// *****************************************************************************
// *****************************************************************************


*******************************OVERVIEW*********************************

    Purpose of this read-me file.
This "read-me" file summarizes the output for the the empirical results in:
   Neil R. Ericsson and Steven B. Kamin (September 2008)
   "Constructive Data Mining: Modeling Argentine Broad Money Demand",
    International Finance Discussion Paper (IFDP) No. 943,
    Board of Governors of the Federal Reserve System, Washington, D.C.
A virtually identical version of this paper will be published as:
   Neil R. Ericsson and Steven B. Kamin (2008, in press)
   "Constructive Data Mining: Modeling Argentine Broad Money Demand",
    in Jennifer L. Castle and Neil Shephard (editors)
   "The Methodology and Practice of Econometrics:
    A Festschrift in Honour of David F. Hendry", 
    Oxford, Oxford University Press.
The included output also replicates some results that appeared in:
   Steven B. Kamin and Neil R. Ericsson (November 1993)
   "Dollarization in Argentina",
    International Finance Discussion Paper (IFDP) No. 460,
    Board of Governors of the Federal Reserve System, Washington, D.C.
and:
   Steven B. Kamin and Neil R. Ericsson (April 2003)
   "Dollarization in Post-hyperinflationary Argentina",
   Journal of International Money and Finance, 22, 2, 185-211.

    Distribution and citation:
These files may be used freely for non-commercial purposes, 
provided that proper attribution is made in any report or publication 
involving their application.  To facilitate replication and validation 
of empirical findings, please cite Ericsson and Kamin (2008), as
referenced above.  E.g., in the acknowledgements footnote of an article 
reporting estimates of a model using this data, include the sentence: 
"The model estimated herein uses data from Ericsson and Kamin (2008)."
The data, batch, and output files, either in whole or in part, may not 
be modified or be incorporated into any report, publication, computer 
program, or other media without the express, written consent of both authors.

    Feedback:
The authors welcome feedback regarding the use of this data.  
Please send comments, suggestions, and new results to the contact address 
given below.

This is file:                    IFDP943-EricssonKamin-ReadMe-Sept2008.txt.
Author/Contact:                  Neil R. Ericsson (ericsson@frb.gov)
                                 Stop 15, Federal Reserve Board
                                 Washington, D.C. 20551 U.S.A.
                                 (1) (202) 452-3709 (office)
                                 (1) (202) 263-4850 (fax)
Date:                            September 3, 2008


*******************************THE DATA*********************************

Argentine broad money demand data, monthly, January 1977-January 1993.
        Here are Excel datasets (.xls and .csv formats) of the raw data.
        The .xls and .csv files contain identical sets of data.
        The .pdf file is a convenient viewable and printable version.
   08/01/2008  10:26 AM    74,240 ArgentineMoneyDemand.xls
   08/01/2008  10:27 AM    11,116 ArgentineMoneyDemand.csv
   08/01/2008  10:28 AM    15,186 ArgentineMoneyDemand.pdf
        Here are the OxMetrics 5 transformations applied to the raw data.
        The .alg file is the set of transformations.
        The .out file is the corresponding OxMetrics output.
   08/01/2008  11:21 AM     5,911 ArgentineMdDataTransformations.alg
   08/01/2008  11:25 AM     6,381 ArgentineMd-DataTransformations-Aug1.out
        Here are OxMetrics (.in7/.bn7) and Excel files of the transformed data.
        The OxMetrics and Excel files contain identical sets of data.
   08/01/2008  11:23 AM   123,136 ArgentineMd.bn7
   08/01/2008  11:23 AM     5,442 ArgentineMd.in7
   08/01/2008  11:24 AM   289,876 ArgentineMd.xls
   08/01/2008  11:24 AM   190,809 ArgentineMd.csv

Notation for the data.
        Uppercase denotes the level of the variable.
        Lowercase denotes the log of the variable.
        A suffix "_j" denotes the j-th lag of the variable.
        "D" denotes the first difference (Dx = x_{t} - x_{t-1} for x).
        "D6" denotes the six-month difference (D6x = x_{t} - x_{t-6} for x).
        "#NA" in .xls and .csv files denotes a missing value.

Raw data and transformations.
        M1  Narrow measure of money ("M1")
        M3  Broad measure of money ("M3")
        P   Consumer price index (CPI)
        R   Interest rate on domestic peso-denominated 30-day fixed-term 
               bank deposits
        Y   Gross domestic product
        E   Free-market exchange rate
        See Ericsson and Kamin (2008) and Appendix 1 of 
        Kamin and Ericsson (1993) for further details on
        the data, including their published sources.  Details of the 
        transformed data appear in the .alg file above.
        The variables M1 and Y are not used in Ericsson and Kamin (2008).


*********************THE BATCH AND OUTPUT FILES*************************

Output is from OxMetrics 5.00, PcGive 12.00, PcGets 1.02, and Autometrics 1.50.
Most output files (.out) are generated by OxMetrics batch files (.fl).

The augmented Dickey-Fuller (ADF) statistics (Table 1).
   08/01/2008  04:03 PM    28,965 ArgentineMd-ADFs-Aug1.out

The cointegration analysis (Tables 2 and 3).
   08/01/2008  05:12 PM    28,068 ArgentineMd-Cointegration-Aug1.fl
   08/01/2008  05:59 PM   196,642 ArgentineMd-Cointegration-Aug1.out

The autoregressive distributed lag (ADL) and related models (Table 4).
   08/06/2008  05:37 PM    11,806 ArgentineMd-ADLs-Aug6.fl
   08/06/2008  07:23 PM    97,217 ArgentineMd-ADLs-Aug6.out

The PcGets output (Table 5).
   08/07/2008  12:18 PM    12,697 ArgentineMd-PcGets-Table5-upper-Aug7.fl
   08/07/2008  05:38 PM   850,387 ArgentineMd-PcGets-Table5-upper-Aug7.out
   08/07/2008  07:29 PM     9,124 ArgentineMd-PcGets-Table5-lower-Aug7.fl
   08/07/2008  07:55 PM   165,709 ArgentineMd-PcGets-Table5-lower-Aug7.out

The Autometrics output (Table 6).
   08/08/2008  11:38 AM    13,059 ArgentineMd-Autometrics-Table6-upper-Aug8.fl
   08/08/2008  11:40 AM   287,432 ArgentineMd-Autometrics-Table6-upper-Aug8.out
   08/08/2008  06:34 PM     9,027 ArgentineMd-Autometrics-Table6-lower-Aug8.fl
   08/08/2008  07:05 PM   133,749 ArgentineMd-Autometrics-Table6-lower-Aug8.out

Replication of the final equation (Equation (7)) and related calculations.
   08/08/2008  08:15 PM     9,197 ArgentineMd-NewECM-Aug8.fl
   08/08/2008  08:24 PM    51,257 ArgentineMd-NewECM-Aug8.out


FINISHED.
// *****************************************************************************
// *****************************************************************************