Forecasting and the database: An analysis of databases for international business

9
Journal of' Forecasting, Vol. 4, 89-97 (1985) Forecasting and the Database: An Analysis of Databases for lnternat Business onal GlLLlAN RICE ESSAM MAHMOUD Concordia University, Montreal ABSTRACT A major consideration in the selection of a forecasting method for a specific situation is the type of pattern in the data. Before the data pattern is identified, the forecaster must recognize the dependence of any forecasting method upon the accompanying reliable database. This issue is discussed in the paper with reference to databases for international business. KEY WORDS Database International data sources A major consideration in the selection of a forecasting method for a specific situation is the type of pattern in the data. Before a data pattern is identified, it is important that the forecaster recognizes the dependence of any forecasting method upon the accompanying reliable database (Lackman, 198 1 ; Mahmoud, 1982). Proper operation and maintenance of an accurate and timely data system gives the forecaster an instrument with which to control and minimize the shortcomings of various forecasting methods. It is, therefore, essential to evaluate the databases available to verify the reliability of the data before studying the data pattern. This paper illustrates the evaluation of databases. The emphasis here is placed on international business applications. A list of international business data sources is presented with their possible forecasting applications. Also, selected data sources are further evaluated in terms of their reliability, coverage, time periods and forecasting application. FORECASTING AND THE DATABASE Normally, the database consists of two types of data: external and internal data. It is useful for the company to have access to detailzd descriptions of data with respect to various forecasting situations. This enables the managers to consider more than one forecasting model and provides opportunity for choice on the basis of the appropriateness of a model for a given task or data set. To illustrate this, the list of selected data sources in Table 1 provides the forecaster or the manager in the organization with a sample of sources available in the area of international business and their appropriate applications in the area of forecasting. In order to help the forecaster to decide which source will be most suitable in meeting a certain accuracy requirement, Table 2 gives an 0277-6693/85/010089-09$01 .OO Received July 1982 (ci 1985 by John Wiley & Sons Ltd. Reuised August 1983

Transcript of Forecasting and the database: An analysis of databases for international business

Page 1: Forecasting and the database: An analysis of databases for international business

Journal of' Forecasting, Vol. 4 , 89-97 (1985)

Forecasting and the Database: An Analysis of Databases for lnternat Business

onal

GlLLlAN RICE ESSAM MAHMOUD Concordia University, Montreal

ABSTRACT

A major consideration in the selection of a forecasting method for a specific situation is the type of pattern in the data. Before the data pattern is identified, the forecaster must recognize the dependence of any forecasting method upon the accompanying reliable database. This issue is discussed in the paper with reference to databases for international business.

KEY WORDS Database International data sources

A major consideration in the selection of a forecasting method for a specific situation is the type of pattern in the data. Before a data pattern is identified, it is important that the forecaster recognizes the dependence of any forecasting method upon the accompanying reliable database (Lackman, 198 1 ; Mahmoud, 1982). Proper operation and maintenance of an accurate and timely data system gives the forecaster an instrument with which to control and minimize the shortcomings of various forecasting methods. It is, therefore, essential to evaluate the databases available to verify the reliability of the data before studying the data pattern.

This paper illustrates the evaluation of databases. The emphasis here is placed on international business applications. A list of international business data sources is presented with their possible forecasting applications. Also, selected data sources are further evaluated in terms of their reliability, coverage, time periods and forecasting application.

FORECASTING AND THE DATABASE

Normally, the database consists of two types of data: external and internal data. I t is useful for the company to have access to detailzd descriptions of data with respect to various forecasting situations. This enables the managers to consider more than one forecasting model and provides opportunity for choice on the basis of the appropriateness of a model for a given task or data set. To illustrate this, the list of selected data sources in Table 1 provides the forecaster or the manager in the organization with a sample of sources available in the area of international business and their appropriate applications in the area of forecasting. In order to help the forecaster to decide which source will be most suitable in meeting a certain accuracy requirement, Table 2 gives an 0277-6693/85/010089-09$01 .OO Received July 1982 (ci 1985 by John Wiley & Sons Ltd. Reuised August 1983

Page 2: Forecasting and the database: An analysis of databases for international business

Tab

le I

. D

escr

iptio

n of

pos

sibl

e da

ta s

ourc

es f

or i

nter

natio

nal b

usin

ess

\o

Sour

ce

Typ

e of

dat

a

A.

