5 People S Republic of China: Zhang Mingli and Hui Liu

5
Peoples R epublic of China

Zhang Mingli and Hui Liu
I. Fiscal Policy and M anagem ent
In the Peoples Republic of China (PRC), scal policy has three

objectives:
(i) Channel social resources to the departments that the Government

wants to support. This includes the transfer of capital to the state-
owned enterprises (SOEs) as well as the nancing of public
expenditure.
(ii) Incom e distribution This function uses tools such as subsidies to
maintain the states stability. This important function is done by
protecting groups such as workers and peasants.
(iii) Stabilizing uctuations of the econom ic cycle In the short term,
this is realized by adjustment of aggregate demand, and in the
long term by adjustment of aggregate supply. The PRC has recently
been forced to change its policy of weak nance, strong banks,
making debt policy an important ingredient of positive scal policy.
Expenditure and revenue have increased steadily due to rapid eco-

nomic growth, with scal revenue increasing by 5.3 times between 1985
and 1998, and scal expenditure by ve times.
Meanwhile, scal decit increased at a di¬erent pace. Before 1994,
the budget included the income and expenditures of government bonds,
and debt issuance was regarded as a source of scal income. Therefore
the decit was not very high, although it increased year by year. Since
1994, the scal budget has not included income and expenditure of
government debt, and the Ministry of Finance (MOF) has not been al-
lowed to overdraw from the Central Bank. Fiscal decit leaped to more
than Y50 billion in 1997, and to more than Y90 billion in 1998 because
of the positive scal policy introduced after the Asian nancial crisis.
Although the absolute amount of scal revenue/expenditure is in-
creasing steadily, the relative proportion to gross domestic product (GDP)
142 Government Bond Market Development in Asia
is falling as a result of fast economic development and a governmental

tax system, which is not comprehensive. The lessen taxes-add subsidies
policy, designed to support the reform of state-owned enterprises (SOEs)
is another important factor.
TA BL E 1
Fiscal R evenue and Expenditure as Percentage of GDP
Year Fiscal Revenue Fiscal Expenditure

1982 22.90 23.23
1983 23.03 23.75
1984 22.91 23.72
1985 22.36 22.36
1986 20.80 21.61
1987 18.39 18.91
1988 15.79 16.69
1989 15.76 16.70
1990 15.84 16.63
1991 14.57 15.67
1992 13.08 14.05
1993 12.56 13.40
1994 11.16 12.39
1995 10.67 11.67
1996 10.91 11.69
1997 11.62 12.40
1998 12.44 13.60
Source: China Statistical Yearbook 1999.
Central nance in the PRC is rather weak. While the central Gov-
ernment has a relatively stable proportion of overall government revenue,
there is a clear downward trend. Meanwhile, the level of its expenditure
is rising. In 1994, the ratio of central government expenses to overall
government expenditure leaped from 22.02 percent in 1993 to 55.70
percent.
Subsidies to loss-making SOEs are not viewed as expense items
but rather as revenue items, so the growth of subsidies in fact does not
increase the Governments expenditure, but rather reduces revenue. There
is no special mechanism to restrain this except indirect decit pressure.
A . Governm ent R evenue and Expenditure
Government revenue in PRC is classied as follows: (i) budget

revenue; (ii) extrabudgetary revenue, and (iii) extrasupervision revenue.
The biggest source of budgetary income is the taxes and prot
Peoples Republic of China 143
FIGUR E 1
Central Governm ent R evenue and Expenditure
(percentage of GDP)
60
50
40
Percent
30
20
10
0
’85 ’87 ’88 ’90 ’91 ’93 ’94 ’96 ’97
Year
Central Government Central Government

Revenue Expenditure
earned by SOEs. Statistics show that more than 60 percent of govern-

ment income comes from SOEs. In recent years, however, SOEs have
become decreasingly protable and increasingly subsidized, weakening
the basis of government income, and making the Government ever more
dependent on debt issuance. Under conditions of high repayment and
issuance costs, debt issue will demand larger repayment in the future,
creating a cycle of larger and larger debt issuance.
Extrabudgetary revenue is not planned in the central Governments
annual budget, but is controlled by local governments or departments,
respectively.
Extrasupervision revenue is controlled by policy performers inwardly.
It is outside the system of government revenue, and is usually used to
provide for the welfare of small groups, such as peasants or farmers.
The proportion of budgetary revenue has been gradually decreas-
ing, while extrasupervision revenue has become larger and larger, with
much income raised by subordinate authorities in the form of fees or
charges. This has weakened the Governments scal power, and often
creates policy dilemmas. Weak budget nance means central Govern-
ment must depend more on debt issuance, but at the same time it adds
to the burden on central Government, often meaning it cannot support
this larger issuance. Before the Asian nancial crisis, the economy did
not run too badly under this system, but improvements in the scal
revenue system are now more and more urgently needed.
FIGUR E 2
Governm ent R evenue by Classication
(percent)
100%
80%
60%
40%
20%
0%
’91 ’92 ’94 ’95 ’96 ’98
Year
Fiscal Extra-budgetary Extra-supervision
A key problem is the e¹cient use of government expenses. Gross

expenditure is planned by MOF and approved by the Standing Commit-
tee of the National Peoples Congress (SCNPC). The users are state
ministries and other departments. Once they have their quota of budget,
they can decide how to use the money by themselves, without supervi-
sion, resulting in inevitable ine¹ciency.
In Table 3, note that subsidies to loss-making SOEs are not in the
revenue but rather in the expense items.
B. Current Problem s and Future Task s
Compared with the healthy scal systems of modern developed

countries, the Governments position is distorted. To improve this situa-
tion, it must stay out of many business areas and devote itself instead to
molding a healthy business and legal environment.
TA BL E 2
Gross R evenue and Expenditure
(billion yuan)
Current Prices Constant Prices Balance

