10:30 pm
I wanted to share with everyone some things I have been working on with the SRS Swing 2 System. I wanted to first say that I have only back-tested these results to May 1, 2009. I can almost guarantee that my improvements will not hold up when there is increased market volatility like there was before May 1st. The percentages listed below would have to be increased to account for the increase in market volatility. The only downside would be that one would not make as much in gains.
On to the improvements! I have implemented 2 things: Protecting losses and protecting against whipsaws. Now these improvements are not 100% but it does reduce each of these areas. As you can see in the comparison charts below, I have reduced the losses of 190k to only 90k. This has increased the overall profit from 76k to 112k. The only negative I can see on my improvements is that the Percent Profitability has dropped from 50% to 30%. This is because of all the breakeven trades.
Settings:
---WIPSAWS------the past couple weeks we have had several large whipsaws including today's action. V-type action is what doesn't work well with these systems. For my settings I took the average between the Fast and the Slow moving averages. I then tried to find the percentage above and below the MA's where a trade does not become profitable to take. I back-tested this and came up 3.2%. So when the current price of the closing bar MA cross is 3.2% or higher than the averaged MA's price then I don't take the trade.
---Breakeven stop-loss-----I then applied a breakeven stop-loss. As soon as the price of SRS is 2% above the entry price I applied a breakeven stop-loss. This reduces the amount of profitable trades but greatly increases the total dollar amount of profit.
I welcome comments, concerns. Any additional ideas would be greatly appreciated.
thank you for your comments
Posted by cwa82675 on 3rd of Nov 2009 at 12:00 pm
thank you for your comments rank. I like the overall idea of setting stops to break even when you are up a certain percentage. However, it seems all too many people know about the parameters for the 15min srs strategy. I discontinued my use of it today. After I shut it down I ran a simple optimization report and it came up with 3,34 instead of 9,39, It has a profit factor of 1.97. I temporarily turned it on to watch it trade, and will probably give it some money once we get into the sp 1060-1070 area. srs in general just seems to be dead and manipulated here, its not trading as well as other stocks and can quickly take away your profit (like it did on friday). just my thoughts...
Title: rank10's SRS work and
Posted by bkout3 on 3rd of Nov 2009 at 11:57 am
Nice work rank! Perhaps (?) if the % from the average band was based on ATR (or ATR% meaning ATR as a % of price) it could move with the market volitility.
In the meantime since Ninja code is beyond me and I don't use TS I wanted to set something to keep an eye on using Strategy Desk. Here is a chart with "Overbought" and "Oversold" bands above and below the price and MAs. The chart ends 5/27 in order to show a trade that should be eliminated (blue arrow). The crossing signals are indicated in the bottom pane (hard to see the little backtesting symbols and/or the MA cross). This trade was selected because it was the worst in the time period and met the conditions to be eliminated assuming I'm understanding what rank is doing. I'm not worrying about the profit target or the stop even exit for now.
Here are the formulas for the bands if anyone wants to set them up (parameters are replaced with xx):
((ExpMovingAverage[EMA,Close,xx,0,15]+MovingAverage[WMA,Close,xx,0,15]) / 2)*1.032
((ExpMovingAverage[EMA,Close,xx,0,15]+MovingAverage[WMA,Close,xx,0,15]) / 2)*.968
Now have to see if I can automate the filter -- not sure info was right about not being able to do this in SD.
StrategyDesk Code
Posted by davetrader on 3rd of Nov 2009 at 08:34 am
I just got off the phone with StrategyDesk and there seems to be no way to automate these settings. He said the program would need to be able to do a if/then caculation and currently SD is not able to do that. He said that as the ThinkorSwim trading platform developes that this may be posible, but that will not be for several months.
I would appreciate comments if anyone is able to set this up in another trading platform.
its pretty easy in tradestation
Posted by rank10 on 3rd of Nov 2009 at 09:02 am
its pretty easy in tradestation as long as you know how to use Easy Language.
all the settings above I have already automated into my strategy. If you switch to trade station and have any questions on how to set this up I would be glad to help
Great ideas, worth monitory. The
Posted by algyros on 3rd of Nov 2009 at 07:55 am
Great ideas, worth monitory.
The mechanical systems work best when automated. Do you have any idea how to write TS code to automate your modifications?
Nice improvements !!!
Posted by mansilvcas on 3rd of Nov 2009 at 07:11 am
Thanks a lot. Could you share the partial results of each improvement, so we can see how each affect the profit factor and the percent profitable? Many thanks
rank10 -- very nice equity
Posted by Michael on 3rd of Nov 2009 at 06:52 am
rank10 -- very nice equity curve. Thanks for the work and the ideas. Of course, the lower percentage profitable doesn't mean anything -- its just the breakeven trades being added in. winners to losers are almost 50%. Will be interesting to do more bakctesting and optimizing!
SRS#2 StrategyDesk Formula
Posted by davetrader on 3rd of Nov 2009 at 06:28 am
Nice improvements, thanks for sharing. Anyone have any ideas how to write this formula for SrategyDesk for these changes? I have a few ideas how to write it but need some help, I plan on trying to call Ameritrade this morning and see if they can offer any help.
Thanks, really nice. I tried
Posted by peterm on 3rd of Nov 2009 at 12:39 am
Thanks, really nice. I tried improvements with breakeven stops and no entering with high stochRSI but did not succeed to get long term results much better. The MA% difference is a nice angle.
IMO with your improvements the % winning vs % losers is the same (about 50%) if you take out the even trades. Don't see a negative there.
The profit factor is higher which makes it more attractive.
Until I have TS running as second system and third broker (risk reduction) I will try to code it in my NT stategy.
I like your ideas. Psycologically,
Posted by jacob on 3rd of Nov 2009 at 12:20 am
I like your ideas. Psycologically, it will be hard to run a 30% profitable system though. I have a hard enough time with a 50% profitable system...
Jacob
Of course he cut the
Posted by user32 on 3rd of Nov 2009 at 12:37 am
Of course he cut the losses in half, don't forget that. So, a few more losses here and there, but they're smaller than what we're used to. SRS is still a good system as it stands though.
Great suggestions! I think your
Posted by user32 on 2nd of Nov 2009 at 11:28 pm
Great suggestions! I think your idea is quite creative, and the improved results certainly warrant attention.
This is great IMO. One
Posted by mitchellvisa1 on 3rd of Nov 2009 at 12:09 am
This is great IMO. One of the things I was discussing with a fellow full time trader of these systems today is when the MA's have large whitespace between them on the actual chart we get losing trades. This white space I describe is exactly the same as what you are talking about...significant spread betweem the actual price and the MA. I believe a mean reversion program could do well on SRS in this current environment. Thanks for the insight, will try to post other ideas tomorrow.