personally I like your cash cow system better if you can get it to work forward testing.
BWolinsky Trading
- Thread starter bwolinsky
- Start date
- Status
We've severely detracted from my goals here.
Today PTQQS had no trades, and volatility continues to narrow in the NDX despite the spike in the vix.
Fidelity failed me today, but it does not appear that I had any problems missing a trade.
Their streaming quotes were not functioning all day after I returned from the red eye from Savannah to Kansas City. I had monitored the cash cow system while on vacation for the fourth.
I had to refresh the page every 15 minutes manually despite real time quotes being connected in WLP.
This is the output from PTQQS
7/6/2009
QLD Projection: 34.6506368196927 QLD Close: 36.1800003051758
QLD Projection: 38.0258178017356 QLD Close: 36.1800003051758
QLD Projection: 37.4998089744931 QLD Close: 36.1800003051758
RSQ1: 0.874360740184784
RSQ2: 0.905622601509094
<i>The Rsquare value measure the fit of the projection. A .87 reading implies it fits 87% of the tested data.</i>
THE TREND IS 0.0608190112077257 <i>A Snap Trend calculation showing upward movement or slightly mostly to the overbought side</i>
7/6/2009:QLD Predict: 34.4806367433988 QLD Close: 36.1800003051758
<i>A key point here is that it appears that QLD is predicted to go lower. In order for us to trade tomorrow, assuming we continue to sell off, we will need to close below the QLD Predict Value</i>
WASH! <i>Indicates we are in a Wash Sale from the previous trades</i>
QLD Needs to drop -3.06761682788375%
Down Threshold is $ 35.0701365274858 on QLD
<i>A Very Rough Estimate</i>
All of these values will change tomorrow. But the main one I pay attention to are the Linear Regression Projections. You can have the debate for a long time over regressing directly onto prices, but you miss the point that all it does is predict the value for the next day. Seeing the values from today gives a rough idea what will happen tomorrow. I say for trades tomorrow, because when the market closes, I'll refresh this data to make my decision on whether to enter the trade the day after tomorrow. We are slightly on the oversold side, and roughly 3.067% away from the oversold level. It took a lot of math to get to the last two lines.
The last two lines about where the exact oversold threshold is at is only an estimation. It is not exact, but gives you some idea of the magnitude of moves required to cross under an oversold threshold. Last week we were close to overbought thresholds, but have since come down on a normalized basis.
Today PTQQS had no trades, and volatility continues to narrow in the NDX despite the spike in the vix.
Fidelity failed me today, but it does not appear that I had any problems missing a trade.
Their streaming quotes were not functioning all day after I returned from the red eye from Savannah to Kansas City. I had monitored the cash cow system while on vacation for the fourth.
I had to refresh the page every 15 minutes manually despite real time quotes being connected in WLP.
This is the output from PTQQS
7/6/2009
QLD Projection: 34.6506368196927 QLD Close: 36.1800003051758
QLD Projection: 38.0258178017356 QLD Close: 36.1800003051758
QLD Projection: 37.4998089744931 QLD Close: 36.1800003051758
RSQ1: 0.874360740184784
RSQ2: 0.905622601509094
<i>The Rsquare value measure the fit of the projection. A .87 reading implies it fits 87% of the tested data.</i>
THE TREND IS 0.0608190112077257 <i>A Snap Trend calculation showing upward movement or slightly mostly to the overbought side</i>
7/6/2009:QLD Predict: 34.4806367433988 QLD Close: 36.1800003051758
<i>A key point here is that it appears that QLD is predicted to go lower. In order for us to trade tomorrow, assuming we continue to sell off, we will need to close below the QLD Predict Value</i>
WASH! <i>Indicates we are in a Wash Sale from the previous trades</i>
QLD Needs to drop -3.06761682788375%
Down Threshold is $ 35.0701365274858 on QLD
<i>A Very Rough Estimate</i>
All of these values will change tomorrow. But the main one I pay attention to are the Linear Regression Projections. You can have the debate for a long time over regressing directly onto prices, but you miss the point that all it does is predict the value for the next day. Seeing the values from today gives a rough idea what will happen tomorrow. I say for trades tomorrow, because when the market closes, I'll refresh this data to make my decision on whether to enter the trade the day after tomorrow. We are slightly on the oversold side, and roughly 3.067% away from the oversold level. It took a lot of math to get to the last two lines.
The last two lines about where the exact oversold threshold is at is only an estimation. It is not exact, but gives you some idea of the magnitude of moves required to cross under an oversold threshold. Last week we were close to overbought thresholds, but have since come down on a normalized basis.
Well, hopefully Fidelity has their shit together tomorrow. This is the only day I've ever been with them that they had problems with streaming data. It's probably happened once before, but I wasn't quite as dependent on real time data. WL5 seemed to function all right today, though.