Ext

erna

l dat

a I.

N

on-G

ocer

nmet

ital

Sour

ces

Uni

ted

Nat

ions

M

onth

ly B

ulle

tin o

f St

atis

tics

(Uni

ted

Nat

ions

)

Uni

ted

Nat

ions

Y

earb

ook

of I

nter

natio

nal

Tra

de S

tatis

tics

Vol

. I.

Tra

de b

y C

ount

ry

Vol

. 11.

Tra

de b

y C

omm

odity

Uni

ted

Nat

ions

St

atis

tical

Yea

rboo

k

Inte

rnat

iona

l In

tern

atio

nal

Fina

ncia

l M

onet

ary

Fund

St

atis

tics

Pred

icas

ts

Wor

ldca

sts

Dat

a de

scri

ptio

n Fo

reca

stin

g ap

plic

atio

n ex

ampl

es

% s

Prov

ides

mon

thly

inf

orm

atio

n ab

out

-pet

role

um

prod

ucts

in

all

coun

trie

s pr

o-

--ex

tern

al t

rade

ind

icat

ors

for

man

y di

ffer

-

+stir

nate

s of

mid

-yea

r po

pula

tion

stat

istic

s

-rud

e bi

rth

and

deat

h ra

tes

in

som

e

-som

e di

ffer

ent

stat

istic

s on

sel

ecte

d pr

o-

Prov

ides

the

basi

c in

form

atio

n fo

r ind

ivid

ual

coun

trie

s' ex

tern

al t

rade

per

form

ance

s in

te

rms o

f the

ove

rall

tren

ds in

cur

rent

val

ue a

s w

ell a

s in

vol

ume

and

pric

e, t

he im

port

ance

of

tr

adin

g pa

rtne

rs

and

the

sign

ifica

nce

of

indi

vidu

al c

omm

oditi

es i

mpo

rted

an

d ex

port

ed.

Maj

or s

ourc

e of

wor

ld e

cono

mic

dat

a. I

n-

clud

es i

nfor

mat

ion

on p

opul

atio

n,

man

- po

wer

, ag

ricu

lture

, min

ing,

man

ufac

turi

ng,

trad

e, w

ages

and

pric

es,

heal

th,

hous

ing,

ed

ucat

ion,

etc

. M

onth

ly i

ssue

s co

ntai

n w

orld

tab

les

pro-

vi

ding

inf

orm

atio

n on

nat

iona

l ac

coun

ts,

gove

rnm

ent

finan

ce,

inte

rest

rat

es,

pric

es.

and

prod

uctio

n, b

anki

ng, i

nter

natio

nal l

iqui

d-

ity. a

nd e

xcha

nge

rate

s.

Com

mod

ity a

nd p

rodu

ct-m

arke

t da

ta f

or

indu

stri

es s

uch

as c

hem

ical

s,

met

als,

in

- st

rum

ents

. tr

ansp

orta

tion

equi

pmen

t an

d fa

bric

ated

pro

duct

s. I

n ge

nera

l, on

e-, t

wo-

, th

ree-

and

fou

r-ye

ar p

roje

ctio

ns a

re d

evel

- op

ed.

duci

ng p

etro

leum

ent c

ount

ries

in d

iffer

ent

coun

trie

s

coun

trie

s

duct

s in

var

ious

cou

ntri

es.

Mar

ket

pote

ntia

l; tr

ade

deve

lopm

ents

; lim

ited

prod

uct-

mar

ket

fore

cast

ing.

Mar

ket

pote

ntia

l: re

gion

al l

ead-

lag

anal

ysis

to

det

erm

ine

mar

ket n

eeds

: ba

lanc

e of

pay

- m

ents

and

fin

anci

al f

orec

astin

g; i

ndus

tria

l gr

owth

pat

tern

s; p

rodu

ct-m

arke

t an

alys

is.

