Steve Koch 02:05, 7 October 2008 (EDT):I floated the idea today of having students spend one lab (two sessions) on some kind of finance / economics data analysis and presentation lab. I think students seemed very interested in this, and I think we should try to set it up before next week.
- Part of lab should be learning how to calculate net present value (NPV) of future cash flows
- What originally struck me was the need for students to learn where to find economic / government data (such as IRS revenue data; unemployment numbers; GDP; inflation; census; etc.), and being able to crunch the numbers themselves instead of relying on media and politicians.
- I want to do this without injecting politics, as much as possible. I think that should be achievable if both student and instructor have that goal.
- A goal is to find primary sources of information, and develop ways of assessing the reliability of the information and the underlying assumptions.
- Darrell Bonn had a related good idea of having students re-analyze data from a published paper. I like this idea, but am not sure (a) whether it fits in this lab and (b) how to narrow the scope or make sure it's worthwhile without picking the paper for the student.
- While this won't have anything to do with modern physics, I think the net benefit to student will be positive. First, the practical skills learned will be very useful, I think. Second, the analysis, number-crunching techniques will be applied things they are learning in 307L.
- Session 1
- Net Present Value and other calculations to understand some finance things
- Searching for reliable sources of data related to an economic / finance / or other kind of issue with data involved.
- Begin downloading and "analyzing" data
- Session 2
- Continue downloading analysis.
- Refine question being addressed
- Seek more data
- Final analysis and presentation (lots of graphs, etc.)
Comments from anyone
- Alexander T. J. Barron 17:29, 10 October 2008 (EDT): I don't know if this fits with where you're going on finance, but those of us who took Dr Boyd's PHYC 290 [here] last Spring crunched some Yahoo! stock price data over a couple decades as I recall. His MATLAB code and lecture is in the 4/11/2008 table cell. We filtered out the "noise" of rapid price fluctuations via differently shaped boxcar methods in order to get a more recognizable price trend over the long term.
- Steve Koch 01:09, 11 October 2008 (EDT): Wow! I didn't know Boyd was so awesome! I give him a lot of props for having you students do that. It's definitely along the lines of what I'm thinking: gaining practical real-world experience that is important, and learning how to do it the right way and learning skills that apply to physics along the way. Analyzing stock price data is definitely fun and something a Ph.D. physicist is likely to end up doing if he/she leaves science (I have several friends who do this). I love the stock market because there's so much data to analyze. I guess we'll have to talk about it -- I was thinking more economic or "fundamental" things (versus what they call "technical" stock price analysis). I think the student and I should discuss the goal and make sure we both think it's worthwhile.