Monday, March 14, 2011

How to Do Curve Fitting with Polynomial Models: Finance Application

1. Read over Example 3

2. Go to
  Choose a stock (make sure it is publicly traded).

3. Choose one of three indices: Dow, Nasdaq, S&P 500
Choose Nasdaq if your stock is traded on the Nasdaq.
Choose S&P 500 if your stock is listed in the S&P 500 (google S&P!)
Choose the Dow if your stock is not traded on the Nasdaq or your stock is not in the S&P.

4. Do example 3 for your stock and for your index using the same years as in the example. Choose January 5th for each your when looking up historical prices. If January 5th falls on a weekend use January 6th (or 7th). Please use the OPENING PRICE.

5. Try a quadratic, cubic and quartic model. Use the correlation coefficient r squared to determine which polynomial model is the best. (Remember the closer to 1 the better!).

6. Using your regression model, predict the opening price for 2002 for your stock and your index. Look at the historical prices for 2002.

1. Does the price you got using your model match the actually prices? Why or Why not.
2. How over what intervals does your stock and your index both increase? Both decrease? What might account for this?
3. Over what interval does your stock increase and your index decrease?  What might account for this? 4. Over what interval does your stock decrease and your index increase?  What might account for this?

Research news about your company and about the index in the year 2002 to answer the questions

7. Repeat step 6 for the years 2005, 2008, 2010, 2011.

8. How effective do you think mathematical modeling is to predict the stock market? Explain.

1 comment:

Xiaodi said...

YAY! thank you for putting it up!