Mar

ket p

oten

tial;

indu

stria

l gro

wth

pat

tern

s;

labo

ur av

aila

bilit

y; te

chno

logi

cal f

orec

astin

g;

econ

omic

tren

ds.

Fina

ncia

l and

fore

ign

exch

ange

fore

cast

ing.

-i

0

?

Q

Prod

uct-

mar

ket

anal

ysis

; de

man

d pa

tter

n "

anal

ysis

bas

ed o

n in

dust

rial

gro

wth

pat

tern

s; $ 2:

te

chno

logi

cal

fore

cast

ing.

?

.-.

Page 3: Forecasting and the database: An analysis of databases for international business

The

Eco

nom

ist

Inte

llige

nce

Uni

t

The

Eco

nom

ist

Inte

llige

nce

Uni

t

Bus

ines

s In

tern

atio

nal

Cor

pora

tion

McG

raw

-Hill

Inte

rnat

iona

l B

ank

for

Rec

onst

ruct

ion

and

Dev

elop

men

t

Org

aniz

atio

n fo

r Ec

onom

ic

Coo

pera

tion

and

Dev

elop

men

t

E.I

.U.

Wor

ld O

utlo

ok 1

982

Qua

rter

ly E

cono

mic

Rev

iew

Se

rvic

e (8

3 se

para

te re

view

s co

ver o

ver

160

coun

trie

s).

Wor

ldw

ide

Econ

omic

In

dica

tors

Com

para

tive

Sum

mar

y fo

r I3

I co

untr

ies

1982

Ann

ual

B I/ D

ATA

(c

ompu

teri

zed

vers

ion)

Ec

onom

ic H

andb

ook

of t

he

Wor

ld:

1981

Wor

ld T

able

s

OE

CD

Eco

nom

ic S

urve

ys

A fe

w se

lect

ed s

tatis

tics,

but p

rim

arily

ess

ays

desc

ribi

ng p

oliti

cal

and

econ

omic

tren

ds in

- cl

udin

g sh

ort-

term

fore

cast

s. Se

para

te en

trie

s fo

r I6

0 co

untr

ies.

10

00-3

000

wor

ds d

iscu

ssin

g m

ain

tren

ds in

th

e ec

onom

y an

d fo

reca

stin

g the

m fo

r a y

ear

ahea

d. 5

000-

10,0

00 w

ords

of

new

s an

alys

is

cove

ring

pol

itica

l de

velo

pmen

ts re

leva

nt t

o an

und

erst

andi

ng o

f th

e ec

onom

y, g

over

n-

men

t eco

nom

ic po

licie

s, tr

ends

in in

vest

men

t an

d co

nsum

er sp

endi

ng; p

erfo

rman

ce o

f key

bu

sine

ss

indi

cato

rs;

eval

uatio

n of

for

eign

tr

ade

data

; as

sess

men

t of

de

velo

pmen

t pl

ans.

Cha

rts,

sta

tistic

al a

ppen

dice

s.

Each

cou

ntry

pro

file

prov

ides

key

eco

nom

ic

data

, de

mog

raph

ic a

nd l

abou

r fo

rce

data

, w

ages

and

pric

es,

fore

ign

trad

e st

atis

tics.

pr

oduc

tion

and

cons

umpt

ion

data

.

For

each

of

264

coun

trie

s/re

gion

s, l

imite

d st

atis

tics

incl

udin

g po

pula

tion.

and

ess

ays

desc

ribi

ng

econ

omic

st

ruct

ure,

do

mes

tic

tren

ds, t

rade

and

fore

ign

inve

stm

ent,

futu

re

dire

ctio

n, m

embe

rshi

p of

inte

rnat

iona

l org

an-

izat

ions

cur

rent

as

of I

Jul

y 19

80.

Econ

omic

, de

mog

raph

ic a

nd s

ocia

l da

ta

rela

ting

to p

ract

ical

ly a

ll co

untr

ies

in t

he

wor

ld.

Ann

ual

book

lets

on

any

one

of t

he 2

1 co

untr

ies,

con

tain

ing

info

rmat

ion

on re

cent

tr

ends

of

dem

and

and

outp

ut;

pric

es a

nd

wag

es,

fore

ign

trad

e an

d pa

ymen

ts,

econ

- om

ic p

olic

y an

d pr

ospe

cts.

