Download Revising Hong Kong's Money Supply: Including Short-term Exchange Fund Placements and more Summaries Statistics in PDF only on Docsity!
QUARTERLY BULLETIN
5/
The current practice of excluding short-term Exchange Fund placements from
the monetary aggregates has introduced minor statistical distortions. In order
to address this issue, a slight revision will be made by including all Exchange
Fund placements in money supply statistics. The revised data, as well as
their relationship with macro-economic variables, are little affected by the
change.
M3 = M2 + deposits with restricted licensed banks and deposit taking companies
- NCDs issued by restricted licensed banks and deposit taking companies and held by the public
These aggregates can be further broken down into Hong Kong dollar and foreign currency components.
The types of liabilities included in the money supply are similar to those in other economies. There are, however, a number of differences (see Annex 1 for definitions of money measures in other economies). First, for a number of industrial economies, the monetary aggregates include a larger set of instruments, such as repos and money market funds. However, such instruments are considered inappropriate for inclusion in the money supply in Hong Kong as they lack liquidity. Secondly, in most other economies, the monetary aggregates are compiled on a resident-holding basis. In contrast, in gathering Hong Kong’s money supply figures, we do not differentiate between holdings by resident and non-resident entities. 1
The third difference relates to the treatment of Government and Exchange Fund placements with banks. Government’s deposit holdings are included in the Hong Kong dollar broad money, whereas in most other economies they are
DEFINITION OF MONEY SUPPLY
I. Introduction
The HKMA compiles several measures of monetary aggregates. Nevertheless, because of the peculiar treatment of the Exchange Fund’s placements with the banking system, minor statistical distortions would arise under certain circumstances. This paper reviews the definition of money supply in Hong Kong, and sets out a proposal for revising it.
Section II reviews the current definition of money supply in Hong Kong, and shows how the treatment of Exchange Fund placements in the monetary data may give rise to statistical distortions. Section III discusses some conceptual considerations regarding the treatment of Government and Exchange Fund’s deposit holdings in the money supply. Section IV examines the empirical considerations regarding changes in the money supply definition, and a revision is proposed in Section V.
II. Definition of Money Supply
In Hong Kong, several measures of money supply are compiled:
M1 = currency held by public + demand deposits M2 = M1 + savings and time deposits with licensed banks + NCDs issued by licensed banks and held by the public
1 Nevertheless, in respect of Hong Kong dollar money supply, it is reasonable to assume that most holdings are held for domestic use, i.e. by entities with an economic interest in Hong Kong. As for the foreign currency component, its interpretation is complicated by the lack of information about the size of non-resident deposits in our banking sector.
QUARTERLY BULLETIN
5/
e x c l u d e d. F u r t h e r m o re , E x c h a n g e F u n d placements with a maturity of over one month are counted as customer deposits, and hence included in the money supply figures (see box below).
Exchange Fund placements with a remaining maturity of one month or below have been excluded from the monetary aggregates since May
- This followed a decision to deduct such placements from banks’ liquefiable assets in calculating their liquidity ratios. Against a background of rapid credit growth, the measure was introduced with the aim of preventing short- term Exchange Fund placements from contributing to banks’ credit expansion. Nevertheless, the effect of the measure on credit growth appeared to be minimal: the liquidity requirement was a prudential, rather than a monetary tool; and most banks had a liquidity ratio considerably higher than the required minimum.
As the treatment of short-term Exchange Fund placements is different from that of longer-term Exchange Fund placements and Government deposits, small statistical distortions arise under two situations. First, it may result from fiscal transfers. There will be a decline in money supply when funds are transferred from the Treasury to the Exchange Fund, and placed in the money market at the short-end. Distortions also occur when longer-term Exchange Fund placements mature, and fall within the maturity bracket of one month and below.
III. Conceptual Considerations Regarding the Treatment of Treasury and Exchange Fund Placements
The distortions noted above can be removed by (a) including shor t-term Exchange Fund placements in the monetary aggregates, or (b) excluding all Exchange Fund and Government placements from the monetary aggregates. The fundamental issue is thus whether deposits placed by the Treasury and the Exchange Fund should be included in the money supply statistics. The paragraphs below examine this issue in detail.
Treasury Placements
The IMF Manual of Monetary and Financial Statistics (2000) provides commonly used guidelines for the treatment of government deposits. It does not offer a prescription for individual economy’s definition of money. Nevertheless, it notes that “deposit holdings of the government are usually excluded from the monetary aggregates. It is argued, at least for some countries, that central government deposit holdings do not respond to macro-economic influences (i.e. changes in economic activity, interest rates, exchange rates, etc.) in the same way, or to the same degree as deposits of the money holding sectors”.
The distinction between deposits from government and other sectors is crucial in a regime where money supply statistics serve as a guidepost for monetary policy, as central banks wish to ascertain the impact of policy changes on private holdings of monetary instruments. In most cases, such actions have limited leverage on government deposits, which are often influenced by the government’s cash management policy, such as debt management and tax collection. Therefore, there are merits for excluding government deposits from monetary aggregates that serve as an intermediate target or an indicator of the stance of policy.
