Strategy Performance Report - Should this system be traded?

[tradesmyth]

Problem with gaps is it is hard to quantify them, the SD of gaps (if you calculate and plot them over a period) is high and certainly skewed distribution thus randomness (read chaos) is high. It hurts specially when trading lower TF. I was once working on overnight system (buy around market close to sell next day morning), a gap strategy, and lost some hairs, he hehe

So, your suggestion would be to either trade purely intra-day, if trading must be done on a lower TF; or trade larger TF with overnight positions. Is that correct?

my 2 cents, don't look at stock level volatility, rather market level or if you can sector level, pays good dividend, !

hmm ... Can you give an example of a system that works on sector volatility? Forgive me if this is too many questions going on different tangents.

Our forte is our system, platform, code, these are our weapons. As they say in fitness training, more you sweat more you lose, here I say, more you sweat less you lose. (pun intended)
there is no hurry, take your time with mastering the code (unfortunately will not be able to help here since have no knowledge of Easy Language, if you switch to Amibroker or Strategy Trader, feel free to ask anything)

The trouble with Easy Language was no built-in ability to Walk Forward Test. So I've been struggling with AFL off-and-on for the past two months. Right now I'm at the stage where I'm awed by the thousand-line-code AFLs strewn around the internet; and have been resisting the temptation to blindly start testing all of them on various instruments. I understand there's no short cut to this, and I'm gradually putting in the hours to get better at AFL.

I think may be my comments aren't making you feel too happy, so for all your serotonin and feel good, a small gift.
http://www.youtube.com/watch?v=MwKYjZ_8EcE&feature=share

Hah! your comments make me feel very enthused actually. I feel like I have so very much to learn from you, and you do a fantastic job of explaining. Thank you.

 

[iamaaditya]

So, your suggestion would be to either trade purely intra-day, if trading must be done on a lower TF; or trade larger TF with overnight positions. Is that correct?

Yes, but I am only showing you the challenges involved in gap trading. Everything has a trade off. Pros n Cons. Gap strategies generally have very very high win% (sometimes upto 90%), (especially when coupled with data from other markets, S&P, DJIA, FTSE, Crude Oil and USD Index) (these factors mainly decide opening level) ya but low pay-off ratio.
You can still go ahead , do more research and better your system, but you will have to step up your sophistication and be state-of-the-art to profit from what I call less juicy fruits. (Sure you can leave and go for more juicy fruit, where everybody is there but since there are many already here battle is no less difficult). (I hope you are getting what I am trying to say, !! If not, just try a simple MA crossover system and run WFA by optimizing the parameter for both periods on NSE (CNX 100 only for liquidity reasons) for around last 10 years). (you will see what I am saying, the current scenario is such that it selects only antitrend (MA1 > MA2) system, because trend crossovers are all "maxed out")

hmm ... Can you give an example of a system that works on sector volatility? Forgive me if this is too many questions going on different tangents.

Never mind the questions, that is how we learn.
Yes, one system on which I am currently working involves sector analysis. This system takes a sector (comprising less than 10 stocks) and for 5 year period finds out the beta of each stock (vs the stock index, if not available, compose one, in AFL, use Add to Composite), and once we have the data, when there is an aberration on the stock volatility vs the index on last 3 months, it trades against the move. for eg. if stock A had beta of 1.5 but currently trading at beta 1, it goes long and vice versa. It is a good profitable system (on backtest and WFA), but haven't had time to move it to Live, since occupied with other strategies currently.

The trouble with Easy Language was no built-in ability to Walk Forward Test. So I've been struggling with AFL off-and-on for the past two months. Right now I'm at the stage where I'm awed by the thousand-line-code AFLs strewn around the internet; and have been resisting the temptation to blindly start testing all of them on various instruments. I understand there's no short cut to this, and I'm gradually putting in the hours to get better at AFL.

Information overload is a serious problem in this business. Don't search internet for more afls. It will drive you crazy (personal experience, I have 20K AFL), rather look at built-in examples, and take it from there. If stuck at any point, I will be glad to help.