Fore

cast

ing

polit

ical

ris

k an

d ec

onom

ic 9

tren

ds u

sing

qua

litat

ive

met

hods

. $ n n 3

Fore

cast

ing

polit

ical

ris

k an

d ec

onom

ic

tren

ds u

sing

pri

mar

ily q

ualit

ativ

e m

etho

ds

with

lim

ited

quan

titat

ive

anal

ysis

. F?

%

n Y

Mac

roec

onom

ic f

orec

astin

g; m

arke

t po

ten-

tia

l; pr

oduc

t-m

arke

t an

alys

is;

labo

ur a

vail-

ab

ility

; for

eign

tra

de tr

ends

.

Qua

litat

ive

anal

ysis

of

econ

omic

tre

nds;

po

pula

tion

grow

th.

b

2 cp Q 5

trad

e; re

gion

al m

arke

t for

ecas

ting.

%

P 2

Econ

omic

tren

ds;

soci

al c

hang

e.

s. 0s

Q

Det

aile

d in

divi

dual

cou

ntry

ana

lysi

s-m

acro

- &

econ

omic

for

ecas

ting;

tre

nds

in i

nter

natio

nal

- i3 (c

onti

nued

) *

Page 4: Forecasting and the database: An analysis of databases for international business

Tabl

e 1-

cont

inue

d

Sour

ce

Org

aniz

atio

n fo

r Ec

onom

ic

Coo

pera

tion

and

Dev

elop

men

t

11.

Gov

ernm

enta

l So

urce

s N

atio

nal

Gov

ernm

ents

U

.S.

Dep

t of

Com

mer

ce a

nd

U.S

. Fo

reig

n Se

rvic

e

B.

Inte

rnal

dat

a

C.

Com

petit

or d

ata

Type

of

data

D

ata

desc

ript

ion

Fore

cast

ing

appl

icat

ion

exam

ples

OE

CD

Fin

anci

al S

tatis

tics

Info

rmat

ion

on

finan

cial

m

arke

ts

in

16

Euro

pean

cou

ntri

es,

U.S

.A.,

Can

ada

and

Japa

n. F

ocus

es o

n ca

pita

l op

erat

ions

and

fi

nanc

ial t

rans

actio

ns w

ith fo

reig

n co

untr

ies;

ne

w s

ecur

ity is

sues

on

natio

nal

and

Eur

o-

mar

kets

; se

curi

ty p

ortf

olio

s of

the

diff

eren

t ca

tego

ries

of

inve

stor

s; in

tere

st r

ates

for

10

to 1

5 di

ffer

ent

finan

cial

ins

trum

ents

in e

ach

coun

try,

etc

.

Fina

ncia

l and

fore

ign

exch

ange

fore

cast

ing.

Nat

iona

l St

atis

tical

Abs

trac

t M

ajor

eco

nom

ic a

nd s

ocia

l ind

icat

ors

for

a pa

rtic

ular

cou

ntry

. Fo

reig

n Ec

onom

ic T

rend

R

epor

ts

Ann

ual o

r sem

i-an

nual

rep

orts

pre

pare

d on

al

mos

t ev

ery

coun

try

in t

he w

orld

. U

p-to

- da

te e

cono

mic

sum

mar

ies

incl

udin

g su

ch

topi

cs a

s ba

lanc

e of

pay

men

ts, c

redi

t ava

il-

abili

ty, i

nfla

tion,

and

inve

stm

ent

clim

ate.

Diff

eren

t fr

om o

ne fi

rm t

o an

othe

r. I

t de

pend

s on

the

oppo

rtun

ity a

nd t

he s

ize

of

the

com

pany

Mos

t fir

ms

colle

ct:-o

mpa

ny

sale

s da

ta

(a)

hist

oric

al s

ales

rev

enue

. (b

) uni

ts-fo

r to

tal c

ompa

ny an

d/or

spec

ific

prod

ucts

. T

he f

requ

ency

of

the

data

va

ries

from

one

com

pany

to

anot

her.

