• C++ Programming for Financial Engineering
    Highly recommended by thousands of MFE students. Covers essential C++ topics with applications to financial engineering. Learn more Join!
    Python for Finance with Intro to Data Science
    Gain practical understanding of Python to read, understand, and write professional Python code for your first day on the job. Learn more Join!
    An Intuition-Based Options Primer for FE
    Ideal for entry level positions interviews and graduate studies, specializing in options trading arbitrage and options valuation models. Learn more Join!

Need Input Please - Weather SEA application

finssite

Joined
11/17/07
Messages
21
Points
11
Hello Everyone,

Please provide your thoughts on the following "Weather SEA" application that I plan to implement.

My goal with this application is to develop an options-trading rules engine that adapts to the dynamics of weather projections, price movements, and mother nature. It is possible that the results show nothing spectacular at the end but then it could also just be me. Hence, I would like to stay close to the Quantnet community for feedback and suggestions. I have several other ideas like how to incorporate emissions, and energy markets into this application.


PHASE Scientific -
  1. On a daily basis, weather data is mapped to the stock open, high, low, and close prices of several companies.
  2. On a daily basis, statistics is gathered on how the stock open, high, low, and close prices changed over the previous two weeks of actual vs predicted weather data.
PHASE Entrepreneurial -
  1. Develop a rules engine to trade options.
  2. Integrate the rules engine into a demo trading application.
PHASE Artistic -
  1. On a regular basis, companies are characterized into several groups based on geographic, business, financial, economic, and other factors.
  2. On a regular basis, statistics within these groups are analyzed.
I would implement these phases in parallel for best feedback to each other.

I am not trying to study how weather affects business or else I would have mined through years of data and done relevant research to find seasonality and weather-related irregular components. However, I would load the historical data to test my current findings in simple ways and that if it turns out useful I might even consider a neural network application. All this would be a lot of work particularly in creating the model compared to what the above described application requires.

I plan to use Java and Oracle to build this application on a Linux machine. I use Eclipse and have come across some charting libraries on sourceforge.net earlier. I have taken Probability and Statistics coursework and perhaps need to learn a lot more mathematics to make this application of any relevance. I have been studying on Time Series Analysis a bit and realized that I have already read on so many different topics in so many different papers but that without applying them it feels like eating protein without a workout. I also don't know from where I can download all the information required for this application.

A few specific questions are:
  1. Where can I find stock open, high, low, and close prices data to download on a daily basis?
  2. Where can I find stock open, high, low, and close prices historical data over the past several years to download?
  3. Where can I find weather data around the world to download on a daily basis?
  4. Where can I find weather historical data around the world over the past several years to download?
  5. What would be good strategies to map weather to stocks?
  6. What would be useful statistical and other results to pursue?
  7. Which are freely available Java-based trading engines?
  8. Where can I find options data to download on a daily basis?
  9. Where can I find options historical data over the past several years to download?
  10. What are potential problems or points of failure I could face while implementing this application? What can I do to overcome them?
  11. Did I make sense above? If not, where did I lose you?
Thank You,
Shakti
 
finssite said:
A few specific questions are:
  1. Where can I find stock open, high, low, and close prices data to download on a daily basis?
  2. Where can I find stock open, high, low, and close prices historical data over the past several years to download?
  3. Where can I find weather data around the world to download on a daily basis?
  4. Where can I find weather historical data around the world over the past several years to download?
  5. What would be good strategies to map weather to stocks?
  6. What would be useful statistical and other results to pursue?
  7. Which are freely available Java-based trading engines?
  8. Where can I find options data to download on a daily basis?
  9. Where can I find options historical data over the past several years to download?
  10. What are potential problems or points of failure I could face while implementing this application? What can I do to overcome them?
  11. Did I make sense above? If not, where did I lose you?
Thank You,
Shakti
Click to expand...

Some answers:

1) If you want free, try Yahoo. If you want a paid service, there are multiples. Try eSignal. I use Primate (www.primate.com)

2) See #1

3) I think the US weather service has this information, at least about US. NOAA's National Weather Service

4) See #4. I dont' think you will find worldwide info very easy though.

5) I don't know much about this but I think you should assign some weights and factors for your model. This has to be your own creation.

6) See #5

7) Trading engines? Nobody offers their engines for free. Keep that in mind... and if somebody gives you a trading engine for free... BE AWARE!!

8) You will have to pay for this data. It might be free but you will have to do a major cleanup.

9) See #8

10) Your main issue initially will be data. It is very hard to find. You will need to do major research but if you succeed, I guarantee you the reward will be awsome. Just by re-packaging the data you will be able to make bundles.
 
Back
Top