I tried adding the reversal, and there's too much chop. On some days you're hovering around 50 for most of the day and keep crossing over and under and exiting your position.
Volume, Jack, is the key to the whole system. 20000 .SPX contracts is not 20000 ES contracts, as I've tried to explain to you. I have no trades on 5 minute bars that trade 20000 contracts in WL5.
I guess I need to go back to the TS code from Scottd. I had submitted a TS program for review from a third party that has been tracking it after optimizing the volume conditions some for 15 minute bars.
What you're saying is that Cash Cow was built for 5 minute bars on obviously ES data, and was shown to be unprofitable when working in TS.
20000 contracts on the ES is nothing.
You've screwed-up jack's over-optimized programming.
LOL, this place is a madhouse.
You've screwed-up jack's over-optimized programming.
LOL, this place is a madhouse.
Quote from MandelbrotSet:
20000 contracts on the ES is nothing.
You've screwed-up jack's over-optimized programming.
LOL, this place is a madhouse.
Yeah, well, SpyderTrader was a WL programmer, not a TS coder. This system only works on .SPX. The results don't indicate failure, however, and I'm still watching the weather change.
I don't really do anything. It's all automated once you write the program.
It really depends on the time frame, Mandy. The first forward trade was a winner. Little drawdown.
The trading stategy of the Cash Cow is to build a basic core intraday trading core that trades channels on the 5 minute ES using reversal trading. Raw price/volume data (P, V) was converted to about 31 information components. These were assembled on four levels of logic and iteratively refined (five versions).
The basic core corresponds to beginner and in May version 1.3.3 was doing 25K/month with 5 contracts, 7,500/Month in all of June with 5 contracts.
Now, with versions 1.3.4 and 1.3.5 we are backtesting May and June. The basic core is done (complete for channel trading) as we see it.
Three more shells will be added: basic supreme (channel trading using "order of events"), intermediate (traverse trading with leading indicators of ES) and expert (tape trading, with leading indicators of ES).
I checked Fidelity. They do not offer trading on ES.
I do not know what TS stands for regarding the Cash Cow. Usually it means Trade Station. Trade Station is unfamiliar to any of us.
Regarding the applicability of the cash cow to other futures indexes we are totally uninformed. Anyone trading any other market and wishing to experiment, let us know.
There is a lot of proof at this point that the cash cow is not being used on the SPX on the Fidelity platform for intraday trading.
You can't play around with volume relationships like that.
This is amatuer hour, right?
I'm going to respond to this because I know you're an older aged man.
TS is tradestation and refers to easy language code, a C++ derivative. WL is Wealth Lab Perl in V4, and .NET in V5, pro or developer depending on where you live. Fidelity offers a unique dataset in .SPX native only to Fidelity Wealth Lab Pro V.5. This is probably where you're getting confused. I trade SDS and SSO <i>BASED</i> on .SPX indicators. I learned this writing PTQQS, Pairs Trading QID QLD Scalper after years of working with trading systems. At the end of the day, your logic gates are just if then else statements, and you cling to this cryptic method of presenting your trades with weird charts with straight lines.
It makes zero sense to me for you to say TS is unfamiliar to you, because Scottd's version <i>WAS</i> written in easy language code.
Jack, if you're serious about 3rd party verification, rather than this cryptic language you can send me whatever program you have, and I'll verify it. I've done this for two people before, and they haven't turned me down, so don't be the first.
If I were to discover your claims are true, just imagine how good that would be for you. If it's true, your cult wins, and if not, nothing's changed. I'm just one more person willing to give you a shot.
I'm sure Fidelity had a great time talking to you, Jack.
Anyway. I'm a disinterested 3rd party here. As 3rd party as it gets, and took a Cash Cow ATS from Scottd and converted it from Tradestation easy language code into Wealth Lab code.
I've looked at iterative refinement and do not see anything but multi-alias posts of daily charts with non-sensical lines with no predictive value.
All we really need is the program. The Ergodic Oscillator from the other pariah has not been traded or verified succesfully either, and he gets the same invitation.
For you, though, I don't see any technical way you can trade off the shelf out of excel unless paper trading. You do not appear to speak about FIX protocol ever in your statements, therefore there's no way you're trading with real money out of VBA or Excel or any other non-functional platform.
So what do you say, Jack? PM me, and we'll get the verification started.
I suspect, though, that you'll go on with more of the same. My version of Cash Cow was found to be profitable in WL5. Take the WL script translator to the WL4 code, do a little optimization and you get the results in this thread.
I know SpyderTrader wrote your scripts, so you can continue to work with him, or work with me. It doesn't look like Spyder ever made real time calls, which does call into question why you're so concerned with Cash Cow being in this thread.
- Status