Welcome to Scribd, the world's digital library. Read, publish, and share books and documents. See more
Download
Standard view
Full view
of .
Save to My Library
Look up keyword
Like this
8Activity
0 of .
Results for:
No results containing your search query
P. 1
bu-0612-1

bu-0612-1

Ratings: (0)|Views: 4,489 |Likes:
Published by Kostas Georgioy

More info:

Published by: Kostas Georgioy on Jun 23, 2012
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less

09/10/2014

pdf

text

original

 
1
Bulletin
|JUNE QUartEr 2012
Introduction
Assessing current economic activity is an importantpart of macroeconomic policymaking. However,official economic statistics can take some time tocompile and to be published. For example, quarterlygross domestic product (GDP) figures are releasedaround nine weeks after the end of the relevantquarter. As a result, the Reserve Bank looks at a rangeof more timely, but less complete, indicators togauge current conditions in the economy, such asthe various business surveys and the Bank’s businessliaison program.
1
 Over recent years, however, technologicaldevelopments, and the digitisation of informationand activity, have generated a vast array of electronicdata, which can potentially be analysed on a dailybasis, or even in real time. Some of these data coververy large numbers of individuals and businesses– far more than many traditional surveys used bystatistical agencies – and have the potential to beuseful for monitoring and measuring aggregateeconomic conditions. While official statisticians areincreasingly using electronic data in the production
1 For an overview of the main business surveys in Australia and howthey are used by the Reserve Bank, see Park (2011).
of economic indicators, this is still very much in itsinfancy.
2
Economists and policymakers are alsomaking greater use of electronic data to understandeconomic developments and as a cross-check ondata from official agencies. This article examines the usefulness of wholesale andretail electronic transactions data and internet searchdata in assessing current economic activity. Giventhe growth of electronic payments and internetuse by Australian households and businesses, thesedata can help to track economy-wide spending andactivity. While wholesale and retail payments dataalready provide some additional information onnational accounts aggregates, and internet searchdata also appear promising as economic indicators,these sources are expected to become even moreuseful in the future, as new technology is adoptedand electronic means of payment evolve further. Assuch, these data are worth monitoring more closely.
Wholesale Payments
Payments generated by corporates and financialinstitutions reflect a wide range of activities such aspurchases of goods, business investment, importsand exports, and financial transactions. Recently,wholesale payments data have attracted attention
2 For example, the Australian Bureau of Statistics (ABS) makes use of electronic tax collection data and Medicare data.
Electronic Indicators of Economic Activity
Troy Gill, Dilhan Perera and David Sunner*
* The authors are from Economic Analysis Department and PaymentsSettlements Department. The authors would like to thank TapasStrickland and James Bishop, formerly of Economic AnalysisDepartment, for analysis featured in this article.
Thr i  rich rry  timy high-rqcy ctric dt tht pttiy i irmtit crrt cmic cditi. I prticr, dt  ctric trcti d itrtrch c   cmpmt t mr tdrd idictr  cmic ctiity. Whi imitd cti  ctric dt i crrty d y ti ttitici i th prdcti cmic idictr, ctric dt r iy t cm  icrigy imprtt rc  irmti.
 
