Corporate Model Licenses Available Only

Note:  This offer is not about this particular model.  It is all about providing financial neural net modelers with a combination of inputs, target variable and data transforms that allow for their own model building.  Naturally, this model is part of the package if you wish to track its out-of-sample performance on your own.

Professional and/or Corporate Model License

- Negotiated -

 Minimum $20,000

Contact

In- and Out-of-sample Equity Surface 

Post-Optimization would not improve this model in that it is already at its optimal profitability.

 Out-of-sample Equity Surface

Post-Optimization would not improve this model's Out-of-sample performance in that it is already at its optimal profitability.

What's included as part of the Nasdaq-100 Neural Modeling Package? 

Click Here

Click the Chart below to View the "Competing Model's" In- and Out- of Sample Equity Curve

See this Model's Out-of-Sample Equity Performance Here

And View this Model's Hypothetical In- and Out- of Sample Trades Here

 In- and Out-of-Sample Equity Curve

 Out-of-Sample Equity Curve

In- and Out-of-Sample Hypothetical Trades

Click Here

Modeling Data

Pinnacle, Reverse Adjusted

Neural Model Software Used

Profit Pro 2001

% Data Split

Train/Test/Validate

50/25/25

Data Start Date: 9/16/1999

Holdback Date: 4/1/2001

Development Date: 6/19/2001

Total Inputs

8

Passed Future Leak Test?

Yes

Neural Model Type

Continuous Adaptive Time Delayed

Trading Frequency

Very Frequent

A large number of trades assist in validating that the results achieved are not due to chance alone.

Dollar Amount used for Commissions/ Slippage on All Trades

$100

Profitability of Each Data Segment?

Train - Yes

Test - Yes

Validate - Yes

Out-of-Sample - Yes

Optimized Inputs

None

Optimized Target?

No

Post-Optimized Neural Signal?

No

Number of Models in System Model

N/a

Max % of Perfect Individual Model in System Model

N/a

Min % of Perfect Individual Model in System Model

N/a

Special Inputs, Techniques, Software

Jurik's JMA indicator is required if you desire to track the exact model from the home page; however, experimentation with the data shows that an adaptive moving average (AMA) produces  comparable models to those which make use JMA for this Nasdaq-100 futures model.  The Excel spreadsheet data documentation includes all calculations necessary to produce an AMA indicator.  Naturally, the use of a robust neural net  financial software package with genetic modeling capabilities is required.

Complexity of Input Data Transformations

Simple

Complexity of Target Variable Transformation

Simple

Medium for Detailing Data Transformations

Excel Spreadsheet

Non-Disclosure Statement Required?

Yes

Product Availability

Now Available for Purchase by Corporations Only.

The beauty of the data employed in this model is that is requires no input optimization, target variable optimization or post-optimization of the neural signal.  In addition, it's very easy to develop great models (inclusive of true out-of-sample data).  Bottom-line, simple models such as this tend to be more robust than their optimized counterparts.  Lastly, you can attempt to build models with the data prior to purchase in order to gain confidence in the modeling qualities of the data.  In fact, it is highly encouraged.  If you fail to attain good to great models fairly easily with the data provided, do not purchase a license for this product.

Another important aspect to note is that you are not merely purchasing the perpetual license for a single model.  With the actual data and transformations supplied, you can develop as many models as you desire to fit your personal needs and re-train as often as desired.  The blackbox mystique of the data and transforms required to construct this model is completely unveiled.  There is, however, no guarantee future model performance will in any way mimic the performance(s) achieved by this model or others depicted on this website.  Retraining of models is also not guaranteed to produce models similar in performance to those depicted on this website due to the nonstationarity (ever changing data characteristics) of financial data.