forecasting method

#1
Please I wanted to determine forecasted prices of coton which are as follow
year Prices
january 2008/april 2009 636.921
january 2009/april 2010 635.219
january 2010/april 2011 659.557
New york Term(january-april 2011) 806.785
The prices are in franc CFA the new york term is the price at which potential buyers are willing to buy and it is also used to determine the forecasted future price. My problem is that I don't know which method of forecasting is suitable for these types of data .
Thank you for your help
 
#2
Hi euclude,
I have some questions about the info you gave.

First;
- Is it necessary to know the NY term? Or do you just want to estimate the real prices of cotton?
- How many datapoints do you have? Like, do you have the prices of cotton for 40 years, or just 3 years?

Supposing you are not really interested in the NY term and you can find prices of let's say 40 years, there are in my opinion 3 ways of calculating the price of cotton. From simple to advanced

1. Decomposition method (easy):
http://www.business.unr.edu/faculty/rtl/Seasonality-Final16.pdf and http://home.ubalt.edu/ntsbarsh/Business-stat/stat-data/DecomposePaper.pdf
2. Triple exponential smoothing, or ARIMA models (medium)
(depending on your skills with statistics and statistics software, this is i guess medium)
3. Econometric model building (advanced),
in which you explain the variation in price by a number of independent variables. Those variables you have to determine beforehand and data of those variables should be present.

Thus, I recommend you try the decomposition method if you're skills in statistics are limited. If you know how to work with statistical software, I would recommend the exponential smoothing or Arima models.