When to pull money on the line??

[TerryS]

When to put money on the line??

I have been forward testing a simple short term trading strategy alongside my other trading. The sample tested so far is 69 trades since Jan 05. I have been surprized at the results.
58% profitable trades
Average profitable trade 1.73 R (R =$ risked)
Average losing trade 1.0 R
Doesn't sound that amazing but because of the reasonably high frequency of trades, low cost commissions and the position-sizing formula ,the return on capital has been 57.71% (YTD)with a reasonable equity curve (at least liveable). It looks to be heading for a 100+% annual return. I don't know what you other guys are getting, but it beats what I have been doing on the eminis in that period with a lot less screen time.
My question is, how big a sample do I need to go "live" with the system and put my money where my mouth is?
What size sample do you guys use to validate a system? I was thinking 100 trades but am I being too conservative?
My past experience makes me sceptical about such returns, however so far, the system seems robust with positive returns through an up and down phase in the market. since January. (the system only goes long).

Any comments from experienced system developers

Terturk

 

[Tuffty]

How about backtesting it to see how it would have faired in previous periods/years?

 

[TerryS]

Tuffy
Thanks for the suggestion but I don' t have backtesting capability and there is a discretionary aspect to the entry , it might be hard to backtest.

Terturk

 

[phoenixforex]

so what are u trading, eminis? you should find a broker who offers the possibility to trade mini accounts and start with that, if it works out just increse your account size by letting the profits compound.

 

[TerryS]

I'm swing trading NASDAQ and NYSE stocks.

 

[options]

I'm swing trading NASDAQ and NYSE stocks.

Don't know how long/what experiance you have. But I would say unless you really know what you are doing, leave the nyse alone. You really need to be on the ball with this one.

 

[BroadSword]

I wouldn't bother back testing it, you've effectively forward tested it and it works!

IAssuming the drawdowns are not too bad (you don't comment on this I believe) then I would just trade it, at first with small size.

Just don't do anything differently when there is real money involved.

Relax - you know it works, you've got 69 trades over 6 months telling you so!

 

[TWI]

If you have been honest about the testing i.e. actually taken signals as they happened real-time then it is worth tradig it in minimum size. If you have a position sizing algorithm as you state then this should increase risk as your account compounds. I do not know why you need to ask the question. Get in there.

 

[commanderco]

Hi TERRY S

At the risk of raining on your parade, I dont get out of bed for 1.73 & 58%.

However the good news is that you are probably on the right track.
My suggestion is to break your system into 3 parts
1 entry
2 managing the trade
3 exit

Apply a cast iron set of rules to these 3 parts and you will see a profound
difference in your trading results.

Oops, I almost forgot.
This will require the use of your discrection and so now
you are banging up against the wall that sorts the good traders from the rest.
Good luck .... you can do it.

 

[TerryS]

Thanks for the advice guys
options: any reason why NYSE is more dangerous than NASDAQ? Have I been lucky so far?
Broadsword: Largest drawndown so far is 10.3% with max conseq losers = 5. You are right I am aware that the biggest problem will be doing it consistently with money on the line.
twalker: Yes I have been very honest and disciplined with the signals and the figures. I have been down that track of cheating myself mentally by fudging the figures. It doesn't work in real world.
Commanderco: Thanks for the advice. I have strict rules for everything. The entry is based on a couple of patterns (retracements in a strong trend, nothing original) I don't see entry as important as the position-sizing formula and the risk management parameters.

The general opinion is that the sample test size is big enough to give a reasonably reliable result. (I know nothing is guaranteed in the market). Is that right?

 

[BroadSword]

...The general opinion is that the sample test size is big enough to give a reasonably reliable result. (I know nothing is guaranteed in the market). Is that right?

In my opinion, yes.

If I were you, I'd start with small stakes and see where I was in say 20 trades time...