Hi all,I was looking for a good book on econometrics, which does not assume any prior knowledge.My end goal is to implement cointegration and would like to learn all basics before embarking on the development. Also are there any libraries (open source) which have an implementation for cointegration. This would help me to explore the algorithm and use it to verify my implementation. Thanks in advanceregards

http://cran.r-project.org/src/contrib/V ... trics.html "Base R ships with a lot of functionality useful for computational econometrics, in particular in the stats package. This functionality is complemented by many packages on CRAN"Wilmott Bookshop:Search790 search results: Econometrics books

QuoteOriginally posted by: dejavuHi all,I was looking for a good book on econometrics, which does not assume any prior knowledge.My end goal is to implement cointegration and would like to learn all basics before embarking on the development. Also are there any libraries (open source) which have an implementation for cointegration. This would help me to explore the algorithm and use it to verify my implementation. Thanks in advanceregardsCheck out "introduction to econometrics " by Chris brooks. Or "market Models" by Carol Alexander. The first is more introductory but the second will give you a deeper knowledge of the equations. I think there is also a book called "cointegration" with papers on the subject. Its worth to check that too.

dejavu,It all depends on what you mean by prior knowledge and the amount of time you are willing to put on this... If you are REALLY committed to learning cointegration, you should go for the Johansen (95) and the Juselius (07) books. Johansen is the theory and Juselius the application (mainly to macro questions). I found their approach the most complete and, well, they are the authorities on the subject. If you are not willing to spend much time, well, you could use a chapter of a general time series book. Like Tsay, for example. The best software for cointegration analysis, as far as I am concerned, is RATS (using CATS). Unfortunately, it isn´t free (I don´t know much about R, but it should have a decent cointegration package) and would only be worth the price if you are willing to spend a lot of resources on this.

Gujarati's book is Ok, it covers all the standard stuff. it doesnt specifically require a prior knowledge, but u should know basic probability concepts.

Hi, An introductory book on time series analysis is Applied Time Series Modelling and Forecasting by Richard Harris and Robert Sollis.You can do cointegration analysis even in EViews without any programming, just point and click. If you are looking for more than one cointegrating vector you will need to use Johansen's procedure. If not, simpler methods like the Engle-Granger method are possible, but not always recommendable. In case you use the Johansen test, you will need to specify whether your series are trended, etc. Getting this wrong invalidates the test. If you only need the very basics the EViews help file is probably sufficient. For more theory you will have to understand much more econometrics because time series econometrics is based on many results of regular, i.e. cross-sectional econometrics results. The standard text on time series econometrics is Hamilton's Time Series Analysis, but this is definitely not an introductory text but it covers everything you need.Cheers.

Boy..oh Boy!! Hamilton's book is more of an handbook..not a textbook!! and the topic here is econometric..not time series...relate stuff but not the same stuff.ok...disregarding the time series disgression stuff...and ...taking into consideration the issue started with "...a good book on econometrics, which does not assume any prior knowledge" and follows that "My end goal is to implement cointegration..."Guys...if you asked for an basic text...coitegration is not likely to be discussed there...learn to walk first...then running...Excellent introductory texts:A) Econometric Models and Economic Forecast. Pindyck & Rubinfeld.McGraw-HillB) Introduction to Econometrics. Dougherty. Oxford. (This one does have a little section on cointegration)Once you master that...we'll talk about coitegration.

Gujarati's 'basic econometrics' is indeed basic, but does discuss cointegration to the extent sufficient to understand what it's about and to start playing with it. If I could only read one book on econometrics, that'd be it. Yes the book is silly, but therefore accessible.

Assuming you have no prior knowledge of econometrics but you know some very basic mathsand probability you could use "Econometric Analysis" by William H. Greene, 6th edition.It is very readable and complete, never silly.You can take a look at the table of contents here:http://pages.stern.nyu.edu/~wgreene/Tex ... tmChapters 1-8 + chapters 16, 21 and 22 should give you a basic econometrics background.Then you might use "Applied econometric Time Series" by Walter Enders, which is very readable, orthe more advanced "Time Series Analysis" by Hamilton.I have used E-views,PcGive, Matlab and Gauss for cointegration analysis, there are many free alternatives and the choice really depends on the particular model you have to build and how much do you want to program.Have a look at GRETL: http://gretl.sourceforge.net/.Good luck!

Start with Marno Verbeek's A Guide to Modern Econometrics.It's quicker and faster than basic intro books, but gives great intuitive and mathematical explanations over everything from the Gauss-Markov Assumptions up through Co-integration and Vector Error Correction Models (VAR's that deal with cointegration). I read it along side Green and Hayashi for my first PhD level Econometrics Course. It gets to the point much quicker than either of those books and definitely gets to the point much quicker than Hamilton. It is by far the best first introduction to Econometrics I've read (I TA econometrics, now).After you read it (it won't take more than a week if you put some time into it, you can go back and read the others for a more complete understanding of the issues involved). Also, check out LeSage's Spatial Econometrics tool box for matlab. It can estimate VECM's pretty easily.

If you're also interested in cross section stuff, I'd recommend Wooldridge "Econometric Analysis of Cross Section and Panel Data". But of course there's nothing on cointegration in there...