Year Index (preceding year = 100) (current
Revenue Expenditures Revenue Expenditures prices
1985 186.64 184.48 124.3 119.3 2.16
1986 226.03 233.08 121.1 126.3 7.05
1987 236.89 244.85 104.8 105.0 7.96
1988 262.80 270.66 110.9 110.5 7.86
1989 294.79 304.02 112.2 112.3 9.23
1990 331.26 345.22 112.4 113.6 13.96
1991 361.09 381.36 109.0 110.5 20.27
1992 415.31 438.97 115.0 115.1 23.66
1993 508.82 528.74 122.5 120.5 19.92
1994 521.81 579.26 120.0 124.8 57.45
1995 624.22 682.37 119.6 117.8 58.15
1996 740.80 793.76 118.7 116.3 52.96
1997 865.11 923.36 116.8 116.3 58.24
1998 9,876.00 1,078.92 114.2 116.9 92.23
Note: Since 1994, scal revenue and expenditure no longer include government debts.
That means data before 1994 include debt revenue and costs. The inuence of price has
been excluded from all data.
Source: China Securities and Futures Statistical Yearbook 1999.
A major problem in scal revenue management is the fact that

extrabudgetary and extrasupervision revenue have reached such an ex-
tent that they a¬ect normal budgetary constraint and encourage
subauthorities to abuse their power. This latent revenue cannot be used
to support central Government, which in turn increases scal riskone
of the impediments to enlarging government debt issuance. Regulating
extrabudgetary income is an important task for the future, which re-
quires these fees and charges to be converted into taxes.
Meanwhile, tax income has not kept up with the increasing pace
of GDP, and there is a lag in tax collection from SOEs. Extension of the
tax base is urgently needed, to include whole economic entities, espe-
cially private and collective entities, as well as personal income taxes.
Another key problem is the system of supervision, which currently
only exists at the stage of plan-making, and means e¹ciency is well
below average. Within the central government scal system particularly,
users have no special cost constraint, the cost burden being wholly borne
by the central Government resulting in scal misappropriation. This oc-
curs very often in the use of capital raised by debt issuance. Only the
TA BL E 3
Governm ent R evenue by Source
(billion yuan)
Revenue from
Funds for Key Revenue Revenue
Subsidies Construction from from
Revenue to Loss- Projects in Budget Extra Other
Year Revenue Taxes from Making Energy and Adjustment Charges
SOEs Enterprises Transportation Fund for
Industries Education
1985 20.05 20.41 0.044 0.51 0.146 0.280

1986 21.22 20.91 0.042 0.32 0.157 0.157
1987 21.99 21.40 0.043 0.38 0.180 0.212
1988 23.57 23.90 0.051 0.45 0.186 0.176
1989 26.65 27.27 0.064 0.60 0.202 0.091 0.179
1990 29.37 28.22 0.078 0.58 0.185 0.131 0.299
1991 31.49 29.90 0.075 0.51 0.188 0.138 0.028 0.240
1992 34.83 32.97 0.060 0.44 0.157 0.117 0.032 0.265
1993 43.49 42.55 0.050 0.41 0.118 0.102 0.044 0.191
1994 52.18 51.27 0.37 0.054 0.059 0.064 0.280
1995 62.42 60.38 0.33 0.017 0.035 0.083 0.396
1996 74.08 69.10 0.34 0.004 0.004 0.096 0.724
1997 86.51 82.34 0.37 0.103 0.682
1998 98.76 92.63 0.33 0.113 0.833
Note: The gures do not include revenue from debt issuance.

central Government can issue T-bonds, and it bears the entire cost, but
lack of supervision drains the e¹ciency of debt issuance. The cost-ben-
et constraint on each user of scal capital must be improved.
Discrimination against the nonstate-owned economy is another serious
cause of the low e¹ciency of positive scal policy.
C. Seigniorage Incom e
The Central Banks seigniorage income (the accretion of surplus of

reserve money in the balance sheets of monetary authorities) reached
Y612.87 billion in 1996almost as much as income tax.1 It is created
by mintage approved by the State, and so should actually belong to
scal income. However, most of it is left to the Central Bank. MOF only
receives about Y20 billion from the Central Bank each year, about 10
percent of the supposed level.
In addition, this income is used ine¹ciently. Apart from some used
1. Income tax in 1996 was Y690.982 billion.

by the Central Bank itself, most is transferred to state-owned banks (SOBs).

