BE DO HAVE and Jack Hershey vs Objective Reality
- Thread starter ^^^^^^
- Start date
there is some validity to what jack espouses but its extremely difficult for the common man to do this, the psychology and his starting capital aren't enough.
I'm up a considerable amount, intraday if I didnt swing with the price action I would be down substantially.
the equity curve from attempting to do what is implied if I'm reading it right looks like this.
I'm up a considerable amount, intraday if I didnt swing with the price action I would be down substantially.
the equity curve from attempting to do what is implied if I'm reading it right looks like this.
I mentioned back testing briefly and that the Blocks Player was coming online and it looks pretty good as a possibility.
Some of the US people have done some back testing and forward testing. Some have also offerred to help people who want to do their own.
As I have also mentioned before there is usually an order of magnitude difference in results when different people post testing in the same thread.
One feature of blocks player is that it produces the conventional results of the testing. Several people on ET post back testing results and the results are cited in the conventional way.
The equity curve of the results being cited now looks like it has a fairly consistent loss showing per trade. Just under 20 dollars. There were about 32 trades a day done on a particular universe that was cited. It was a universe of 1,000 stocks taken from the S&P 500 between 2000 and 2005.
All trades took the same number of days. Four days.
It was suggested that the equity curve could be turned upside down and the process would make money. That was checked out and reported. It still did not make money as reported out by the tester.
A 0 to 7 test involves three measures as everyone knows: Price going from declining to advancing during the morning. Volume going from declining to advancing in the morning. And a change of sentiment going from distribution to accumulation in the morning.
Long ago EOD data was the only thing available and the measurements were taken in the evening and the resulting equity curve was about 20 less in money velocity.
The stocks being tested in the universe have been determined in different ways at different times.
Today they pop up literary by the press of a button and some bulking using IAS sheets.
Or a larger universe may be obtained by using a set of six criteria.
What is neat about all of this is that it filters two things. Stocks and people. We all can see that Trader666's backtest is perfect and his posting it is very helpful too. It gets double duty.
US people made a nominal pass with him to "fix" his back test. He was not receiving any input at the time.
400,000 Monte Carlo runs show the exit is 8.1 days after the entry on valid universes.
On the first mechanical system used by a coder in C language before Y2K when version 3.0 of TC2000 was used to pick universes using the 7 Boolean equations at that time , the results were a hold of 6.6 days for a profit of 11.1 percent. After 4.5 hours of 1 one 1 review during a day of real trading, the results after tweaking improved approximately 20%. One side effect must be noted. The perosn switched from working to not working and trading full time. Elsewhere, a similar shift has been noted when people switch from EOD to real time charts.
All THEM people can review all the content of Trader666's backtesting by using his links. Some additional content would be helpful: posting the 1000 IAS sheets he used to qualify the 1000 stocks. Posting how he coded the add/delete on the universe using whatever period or frequency of updating he used. Posting the coding he used to get the price change straight. Posting the coding he used to get the volume change straight. Posting the coding he used to get the A/D sentiment shift straight. Posting the coding he used to merge the streams of data for P, V and A/D.
With regard to the hold period, there are many factors among them stops, period of hold criteria. The way he handled the scoring during hold.
For the exit, he has cited his approach. So that is clear to us. He used an early exit strategy.
If anyone knows how much money was used to get the average loss of 20 bucks, let everyone know. 32 trades a day average must have involved some initial capital. We know it was over 450,000 dollars because that is what he shows the backtest lost.
We also see the drawdown and how the drawdown works. It is reported out directly by the equity chart.
THEM people are not going to backtest the PVT equities trading this time around. The trader666 filter is doing the job it is needed to do.
Newbies are being protected by virgin and she/he may be preforming for them on behalf of woodie too.
THEM's have not posted on the crayola backtest for SCT as yet. This is not performance but it is a test of the potential that the market's offer. The effectivenss and efficiency and optimization flow from the potential that is there.
It has been noted that at 7X the range, the profits go to zero because of commission costs.
For yesterday, we can note on SCT that a 200% result obtained is not good enough to get a 3x's H-L result. It was a tough day for getting 3x high-low. 200% of margin happened however.
Some of the US people have done some back testing and forward testing. Some have also offerred to help people who want to do their own.
As I have also mentioned before there is usually an order of magnitude difference in results when different people post testing in the same thread.
One feature of blocks player is that it produces the conventional results of the testing. Several people on ET post back testing results and the results are cited in the conventional way.
The equity curve of the results being cited now looks like it has a fairly consistent loss showing per trade. Just under 20 dollars. There were about 32 trades a day done on a particular universe that was cited. It was a universe of 1,000 stocks taken from the S&P 500 between 2000 and 2005.