Broad monetary aggregates Including: Treasury deposits Exchange Fund deposits with maturity above one month
Excluding: Exchange Fund deposits with maturity of one month or below
QUARTERLY BULLETIN
5/
0
5
10
15
20
25
30
yoy%
Existing Definition Variant A Variant B
Apr-98 Oct-98 Apr-99 Oct-99 Apr-00 Oct-00 Apr-01 Oct-
Chart 1
Hong Kong Dollar M
A. Year-on-year changes
B. Month-on-month changes
0
2
4
6
8
10
Apr-98 Oct-98 Apr-99 Oct-99 Apr-00 Oct-00 Apr-01 Oct-
mom%
There are also no notable differences in terms of the predictive or corroborative properties for real GDP and prices. Nevertheless, a slightly
higher correlation of Hong Kong dollar broad money and its lags with nominal GDP is found when all the Exchange Fund and Treasur y
QUARTERLY BULLETIN
5/
0
5
10
15
20
25
30
yoy%
Existing Definition Variant A Variant B
Apr-98 Oct-98 Apr-99 Oct-99 Apr-00 Oct-00 Apr-01 Oct-
Chart 2
Foreign Currency M
A. Year-on-year changes
B. Month-on-month changes
0
2
4
6
8
10
Apr-98 Oct-98 Apr-99 Oct-99 Apr-00 Oct-00 Apr-01 Oct-
mom%
improve slightly upon including all placements of the Exchange Fund and the government. 2
2 Specifically, we employed an error-correction model to examine the empirical relationship between various definitions of HK$ M3 and a number of macroeconomic and financial variables such as income, price and interest rate. We also performed a Granger causality test to examine the predictive properties of HK$ broad money on nominal GDP, real GDP and consumer prices. The estimates were very close across various definitions of money supply. The detailed results are available upon request.
placements are included in the monetary aggregate (Table 1 and Chart 3). Our empirical work shows that the estimated money demand functions also
QUARTERLY BULLETIN
5/
Table 1 Correlation between HK$M3 and nominal GDP (Q1/1990-Q2/2001)
Definition of HK$M Existing definition, Variant A, Variant B, i.e. excluding i.e. including i.e. excluding 1-month EF all EF and Treasury and placements Treasury placements EF placements Correlation in terms of: yoy% 0.73 0.74 0. qoq% 0.36 0.39 0.
V. Conclusion
Based on the above analysis, the preferred approach to eliminate the slight distortions in the current definition of money supply would be to include the short-term Exchange Fund placements in the monetary aggregates. The HKMA will release money supply figures based on the revised definition starting from July 2002. Past data will also be adjusted to facilitate comparison.
- Prepared by Priscilla Chiu and Kitty Lai of the Research Department
REFERENCE
International Monetary Fund, Monetary and Financial Statistics Manual , Washington: International Monetary Fund, 2000.
H O N G K O N G M O N E T A R Y A U T H O R I T Y^23
QUARTERLY BULLETIN
5/
ANNEX 1
Definitions of Monetary Aggregates in Various Economies
Hong Kong Argentina Estonia ECB Japan Singapore UK US
M1 Currency + Currency + Currency + Currency + Currency + Currency + N.A. Currency + Demand Current Demand Overnight Demand Demand [No Travellers’ cheques + deposits account deposits deposits deposits deposits satisfactory Demand deposits + deposits in measure of NOW and similar pesos narrow money interest-bearing because many demand deposits interest- bearing deposits are also transactions money.]
M2 M2 = M1 + N.A. M2 = M1 + M2 = M1 + M2+CDs = M2 = M1 + N.A. M2 = M1 + Savings Savings and Time and Deposits with M1 + Savings Savings and and money market time deposits Savings an agreed and time fixed deposits deposits accounts + with licensed deposits + maturity of deposits of with banks + Small time deposits + banks + NCDs foreign up to two banks + CDs S$ NCDs Retail type money issued by currency years and of banks market mutual fund licensed banks deposits redeemable at balances + Overnight notice of up repos + Overnight to three Eurodollars months
M3 M3 = M2 + M3 = N.A. M3 = M2 + M3+CDs = M3 = M2 + M4 = All M3 = M2 + Deposits with Currency + Repos + M2+CDs + Deposits with sterling deposit Large time deposits + restricted Total deposits Specific Deposits and finance liabilities of Wholesale type licensed banks (in both local marketable CDs of companies monetary money market (RLBs) and and foreign liabilities of non-bank financial mutual fund deposit-taking currency) monetary financial institutions balances + companies financial institutions (i.e. banks Term repos + (DTCs) + institutions and building Term Euro dollars NCDs issued societies) by RLBs and DTCs
Government Included Included Included Excluded Excluded Excluded Excluded Excluded Deposits
Sources: Websites of the respective central banks/monetary authorities.
H O N G K O N G M O N E T A R Y A U T H O R I T Y^23
QUARTERLY BULLETIN
5/
