51%

[shadowninja]

Sports trading? CBA with a journal; only doing the PH challenge thing cos it sounded like fun. Doing pretty well considering "it's gambling, innit". I made some dumb errors a fortnight ago by trying to trade a couple of high risk (but high profit) markets and got stung. Now sticking to safer (for my technique) markets. Can make 100% yield per week if I remain cautious. Have taken nearly 500 trades since and all is going well so far.

Will let you know in a month whether I'm still in it or not. It's been an interesting learning experience, anyway, and as suggested earlier, has made me think about my financial trading so it hasn't been a waste of time even if I decide to stop.

I recommend everyone have a go at sports trading. It can be done without risk of real money, too, via the £25 offer. (If anyone's reading, PM me about the £25 offer.)

 

[Vorbis]

I think it's easy to become over-obssessed with strike rate when it's only one part of the equation and probably the least important part.

Here are some figures I produced recently over a 20 Wk period manually testing various parts of an Hourly system. I used a fixed entry, SL, etc. in order to test various exit methods. The final results are almost certainly over-optimistic (it's easy in retrospect) but the figures should still be useful for relative comparitive purposes. The figures show HR then total pips collected by end of testing period.

Exit A - 60% - 976
Exit B - 48% - 2146
Exit C - 39% - 2776
Exit D - 56% - 1997
Exit E - 51% - 2265
Exit F - 47% - 2156
Exit G - 48% - 3065
Exit H - 42% - 3073
Exit I - 43% - 3358
Exit J - 38% - 3624
Exit K - 33% - 3020
Exit L - 21% - 2403
Exit M - 14% - 379

What surprised me is that HR and over-all profitability were almost inversely related. To get over 3000 pips required a HR < 50% and the highest HR (Exit A) made least pips apart from M, explained below.

(The big drop off at the end is almost certainly due to the testing period being too short as these were very loose trailing stop methods that excel over long time periods with very extended draw-downs.)

 

[trader_dante]

These sorts of high percentages may be possible during hot streaks or certain market conditions, but not in the long run.

As a matter of interest, what do you call the long run?

I've been trading a strategy with an 89% strike rate (0.5 R) since November 2009.

Not long enough?

 

[arabianights]

As a matter of interest, what do you call the long run?

I've been trading a strategy with an 89% strike rate (0.5 R) since November 2009.

Not long enough?

stupid question for obvious reasons

 

[trader_dante]

stupid question for obvious reasons

Hi Arabian.

 

[shadowninja]

89% strike rate (0.5 R)

This is pretty much what my conclusion has tended towards - high strike rate, lower R/R... account grows. Aim for 4 R and you'd end up taking more losses than you want.

 

[arabianights]

Hi Arabian.

My point is... you haven't said how many trades you've taken! If you've taken 1000 trades in that time with that record, it's stunning. If you've taken 1, not so much...

It is the number of trades that matters. The time period absolutely doesn't (except in so far as you can say it's seen various different market conditions, although arguably the markets since nov 09 have been pretty similar)

 

[arabianights]

(And it goes without saying, if your system includes any discretionary or non quantifiable aspect it is not a system for the purposes of this thread, but non quantitative types have huge difficulty with this point)

 

[shadowninja]

Yep.

 

[Hotch]

If you've taken 1, not so much...

I'd be quite impressed if he managed an 89% win rate in 1 trade.

 

[arabianights]

I'd be quite impressed if he managed an 89% win rate in 1 trade.

It's Mr P's maths

(Luv ya really T_D! How's about giving me a tip for that gambling account, in enough time so I can trade it.... you pwomissed..... )

 

[shadowninja]

You want a gamble? Go long euro now, hold overnight, stop -100, target +100. Take partial at 1.4190 and stop to b/e.

Note: anyone who takes this trade without doing their own research is a pillock and deserves to lose 100 pips.

 

[trader_dante]

My point is... you haven't said how many trades you've taken! If you've taken 1000 trades in that time with that record, it's stunning. If you've taken 1, not so much...

The actual number of trades in question is 45. (40 wins / 5 losses) and although I started out playing it with discretion, I now play it with none. It is totally systematic

As a matter of interest: what is a statistically valid number of trades? I would have thought that over that time frame (November 2009 - March 2011) the number would be less than for, say, a scalper.

 

[trader_dante]

You want a gamble? Go long euro now, hold overnight, stop -100, target +100. Take partial at 1.4190 and stop to b/e.

Note: anyone who takes this trade without doing their own research is a pillock and deserves to lose 100 pips.

Just to mix it up a little bit, Shadow, would you say it is also a gamble to go short now with the same stop and target?

 

[shadowninja]

The time-frame is irrelevant. The number is all that matters. I would say 1000 trades is a good sample size. If I said 40 out of 45 people I asked thought Justin Beiber was the most talented singer ever, you'd tell me to ask another 955 people. And stop hanging outside girls' schools.

 

[shadowninja]

Just to mix it up a little bit, Shadow, would you say it is also a gamble to go short now with the same stop and target?

To answer that would hamper my recommendation!

But I see no reason technically to be short for that size of stop. You might, of course.

 

[bob1974]

I'm a newbie here on this forum, but I appreciate everyones thoughts and oppinions. Thank you in advance for any help, or suggerstions in the future.
Cheers

 

[the hare]

The actual number of trades in question is 45.

Thats less than the number of trades I take in a week and I constantly worry about my stats being based on an insufficient sample size !

How will you know when your methodology no longer works ? do you have a definate line in the sand that you wont cross ?

 

[robster970]

If TD gives us the mean and Standard Deviation of his 45 trades, I think I can work out the likelihood of him achieving this by accident.

But don't quote me on that.

 

[arabianights]

If TD gives us the mean and Standard Deviation of his 45 trades, I think I can work out the likelihood of him achieving this by accident.

But don't quote me on that.

.