So

me c

ompa

nies

hav

e fa

cilit

ies t

o co

llect

da

ta o

n da

ily, w

eekl

y, m

onth

ly, q

uar-

te

rly,

or

year

ly b

ases

. So

me

com

pani

es c

olle

ct d

ata

abou

t th

eir

com

petit

ors

such

as

com

petit

ors’

sal

es a

nd

pric

es.

Reg

iona

l eco

nom

etri

c m

odel

s.

Det

aile

d in

divi

dual

cou

ntry

ana

lysi

s-fo

re-

cast

ing

of p

oliti

cal,

econ

omic

and

fin

anci

al

tren

ds u

sing

bot

h qu

alita

tive

and

quan

ti-

tativ

e m

etho

dolo

gies

.

Prod

uct-

mar

ket

anal

ysis

; sa

les

fore

cast

ing

at th

e co

mpa

ny le

vel.

Hel

p to

det

erm

ine

com

pany

’s m

arke

t sha

re

and

size

of

com

pany

sal

es f

orce

.

Page 5: Forecasting and the database: An analysis of databases for international business

Tabl

e 2.

Ev

alua

tion

of se

lect

ed d

ata

sour

ces

acco

rdin

g to

som

e cr

iteri

a

Sour

ce

Cov

erag

e A

vaila

bilit

y A

ccur

acy

Tim

elin

ess

9

Fore

cast

s 2 a

I. N

on-G

over

nmen

tal

Sour

ces

Uni

ted

Nat

ions

Y

earb

ook

of

Inte

rnat

iona

l Tra

de

Stat

istic

s V

ol.

I Tr

ade

by

Cou

ntry

V

ol. I

I Tra

de b

y C

omm

odity

Uni

ted

Nat

ions

19

79/8

0 St

atis

tical

Y

earb

ook

Euro

mon

itor

Inte

rnat

iona

l M

arke

ting

Dat

a an

d St

atis

tics

Exce

llent

-ann

ual

stat

istic

s fo

r 16

3 co

untr

ies

or r

epor

ting

cust

oms

area

s. 1

980

Yea

rboo

k gi

ves

time

serie

s I9

7 1 - 1

980.

Exce

llent

-dat

a on

eco

nom

ic

perf

orm

ance

and

pro

duct

ivity

, re

lativ

e le

vel o

f ec

onom

ic

deve

lopm

ent,

stru

ctur

e of

ec

onom

y, in

dust

rial

out

put,

ag

ricu

lture

and

tran

spor

tatio

n,

grow

th a

nd s

truc

ture

of

exte

rnal

trad

e, d

evel

opm

ent o

f en

ergy

reso

urce

s, et

c.,

grow

th

and

stru

ctur

e of

pop

ulat

ion,

em

ploy

men

t, di

stri

butio

n of

in

com

e, h

ousi

ng,

med

ical

and

cu

ltura

l se

rvic

es. U

sual

ly c

over

s 10

yea

r pe

riod

.

Goo

d-an

nual

da

ta o

n I3

2 co

untr

ies

with

res

pect

to

popu

latio

n, e

mpl

oym

ent,

prod

uctio

n, t

rade

, eco

nom

y,

stan

dard

of

livin

g,

cons

umpt

ion,

hou

sing

, hea

lth

and

educ

atio

n,

com

mun

icat

ions

, tra

vel

and

tour

ism

. M

ore

deta

iled

info

rmat

ion

on 2

5 ke

y m

arke

ts.

Goo

d-ha

rd

copy

av

aila

ble

from

U.N

. D

ept

of I

nter

natio

nal

Econ

omic

and

Soc

ial

Aff

airs

Sta

tistic

al O

ffic

e.

Goo

d-ha

rd

copy

av

aila

ble

from

U.N

. D

ept

of I

nter

natio

nal

Econ

omic

and

Soc

ial

Apd

irs

Stat

istic

al O

ffic

e.

Goo

d-ha

rd

copy

av

aila

ble

from

Eu

rom

onito

r Pu

blic

atio

ns L

td,

18

Dou

ghty

Str

eet,

Lond

on,

WC

lN 2

PN,

U.K

.

Fair-

prob

lem

s Fa

ir-

aris

ing

from

exp

ort

cont

inuo

us

and

impo

rt

publ

icat

ion

valu

atio

ns,

for

lag

of

exam

ple.

ap

prox

imat

ely

two

year

s.