2
ReseRve bank of austRaliaelectRonic indicatoRs of economic activity
as a potential economic indicator, with the financialmessage service provider SWIFT releasing an indexthat helps to predict OECD GDP growth using SWIFT payments sent on behalf of corporate customers.
3
 SWIFT (2012) suggests that inclusion of customer-to-customer payment volumes data can improvethe explanatory power of a simple model of GDPgrowth.In Australia, data are available on the SWIFT paymentssettled across the Reserve Bank Information and Transfer System (RITS), which is Australia’s real-timegross settlement system.
4
Although banks can usevarious payment instruments for their customers’transactions, large-value corporate customerpayments will usually be sent using SWIFT,particularly wholesale transactions relating tobusiness investment. These data have several advantages over other moreestablished indicators of economic activity. They arevery timely, with a day’s payment data available atthe conclusion of each business day. They cover avery large number of payments and being actualfund transfers of banks and their customers arefree from reporting error and revisions. However,RITS transaction data do have some limitations. The data include financial transactions and clearlyexclude many small transactions by individuals andbusinesses, while shifts between payment methodscan introduce volatility. Also, payments betweentwo customers holding accounts at the same bank will not normally be sent to RITS for settlement andare therefore not captured in the data. The historicaltime series is also relatively short compared withmore established indicators. While electronicpayments data clearly have their limitations, it isworth noting that existing measures, such as GDP
3 SWIFT uses message types to distinguish between different businesspurposes; the index constructed by SWIFT (2012) includes customer-to-customer payments (SWIFT MT103 payments) but does notinclude bank-to-bank payments (SWIFT MT202 payments).4 Other payments settled in RITS include retail transactions such asdirect entry, cheques and card transactions, as well as transactionsarising from wholesale debt securities, equity and money markettransactions. For further discussion on the settlement of payments inRITS, see Gallagher, Gauntlett and Sunner (2010).
and gross national expenditure (GNE), are alsoimperfect estimates of actual economic activity.Nevertheless, SWIFT payments track changes inthese measures of economic activity reasonablywell (Graph 1, Table 1).
5
Interestingly, the number,rather than the value, of payments is more highlycorrelated with economic activity. This may bebecause volatility in the values series is affected bylarge financial transactions, such as swaps, whichare not directly relevant for measuring economicoutput and demand. The relationships with real andnominal measures of economic activity are similar.Given the greater emphasis on real measures of activity in economic analysis, the following analysisfocuses on the usefulness of electronic payments asan indicator of 
real 
measures of economic activity.Another way to assess the usefulness of wholesalepayments data is to observe whether they canimprove the explanatory power of models of economic activity. A test that represents a relativelylow hurdle is whether wholesale payments containinformation not already provided by a lag of the
5 For the purposes of this article, the SWIFT data were aggregated andseasonally adjusted at a quarterly frequency, after being lagged byone month; the quarterly SWIFT data have a stronger relationshipwith official measures of economic activity when lagged by onemonth, which is consistent with invoicing arrangements that typicallyallow some time for payment after receipt of the service or good.
-2-1012-4-2024-2-1012-4-2024Trend, quarterly growth
SWIFT Payments and Economic Activity*
%%Real GDP
(LHS)
SWIFT payments*
(RHS)
201220102008200620042002 Real GNE
(LHS)
*Number of SWIFT interbank payments settled in RITS, 7-periodHenderson trendSources: ABS; RBA
Graph 1
 
3
Bulletin
|june Quarter 2012
electronic indicators of economic activity
payments variable also slightly improves the models’out-of-sample predictive ability (to a greater extentthan the inclusion of the survey variable), as shownby the fall in the mean absolute error (MAE), which isthe average absolute difference between predictedand actual quarterly growth in the economic variablefor the quarter ahead.A more challenging test is whether SWIFT paymentsdata can improve models of economic activitythat already include a range of timely economicindicators. Principal component analysis can beused to summarise the information provided bysuch other indicators (Gillitzer, Kearns and Richards2005). This technique identifies the movementsof common factors (the principal components)and their importance in driving movements in aset of variables. Two first principal componentsare estimated, one based on various surveys of economic conditions (‘survey variables’), and one ona broader collection of variables including surveys,financial market indicators and official ABS statistics(‘all variables’).
7
Two corresponding baseline models
7 The survey indicators include the NAB business conditions andbusiness confidence indices, the Westpac-Melbourne Instituteconsumer sentiment index, a composite AIG business conditionsindex, and changes in the NAB survey measure of capacity utilisation.In addition to these measures, the broader collection of indicatorsincludes growth in the ANZ job advertisements series; imports; exports;retail sales; dwelling approvals; total credit; real equity, commodity anddwelling prices; and changes in the unemployment rate.
economic activity variable itself, that is, whetherpayments can improve the fit of a baseline modelwhere growth in the economy is modelled as asimple autoregressive process. In addition to thebaseline model, Equation (1) is estimated for eachactivity variable (GDP in this example):Δ
GDP 
= α
0
+
α
1
Δ
GDP 
t-1
+
α
2
Δ
SWIFT 
+
ε
(1)where
SWIFT 
is the number of payments settled perquarter, ε
 
is an error term and ∆
 
denotes quarterlyper cent growth. For comparison, Equation (2) is alsoestimated for each activity variable:Δ
GDP 
=
β
0
+ β
1
Δ
GDP 
t-1
+ β
2
survey 
+
ε
(2)where
survey 
is the NAB survey measure of businessconditions. The results suggest that SWIFT payments data doindeed contain additional information, as the fit of the models improves noticeably, with the modelsexplaining an additional 10–30 per cent of thequarterly movement in broad measures of economicactivity, relative to the baseline model (Table 2). Thisimprovement is comparable to that achieved withthe inclusion of the business conditions surveymeasure in the baseline model.
6
The inclusion of the
6 The results for
nominal 
measures of economic activity are similar tothose shown in Table 2.
Table 1: Correlations between SWIFT Payments and Economic Activity
(a)
March 2001 to March 2012, quarterly
Economic variableSWIFT payments
ValueNumber
Real GDP0.170.49Real GNE0.310.56Real domestic demand0.270.40Nominal GDP0.380.48Nominal GNE0.320.51Nominal domestic demand0.300.35
(a) Contemporaneous correlations based on seasonally adjusted data; RITS data are available from July 1998, but possible structuralbreaks restrict analysis to 2001 onwardsSource: RBA

You're Reading a Free Preview

Download
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->