I trade 1 ES
- Thread starter random trader
- Start date
Well lets do the math.. 90% of daytraders fail. Well nobody knows that number for real, but the majority of traders do fail. This guy is purely guessing on the direction of the ES over a weeks time period. There would be a 50/50 chance of success but lets not forget he does no have enough capital to withstand a bad week. I would say he probably has a 100% chance of failure.
Point being, I will take the bet the average daytrader will outperform certain failure.
You made two points. I agree on one and disagree on the other one.
This account is indeed underfunded for such an experiment. One bad week I will be out--overleverage kills There are several actions I can take and I will take one of them if the experiment gets to a bad start.
1. I can switch to NQ, or QM, or YG and continue experimenting with the same concept.
2. I can add a few K to my IRA account.
3. I can continue the experiment in my normal account, which will not be underfunded for this endeavor.
Your point about 50/50 success rate on a daily or weekly basis is a typical fallacy and I will tell you why. For the discussion that follows, let's ignore commissions for the time being.
In general, if you have a stop of x points and a target of y points, your system has an edge if the success rate is higher than x/(x+y). If you assume my success rate will be 50% because you think in my case x equals y, please go back and read my starting post. If you assume the success rate is always 50% no matter what, you need to take a Statistics course or read a Statistics textbook. I hope you are not one of those who believe the probability of raining tomorrow is a constant 50% regardless if they live in the Amazon jungle or in the Sahara desert.
Let me use two examples to illustrate this point further.
On this board someone used a stop of 1 point (cut loss early) and a target of 10 points (let the winners run) and he's looking for a success rate of 30-50%. Most people know this is unlikely to happen. It turned out his winning rate is a lot closer to 0% than 30% and not surprisingly he quitted after a few futile months.
On the other hand, if someone uses a stop of 10 points and a target of 1 points, he'll be taking profits most of the time. Whether he can win in the long run depends on if his success rate is greater than 91%.
In my experiment, I am shooting for a greater than 80% success rate with a stop of 20 points and a target of 5 points on average. You can see I could have chosen a stop of 1 point with a target of 0.25 point, or a stop of 100 points with a target of 25 points, to achieve the same result. This is why I didn't choose either.
1. If I use 1 point stop and 0.25 point target. I will be churning all day and the leak through spread and commision will kill my account sooner rather than later.
2. If I use 100 point stop and 25 point target, I will trade only a few times a year. This is boring even for me to follow my own experiment.
As a result, 20 point stop and 5 point target was chosen to strike a balance between survival and excitement. These numbers are not concrete as I will try to use them as guidelines and follow them only on average.
I'm reminded of an Abott & Costello show where they are in a situation, and need a horse to win a race. Bud says "We only have a one in seven chance." And Lou says "Naw, we gotta 50/50 chance." "How do you figure that?" asks Bud. "The horse can only win or lose." is Lou's reply.
Anyway, just so you know, a random entry between 9:30 and 4:00 (exits when the targets are hit - can be off hours) with a 5 point profit target and a 20 point stop loss with no comissions or slippage produces the following stats over the last 2.5 years on TradeStation:
Total trades: 1200
Winners: 958
Losers: 242
Even Stevens: 0
That's just shy of your 80% (79.83%). It would appear that you don't need much of an edge to push this into profitability. At 81%, you'd probably make about as much as your broker (if you have a very cheap one).
My "random entry" first checks to see if it's normal hours, then if it went long last time it shorts, else it buys. Close the position at +5 points or -20 whichever occurs first. The very first trade is a long.
Anyway, just so you know, a random entry between 9:30 and 4:00 (exits when the targets are hit - can be off hours) with a 5 point profit target and a 20 point stop loss with no comissions or slippage produces the following stats over the last 2.5 years on TradeStation:
Total trades: 1200
Winners: 958
Losers: 242
Even Stevens: 0
That's just shy of your 80% (79.83%). It would appear that you don't need much of an edge to push this into profitability. At 81%, you'd probably make about as much as your broker (if you have a very cheap one).
My "random entry" first checks to see if it's normal hours, then if it went long last time it shorts, else it buys. Close the position at +5 points or -20 whichever occurs first. The very first trade is a long.
Quote from TGregg:
I'm reminded of an Abott & Costello show where they are in a situation, and need a horse to win a race. Bud says "We only have a one in seven chance." And Lou says "Naw, we gotta 50/50 chance." "How do you figure that?" asks Bud. "The horse can only win or lose." is Lou's reply.
Anyway, just so you know, a random entry between 9:30 and 4:00 (exits when the targets are hit - can be off hours) with a 5 point profit target and a 20 point stop loss with no comissions or slippage produces the following stats over the last 2.5 years on TradeStation:
Total trades: 1200
Winners: 958
Losers: 242
Even Stevens: 0
That's just shy of your 80% (79.83%). It would appear that you don't need much of an edge to push this into profitability. At 81%, you'd probably make about as much as your broker (if you have a very cheap one).
My "random entry" first checks to see if it's normal hours, then if it went long last time it shorts, else it buys. Close the position at +5 points or -20 whichever occurs first. The very first trade is a long.
Thanks for the highly valuable statistics. I intend to push to profitability by doing the following:
1. Have a bias that matches with the market's intermediate term trend.
2. Improve entry points such that they are better than random.
3. Optimize target and stop points on the go rather than use mechanical points.
My broker is IB so I have already minimized the leakage.
Quote from TGregg:
. . . is the ability to curve-fit, AKA optimization.
A profit target of 9 points and a stop of 18 gives you 70% winners for a net/trade (before c&s) of $46.59.
The performance chart looks like crap though - got nasty, wide deviations.
No, no, I don't intend to optimize by curve fitting or back testing. I will decide stop and target points based solely on my mood at the time. Maybe optimize is the wrong word; randomize may be more appropriate.
Quote from TGregg:
I'm reminded of an Abott & Costello show where they are in a situation, and need a horse to win a race. Bud says "We only have a one in seven chance." And Lou says "Naw, we gotta 50/50 chance." "How do you figure that?" asks Bud. "The horse can only win or lose." is Lou's reply.
HAHAHAHAHAHAHAHHA! HAHAHAHHAAA! LAUGHING OVER HERE!
Decided to leave a GTC buy @ 982 order before going to bed.
Last Thursday evening after the blackout, I watched ES spiked down and wanted to buy @978 but of course IB was down. I will leave a GTC order whenever I am not watching the market. I guess I will be happy with 8 points (random) if something similar happens again.
Last Thursday evening after the blackout, I watched ES spiked down and wanted to buy @978 but of course IB was down. I will leave a GTC order whenever I am not watching the market. I guess I will be happy with 8 points (random) if something similar happens again.
3 point range overnight (989.75-993). Japan closed above 10000 the first time in a while. Order still standing. No fear, no greed.