Goo

d-ba

sed

on

Fair-

m

etho

dolo

gica

l co

ntin

uous

re

com

men

datio

ns

lag

of

mad

e by

U.N

. ap

prox

imat

ely

St at

ist ic

al

Com

mis

sion

. Tim

e pe

riod

, bas

e ye

ars

and

pric

es

coor

dina

ted

as fa

r as

po

ssib

le.

May

re

quir

e fu

rthe

r ad

just

men

t be

fore

be

ing

used

for

any

m

eani

ngfu

l ana

lytic

al

stud

ies.

two

year

s.

NIA

* Fa

ir-tim

e la

g va

ries

from

two

year

s or

mor

e de

pend

ing

on

coun

try

and

vari

able

co

nsid

ered

; lim

ited

time

serie

s st

atis

tics

for

a fe

w

vari

able

s on

ly.

Var

ious

- .h %

mar

ket

E

pote

ntia

l, re

gion

al l

ead

lag

anal

ysis

, 2

etc;

tim

e se

ries

& fo

reca

stin

g.

Var

ious

- m

acro

- ec

onom

ic

fore

cast

ing;

ec

onom

etri

cs;

time

serie

s fo

reca

stin

g.

> 2 ?

Econ

omet

ric

09 S'

mac

ro-

%

pote

ntia

l. 2 2

fore

cast

ing

of

econ

omic

va

riab

les;

m

arke

t

Reg

iona

l

3

b

2 an

alys

es.

a a c-

\D

w

(con

tinu

ed)

Page 6: Forecasting and the database: An analysis of databases for international business

W

P

Tabl

e 2-

cont

inue

d

Sour

ce

Cov

erag

e A

vaila

bilit

y A

ccur

acy

Tim

elin

ess

Fore

cast

s

Eur

omon

itor

Eu

rope

an M

arke

ting

Dat

a an

d St

atis

tics

Bus

ines

s In

tern

atio

nal

Wor

ldw

ide

Econ

omic

In

dica

tors

C

ompa

rativ

e Su

mm

ary

for

131

Cou

ntri

es

Ann

ual

Inte

rnat

iona

l B

ank

for

Rec

onst

ruct

ion

and

Dev

elop

men

t W

orld

Euro

pean

vol

ume

cove

rs 3

0 co

untr

ies

Goo

d-ye

arly

da

ta p

ublis

hed

in

annu

al; c

ompu

teri

zed

BI/

DA

TA

incl

udes

ann

ual

time

serie

s on

up

to 3

00 v

aria

bles

for

each

of

131

coun

trie

s fr

om 1

960

to d

ate.

Eac

h co

untr

y pr

ofile

in

clud

es d

ata

on k

ey e

cono

mic

in

dica

tors

, G

NP

/NM

P by

ac

tivity

, dem

ogra

phic

s an

d th

e la

bour

for

ce, w

ages

and

pric

es,

fore

ign

trad

e, m

isce

llane

ous

prod

uctio

n an

d co

nsum

ptio

n da

ta. T

he ra

nge

of v

aria

bles

, ho

wev

er,

is no

t as

ext

ensi

ve a

s th

at p

rovi

ded

by o

ther

sou

rces

.

Goo

d-tim

e se

ries

(1 96

5-1

977,

19

70- 1

977)

on

econ

omic

dat

a fo

r ea

ch c

ount

ry a

nd fo

r

Exce

llent

-ann

ual

data

av

aila

ble

in h

ard

copy

fo

rm. C

ompu

teri

zed

Bl/

DA

TA

is a

vaila

ble

on 3

inte

rnat

iona

l tim

e-

shar

ing

netw

orks

: -G

ener

al

Elec

tric

Info

rmat

ion

Serv

ices

C

ompa

ny’s

Mar

k I1

1 Se

rvic

e,

Serv

ices

Inc

., -D

ialo

g In

form

atio

n

-1.