I have one question, what is your computing power ? This line of business needs serious power ! I am currently looking for beyond Desktop PC but within my range (can't afford IBM servers yet). If you know of any arrangement in between would be glad to know.
Currently I have on my capacity 3 PC (actually 1 laptop), all having 2 cores and almost all the time they are 100% utilised but my needs has outgrown my hardware (pun certainly not intended, he he), don't want to add more PC. (buying VPS or Colocation is like another PC, looking for something in between). Can't use Cluster or DC for optimization process , this will add to overhead rather...

 

[tradesmyth]

I take your points regarding the tradeoffs between strategies with high probabilities and low payoffs, and strategies with higher payoffs too congested with a lot more people trying to profit through them. And also regarding not chasing more AFLs.

I have one question, what is your computing power ? This line of business needs serious power ! I am currently looking for beyond Desktop PC but within my range (can't afford IBM servers yet). If you know of any arrangement in between would be glad to know.
Currently I have on my capacity 3 PC (actually 1 laptop), all having 2 cores and almost all the time they are 100% utilised but my needs has outgrown my hardware (pun certainly not intended, he he), don't want to add more PC. (buying VPS or Colocation is like another PC, looking for something in between). Can't use Cluster or DC for optimization process , this will add to overhead rather...

My computing power, so far, is quite inferior, mostly because of a lack of awareness about how to leverage greater computing power. I have one Laptop which is a 2.53 GHz Core i5. I place and modify orders on an ODIN screen, and generate System Signals through Tradestation 2000i based on 30-min bar data that I receive from my data vendor. And this rudimentary a setup could have done with even lesser computing power.

I was under the impression that Colocation was beneficial in case of High Frequency Trading only. I'm unaware of Cluster or DC.

 

[iamaaditya]

Sooner or later you will need more and more resources, in terms of hardware (cores, memory, faster bus, space, backup) and software (to backtest, for monte carlo, excel won't be sufficient, (starts freezing after 1 lakh rows), may be Access, MySQL), also for analysis you will have to learn Matlab, C (C++), R, and If I may add with little humour, W, T & F......

All this may sound like making trading complex, yes it will be complex, and if you think that you would rather stay discretionary watching charts, only PA, think again, in that case all these complexity will build up inside the cerebrum, (if you disagree check out the progress of Flow Trades on Saint's forum, from trading Pivots, it has gone into detail where length of each candle wrt to previous candle are given different names with different interpretations...)

No trader should be against complexity, but only stay away from making things complicated. or should I quote Einstein "Make everything as simple as possible, but not simpler"

take your time, don't be in a hurry, the task at hand is not going to end soon (or should i say ever, personal experience). These backtest/optimisation/wfa/monte/... isn't you suiting up for a BASE jumping (where by the analogy being that once tied correctly and off the hook, you will enjoy the glide).. this isn't going to happen.

I rather envision it as climbing Everest, where reaching top is yes one of the aim, but mostly enjoying the thrill on the way, fighting challenges all the way and learning from every drop and every nail on the cliff.

so when you are at the base (if you are interested base of mt everest is called Solokhumbu, great place, been there), you would rather not hurry up to start climbing. Run around a little for few days (for stamina), do some weight training (for strength), learn about every gorge and turn (for Survivor-ship, analogy Risk of Ruin) and when ready (don't wait too long, else your age/muscles/strength may let you down, analogy confidence, complacency) start the climb, without the worry that you will ever reach the top, enjoying every small climb and learning from every small slip..

Succes rate of climbing Mt. Everest successfully is 29%. (and Mind you these are the fittest people who dare attempt the feat) I would not be surprised if those who fought the adversity wanted to fight the challenge rather than reach the top (analogy being, those who win (in Trading, say making money) are looking for something challenging and to win over the challenges (primary being self's mind, psychology), and not necessarily making money. (which comes as a buy product on which we aren't that interested) (if you love the challenge there are more things to do after reaching top i.e making money)

may be my metaphorical e.g has gone little too far (& wild) and I should stop before people start calling me crazy (wait, arent' they calling me it already)

In short, my advice, take it slow, learn everything you can, prepare yourself, sweat it out, there will be enough opportunities to make money and market isn't closing down in next few months (unless you are a 2012 aficionado)

may the learning be with you.
phsknm

 

[tradesmyth]

yeah. i like your 'Everest' analogy. And i've been of this opinion as well that the sophistication required to find an edge would, in all likelihood, have a high attendant complexity.

i'll be taking it slow.

thanks. and cheers.