Dr. Ye, November 2013
1. What is a financial use case of Big Data modeling?
Robert Shiller and Eugene Fama won 2013 Economics Nobel by proving twice long-term stock prices are predictable (Irrational Exuberance, Shiller 2000). Can they provide ticker-based repeatable predictions on a regular basis? Big Data modeling advocates a paradigm shift for research to automate traditional research process for repeatable and sustainable predictions. Big Data is defined for finance as population (as opposed to sampling) data (see Greenspan on data sampling to 2008 financial crisis): e.g., all historical price data points for a collection of tickers on a given time interval (days, minutes, seconds etc). Our use case: Sentiment Asset Pricing Engine 4.0 (SAPE4) - new methods and system on auto-quant-modeling and auto-predictions. SAPE4 is part of our Pangu servers.
2. What SAPE4 (SPS) can do for investors and institutions?
Ticker-based repeatable predictions on a daily basis: auto-modeling, auto-predicting, auto-charting, and auto-emailing, auto-back-testing, auto-GARCH-fit-testing, auto-Web-presenting. SPS does not do auto-trade-execution. Details: click here for the new SAPEPredictionServer (SPS) system that implements the new methods as software.
3. What is the SAPE4 theoretical framework for auto-modeling?
4. How to validate the auto-generated predictions?
(See the goodness-of-fit back test result; auto-back-testing on nelachip4: t_rt)
5. How will SPS help investors and institutions?
6. What is the essence of SAPE4-SPS (investor needs)?
Recall "Back to the Future - Part II": Grays Sports Almanac, a book detailing the results of major sporting events from 1950 to 2000, had made one (Biff) rich. The essence of SAPEPredictionServer is to dynamically produce an "Almanac Book for Financial Markets" having future stock prices automatically estimated with auto-generated quantitative models.
7. How to forecast next-day's sentiment like weather forecasting?
Internal use only (as on 3 local Linux servers): Realtime forecast of investor confidence
8. Prescriptive analytics on predictive analytics (solutions to future scenarios)
What will you do if SAPEngine predicts market will crash in next few days?
9. Can SAPE4 validate/assess large set of time-series models automatically?
Yes. Use the auto-back-testing function of SAPEngine for massive model validation and assessment for institutions
10. Will market crash tomorrow?
As a subset of SAPE4-SPS, SAPE4 demo app uses Web and smart devices to present daily auto-forecast of next-day market sentiment. Contact: email@example.com
11. Compare features of SAPE4-demo and SAPE4-SPS
|1 Next-day auto-forecast?||Yes||Yes|
|3 Auto-forecast beyond next day?||No||Yes|
|4 Auto-modeling; auto-prediction; auto-validation?||No||Yes|
|5 Auto-charting email?||No||Yes|
|6 Expandable beyond VIX such as SPY?||No||Yes|