P. S

harp

Ass

ocia

tes

Exce

llent

-har

d co

py

from

The

Joh

ns

Hop

kins

Uni

vers

ity

Goo

d Ex

celle

nt-

Var

ious

- B

I/D

ATA

is

spec

ial

upda

ted

ever

y an

alys

es o

f six

mon

ths.

th

ese

data

can

be

con

duct

ed

usin

g in

-hou

se

or

stan

dard

ized

pr

ogra

ms,

and

th

e da

ta c

an

be m

ixed

with

co

mpa

ny d

ata

7 to

mak

e I

rele

vant

Q

anal

yses

. 5 2

co

mpi

led

and

high

ly

Editi

on 1

980

econ

omic

?

Exce

llent

-car

eful

ly

Poor

-Sec

ond

Var

ious

-

relia

ble.

in

clud

es

tren

ds, s

ocia

l

Page 7: Forecasting and the database: An analysis of databases for international business

Tab

les

The

Sec

ond

Edi

tion

(198

0).

from

tim

e da

ta fi

les

of t

he

Wor

ld B

ank

coun

trie

s an

d co

untr

y gr

oups

, de

rived

eco

nom

ic in

dica

tors

for

se

lect

ed p

erio

ds o

f ye

ars

and

dem

ogra

phic

and

soc

ial

data

fo

r se

lect

ed y

ears

.

Inte

rnat

iona

l G

ood-

-dat

a on

int

erna

tiona

l M

onet

ary

Fund

liq

uidi

ty,

inte

rest

rat

es.

Inte

rnat

iona

l Fi

nanc

ial

exch

ange

rat

es, p

rices

of

wor

ld

Stat

istic

s tr

ade

com

mod

ities

, exp

ort

and

impo

rt p

rice

inde

xes,

con

sum

er

pric

e in

dexe

s, a

nd i

nter

natio

nal

trad

e ta

bles

for

104

coun

trie

s.

Pres

s; C

ompu

ter

tape

s fr

om E

cono

mic

and

So

cial

Dat

a D

ivis

ion,

E

cono

mic

Ana

lysi

s an

d Pr

ojec

tions

Dep

t., T

he

Wor

ld B

ank,

W

ashi

ngto

n D

C 2

0433

, U

.S.A

.

Goo

d--h

ard

copy

pu

blis

hed

by t

he

Inte

rnat

iona

l M

onet

ary

Fund

stat

istic

s up

to

1977

. (Fi

rst

editi

on

publ

ishe

d 19

76.)

Exce

llen t

4a

ta

Exce

llent

- pr

ovid

ed b

y na

tiona

l m

onth

ly

sour

ces

are

scre

ened

pu

blic

atio

n.

and

revi

ewed

bef

ore

bein

g in

clud

ed i

n th

e st

atis

tics.

chan

ge u

sing

9

econ

omet

ric

or

time

serie

s f?

tech

niqu

es.

Poor

tim

elin

ess

h

of d

ata

is a

di

sadv

anta

ge

. fo

r bu

sine

ss.

%

Tim

e se

ries

and

fore

cast

ing

of

pert

inen

t fin

anci

al

vari

able

s.

5 rp

Q $ 0

econ

omet

ric

&

11. G

over

nmen

tal

Nat

iona

l G

ood-

-maj

or

econ

omic

and

G

ood

--av

aila

ble

from

V

aria

ble-

the

qual

ity

Fair-

annu

al

Var

ious

- G

over

nmen

ts:

soci

al i

ndic

ator

s fo

r vi

rtua

lly

natio

nal g

over

nmen

t of

dat

a is

dire

ctly

da

ta. (

Onl

y m

acro

- N

atio

nal

Stat

istic

al

ever

y co

untr

y in

the

wor

ld.

conc

erne

d.

rela

ted

to th

e si

ze

curr

ent

sour

ce

econ

omic

A

bstr

act

N.B

. Q

uant

ity o

f da

ta v

arie

s.

and

wea

lth o

f th

e of

tra

de a

nd

fore

cast

ing,

Sour

ces

econ

omy

that

the

pr

oduc

tion

mar

ket

gove

rnm

ent

data

on

a po

tent

ial,

etc.

re

pres

ents

. co

untr

y.)

