Will this ER2 system work?

Will this backtested ER2 system work going forward?

  • Yes

    Votes: 10 52.6%

  • No

    Votes: 9 47.4%
  • Total voters
    19
Just another cent...

Did you tried some Day of the week filters?
Most of the indices have the lowest range on mondays, the Dow remarkable...not too good for Trend/Breakouttrading.
Fridays also behave a little bit different, may be some late day trailing stop just for fridays makes some sense....mainly on the long side.

I would try to find some stable filters to increase the avgtrade results, personally i prefer less often trades and higher avgtrade/profitfactor and a more smoother equitycurve and general: a change of parameters should not change the natur/shape of the equitycurve completely (danger!)

AND:

You did not realized what a system with 50,-$/day REALLY means,
you should think about Moneymanagement...Such a system makes your equity not twice/year, it (theoretically...) makes you a millionaire in ONE year!

I did for some other forum an example, look here:

http://trader.gmxhome.de/the_es_one_point_trade_thing.htm

(the showed example system does NOT trade every day!)

Michael
 
more MM

If you use Volatility based Stops(What makes the most sense) or something similar, you can adjust your positionsize to the stopsize, what of course works out better(or at all) with bigger accounts. Makes always sense to adjust positionsize to the actual marketrisk.

with Stop: 500,-$ 1 Lot
with Stop: 250,-$ ==> double up 2 Lots

can make a lot of difference in your equitycurve...
 
Sorry for the many posts, my thougts coming step by step...

Final conclusion:

Your edge ( I guess...) are the long intraday Trends after the Openrange we see nowadays in the ER, I would not rely on this to be a stable condition over years, markets are always changing. So, if you think you found a system working the next 12 months, trade the hell out of it with a 10-15% Fixed Fraction MM till you reach 80000 - 100000,- $, cash in, and start again with a small account. Your system will fail for sure anytime in the future but you have the chance of making a fortune anyway...

If you rely on this ER behavior you can also try an openrange breakout thing, there are no indicators to use...

good trading and good luck
Michael
 
steve46 - I appreciate your thoughts, certainly time will tell. I do have concerns that perhaps when I sorted by V1, V2, V3, V4 in my optimization that I "saw" more than was really there. I'm a perfectionist and am susceptible to analysis paralysis!

--------------

mechtrader - thanks for all your wisdom - you gave me a few more ideas to try out to see if I can tilt that equity curve a little more up. I will backtest on some other indexes like YM and ES. And see what happens with a Day of Week filter. Also, I'll experiment with # of contracts vs. stop size - good idea!

Good guesses for most part on trying to figure out my system. As you can appreciate I don't want to reveal much - since my system only trades once a day and liquidity on ER2 is not nearly as deep as ES, if more traders trade my system than just me, I will have to worry about slippage on my entries. I will say that I use a 5 min. chart and use no "traditional" indicators - other than price action of course - that is the BEST indicator really...
 
Quote from dalodoma:

Will you consider testing with older data to see how negatively (or positively) the optimization effect the overall results?
More...

I've exhausted all my data actually (back to 12/11/03) If you have older ER2 data to send me, I'll take a look. In the meantime, I think I'll start backtesting with YM and/or ES data to see if my system has a shot on another index...
 
Very last Posting :-)

If you are really trendtrading try this...

It seems I found some little edge for intraday Trendtrading. An Openrange Breakoutsystem (which profits from Trenddays) of mine, performed better(avgtrade/profitfactor) if:

(daily indicators are more reliable):


Daily VIX is under it`s 10 day average (or some RSI under 50 etc.)

Daily market is not Overbought/sold, measured with a CCI(~20) not over 150, under -150, or try any other oscillator, i prefer CCI.

The logic seems to be the nervoesness/anxiety/fear thing, the traders are more irrational, acting faster, this leads to more reversals and erratic moves, and this kicks you out if intraday-trendtrading.

promising would also daily put/call ratio and any other things pointing toward stressed marketpartitiants...

I also look on daily Volatility and don`t trade Intraday-Trends after a Widespread Day before, if the day before had a large range or was already a nice trendday, chances are bad for trendtrading. Volatility is much more predictible than price.

have fun with testing...

(i`m not always sure with my english, not my homelanguage...)
 
Hi DerferMark,
I started this system to trade, yes. It´s a kind of Openrange Breakout on the YM, no intraday indicators, but i look on a lot of priceaction, gaps, size of openrange, shape of openrange, i don´t take trades in the lunchhour (lot of fakes) and so on, a lot of statistical edges. I take trades only not too late in the day and start to trail only the last hour of trading.

I get relativly good positive results with ES, but ES is the most noisiest thing, very hard to trade at all and very hard to trade mechanical. I just want to see that things work also on the ES.

For any reason i like the YM, YM is also more trendy as the noisy ES. Why the ER2 is trendy, i have no glue. I the case of YM lies some little logic, cause the Dow has only 30 stocks, if just a few "are going" it nicely moves the whole thing, so for me i see a reason why the YM could be permanently more trendy as the ES, but logic is not always (or rarely?) working ;-)
 
@DerferMark

I Would be curios what`s the software you use. You trade with the Trend or ?

If it`s possible for you to use daily Indicators i would like to hear if you get similar results with this VIX/CCI stuff and i got another idea here...

Volatility Standartdeviation as filter

in Tradestation it looks like


Condition1 =(VolatilityStdDev(2)of DATA2) > xaverage((VolatilityStdDev(2)of DATA2),Length);

DATA2 = daily

I get for every number of Length over 3 a far better avgtrade/profitfactor as with the original strategy, but it reduces # of trades/netprofit significantly so i just use it to take a higher # of contracts to trade if this condition = true

if Condition1 then NCon = NCon * 1.5;

that`s another way i use positionsizing, better Risk/reward situation, higher positionsize

In case you use TS i can give you some code for filtering/testing, i`m just curios if this works for other strategies/markets as mine too.
 
Mechtrader - thanks for the additional ideas. And your English is just fine - no worries :D Where do you call home? I'm in Chicago (the futures capital!)

If I'm able to try some ideas of yours I'll be sure to let you know the result. So far I have added the Day of Week to spreadsheet and Fridays are the best result for my system and all other days are about equal.

I do not use Tradestation. I only have eSignal for quotes and charts. They have only 6 months of Intraday data however. I do all my analysis either manually looking at charts or via Excel programming - nothing fancy but it works for me. I do not have a programming background but am a former engineer.

Not sure if I even get the VIX - I'll have to see.

My system trades both trend and reversal (depending on value of my signals).

Thanks again - have a great week!
 
You must log in or register to reply here.
Back
Top