All trades took the same number of days. Four days.
It was suggested that the equity curve could be turned upside down and the process would make money. That was checked out and reported. It still did not make money as reported out by the tester.
A 0 to 7 test involves three measures as everyone knows: Price going from declining to advancing during the morning. Volume going from declining to advancing in the morning. And a change of sentiment going from distribution to accumulation in the morning.
Long ago EOD data was the only thing available and the measurements were taken in the evening and the resulting equity curve was about 20 less in money velocity.
The stocks being tested in the universe have been determined in different ways at different times.
Today they pop up literary by the press of a button and some bulking using IAS sheets.
Or a larger universe may be obtained by using a set of six criteria.
What is neat about all of this is that it filters two things. Stocks and people. We all can see that Trader666's backtest is perfect and his posting it is very helpful too. It gets double duty.
US people made a nominal pass with him to "fix" his back test. He was not receiving any input at the time.
400,000 Monte Carlo runs show the exit is 8.1 days after the entry on valid universes.
On the first mechanical system used by a coder in C language before Y2K when version 3.0 of TC2000 was used to pick universes using the 7 Boolean equations at that time , the results were a hold of 6.6 days for a profit of 11.1 percent. After 4.5 hours of 1 one 1 review during a day of real trading, the results after tweaking improved approximately 20%. One side effect must be noted. The perosn switched from working to not working and trading full time. Elsewhere, a similar shift has been noted when people switch from EOD to real time charts.
All THEM people can review all the content of Trader666's backtesting by using his links. Some additional content would be helpful: posting the 1000 IAS sheets he used to qualify the 1000 stocks. Posting how he coded the add/delete on the universe using whatever period or frequency of updating he used. Posting the coding he used to get the price change straight. Posting the coding he used to get the volume change straight. Posting the coding he used to get the A/D sentiment shift straight. Posting the coding he used to merge the streams of data for P, V and A/D.
With regard to the hold period, there are many factors among them stops, period of hold criteria. The way he handled the scoring during hold.
For the exit, he has cited his approach. So that is clear to us. He used an early exit strategy.
If anyone knows how much money was used to get the average loss of 20 bucks, let everyone know. 32 trades a day average must have involved some initial capital. We know it was over 450,000 dollars because that is what he shows the backtest lost.
We also see the drawdown and how the drawdown works. It is reported out directly by the equity chart.
THEM people are not going to backtest the PVT equities trading this time around. The trader666 filter is doing the job it is needed to do.
Newbies are being protected by virgin and she/he may be preforming for them on behalf of woodie too.
THEM's have not posted on the crayola backtest for SCT as yet. This is not performance but it is a test of the potential that the market's offer. The effectivenss and efficiency and optimization flow from the potential that is there.
It has been noted that at 7X the range, the profits go to zero because of commission costs.
For yesterday, we can note on SCT that a 200% result obtained is not good enough to get a 3x's H-L result. It was a tough day for getting 3x high-low. 200% of margin happened however.
You know what the gotcha is right? We were trying to fix your backtest because your universe was not VALID by your own definition of what it means to be valid.
"Things like qualified universe, 5x+ 20%+ cycling, etc.". Two things you admitted to have not incorporated. In other words, your data set was unqualified (ie. not valid). You verified that it does not work on an unvalidated universe. As a result and by your own definition, you have discredited your own backtest. If a car has no gas, how are you going to expect the car to get you to point B. This is the problem we had tried to fix in your backtest (ie. get valid favorable non-fantastical results). Qualified names have juice (ie. gas). Thats what gets your EQ curve from Negative to Positive...
This is a good point.
Anyone see virgin in the chat room checking out live trades??
PM her/him to invite her/him over.
PM her/him to invite her/him over.
He's no virgin, he's been getting fucked by the market for years now 
No, the gotcha's on you!
Because in Jack's "tomorrow's paper today" there's no mention of any of that garbage so I tested only what Jack wrote and it bombed out bigtime.
And with "rockets," the first thing I tested was again, exactly what Jack wrote. I even quoted it and Jack said those conditions were all that was needed. But when it bombed, you guys keep piling on condition after condition, after the fact... one of your many obfuscation techniques.
Because in Jack's "tomorrow's paper today" there's no mention of any of that garbage so I tested only what Jack wrote and it bombed out bigtime.
And with "rockets," the first thing I tested was again, exactly what Jack wrote. I even quoted it and Jack said those conditions were all that was needed. But when it bombed, you guys keep piling on condition after condition, after the fact... one of your many obfuscation techniques.