*rl 2

* In

form

atio

n no

t av

aila

ble.

s 2 S'

09

a &

Page 8: Forecasting and the database: An analysis of databases for international business

96 Journal of Forecasting Vol. 4, lss. No. 1

analysis of a few of the data sources in terms of coverage, availability, accuracy. time periods and forecasting application. Such an approach facilitates comparison and selection.

“Forecasts are frequently as useful as the data base upon which they are built” (Lackman, 1981). The value of a forecast basically depends on two requisites: accuracy and timeliness. Thus, forecast outcomes generally stated in the form of data must be based on timely, reliable and accurate data. Accuracy depends on the source providing the data and the competence of those compiling the data. Timeliness depends partly upon the form in which the data are stated and their accessibility.

In evaluating international databases, Cateora (1 983) suggests that four questions be asked.

1.

2. 3. How were they collected? 4.

Who collected the data? (Would there be any reason for purposely misrepresenting the facts?) For what purpose were the data collected?

Are the data internally consistent and logical in the light of known data sources or market factors?

For example, the World Bank’s ‘World Tables’ are a by-product of the World Bank’s own statistical and analytical work with population and certain financial statistics obtained from other sources. The United Nations’ figures are gathered for 125 countries, and areas submit their national statistics directly in reply to a questionnaire. In the case of the International Monetary Funds ‘International Financial Statistics’, although the data are provided by national sources, they are screened and reviewed before being included in the statistics. If the compilers have any reason to believe that the data provided by a national government are ‘doctored they simply exclude them from their publication. Another advantage of ‘International Financial Statistics’ is that the data are comparable and provided on a monthly basis. Business International’s BI/DATA is also a timely source, as data are updated every 6 months. With many international business data sources, the time lag for publication is at least 2 years (for example, some publications of the United Nations and the World Bank). From the point of view of forecasting, it is also important to consider whether time series are readily available. Here, the United Nations’ ‘Yearbook of International Trade Statistics’ or the World Bank’s ‘World Tables’, which do include time series data, may be contrasted with Euromonitor’s annual volumes which do not. Such details are described in Table 2. Similar distinctions may be made among the internal types of data available to the international manager.

CONCLUSION

From a practical standpoint, if valuable results are to be obtained from applying forecasting models, managers and forecasters must remember that a forecast is only as accurate as that set of data upon which it is based. In this paper, we have emphasized this using international business forecasting as an example. In addition, we have suggested sources of international data and criteria which can be used in their evaluation. These criteria, of course, can be applied to evaluate any database for any type of forecasting application. Only when the evaluation of the database, in terms of coverage, availability, accuracy and timeliness, has been completed, should the forecaster begin the next step in selecting an appropriate forecasting method: the identification of the data patterns.

Page 9: Forecasting and the database: An analysis of databases for international business

G . Rice and E . Mahmoud

REFERENCES

Forecasting and the Database 97

Cateora, Philip R., International Marketing, Homewood, Illinois: Richard D. Irwin, Inc., 1983. Lackman, C. L. ‘Forecasting information systems: the data base module’, Journal ofsystems Management,

January (1981) 29-35. Mahmoud. Essam. ‘Short-term forecasting: matching techniques to tasks: an integrated framework and

empirical investigation’, Unpublished doctoral dissertation, State University of New York at Buffalo, 1982.

Authors’ biographies: Gilian Rice, B.Sc. (Hon), Ph.D., is Assistant Professor of Marketing at Concordia University. She obtained her Ph.D. from the University of Bradford. She has held teaching appointments in England, the U.S.A. and Canada. Her research interests are political risk forecasting, export promotion, the export of computer software and comparative marketing. E w m Mahmoud, . B.A., MBA, Ph.D., is Assistant Professor of Quantitative Methods at Concordia University. He received his MBA and Ph.D. from the State University of New York at Buffalo (SUNY). He has held teaching and research appointments a t the University of Technology at Cairo and at SUNY. His research interests are applied forecasting, forecasting accuracy, opportunity cost as an accuracy measure and the evaluation and selection of forecasting software.

Authors’ addresses: Gillinn Rice and Essam Mahmoud, Concordia University, 1455 de Maisonneuve Blvd. West, Montreal, Quebec, Canada H3G 1M8.