Some is squandered, and most is transferred to SOEs as loans. SOE in-
e¹ciency results in lavish waste of capital, inevitably leading to poor
nance for the Government. The accumulated nancial risks are left in
commercial banks.
Within this closed and insular cycle, the Government takes on a
great deal of extra expenditure. As well as supporting the banking sys-
tem through seigniorage income and SOEs through subsidies and other
privileges, it must nd a way of providing a healthy economic and legal
environment through public investment.
Positive bond policy has thus had low e¹cacy because capital
from debt issue is used mainly by SOEs and cannot stimulate the major
current investors in the economy to invest. The situation worsened once
the SOBs lost the desire to support SOEs after the Asian crisis. Banks
have not been used to supporting other entities, and so superuous capi-
tal piled up within the SOBs.
This insular and closed cycle must be broken, possibly by devel-
oping a direct nancing market to break the dominance of the bank
system. New and indirect control policy tools must be developed by
cultivating a commercialized market. A developed and integrated debt
market is absolutely crucial.
D. Financing of Fiscal Decits
In terms of debt, the proportion of domestic debt is getting bigger

and bigger, accounting for 98 percent of the total in 1998. Foreign bor-
rowing was once an important factor, in 1993 accounting for 48 percent,
but by 1998 this had fallen to 2 percent.
Although foreign debt still accounts for a large amount of the
balance, the PRC is now more conscious of the huge potential of the
domestic debt market. Particularly since 1993, the Government has wanted
to utilize domestic debt better to decrease the risk of foreign debts.
Fiscal decit increased to Y92.22 billion in 1998 from Y1.765 billion
in 1982. From 1982 to 1993, the amount of T-bonds issued and the amount
of scal decit both increased, but at a di¬erent pace. From 1994 to 1997,
scal decit was kept within a very narrow range of Y52 billion to Y58
billion. The amount of T-bonds issued grew more quickly, however, increasing
to Y241.20 billion in 1997 from Y31.478 billion in 1993. This means scal
decit has not been the main reason for the high growth of debt issuance.
In fact, the scal decit redemption was much more dependent on
overdrawing from the Central Bank until 1994. Since this practice was
banned that year, however, decit has had to be met by issuing debt.
TA BL E 4
Fiscal Expenditure Item s
(billion yuan)
Social, Culture
Year Total Economic and Educational National Government
Construction Development Defense Administration Others
1985 20.04 11.28 4.08 1.91 1.71 1.06

1986 22.05 11.59 4.85 2.01 2.20 1.40
1987 22.62 11.53 5.05 2.10 2.28 1.65
1988 24.91 12.58 5.81 2.18 2.71 1.62
1989 28.24 12.91 6.68 2.51 3.87 2.26
1990 30.84 13.68 7.38 2.90 4.14 2.73
1991 33.87 14.28 8.50 3.30 4.10 3.64
1992 37.42 16.13 9.70 3.78 4.65 3.18
1993 46.42 18.35 11 . 7 8 4.26 6.34 5.69
1994 57.93 23.94 15 . 0 2 5.51 8.48 4.99
1995 68.24 28.56 17 . 5 7 6.37 9.97 5.78
1996 79.38 32.34 20.8 1 7.20 11.85 7.18
1997 92.34 36.47 24.6 9 8.13 13.59 9.45
FIGUR E 3
Com parison of Dom estic and Foreign Debt (1993 and 1998)
1993 1998
Domestic Foreign
Debt Borrowing
9% 2%
Other
Domestic
Debt
48%
Foreign
Borrowing
43%
Domestic
Debt
98%
TA BL E 5
Deficit and Dom estic Debt Issuance
(billion yuan)
Year Fiscal Deficit Domestic Debt Issued

1982 1.765 4.383
1983 4.257 4.158
1984 5.816 4.253
1985 0.057 6.061
1986 8.290 6.251
1987 6.283 6.307
1988 13.400 9.217
1989 15.888 5.607
1990 14.649 9.346
1991 23.714 19.930
1992 25.883 39.564
1993 29.335 31.478
1994 57.452 102.857
1995 58.152 150.186
1996 52.956 184.777
1997 58.242 241.202
1998 92.223 380.870
Source: China Statistical Yearbook 1999, p. 265.
Fiscal policy does not require the annual government budget to

include the income and expense of debts. Thus, the repayment of debt
has to be covered by debt income. In conditions of ever-greater repay-
ment pressure this directly causes the expansion of debt volume.
II. M onetary Policy and M anagem ent
A . M onetary Policy Tools
The Peoples Bank of China (PBC), which began to act as the

Central Bank in 1984, carries out its monetary policy through the mecha-
nism of required and excessive reserves. Besides oating nominal interest
rates and required reserve rates, excessive reserve is very important. Be-
cause of the ine¹ciency of SOBs, required reserves cannot meet the
demand of drawing. Excessive reserves are necessary to maintain the
stability of the banking system. The PRC has experienced ination for
about 20 years, and in such a situation SOBs have an incentive to use
the excessive reserves.
To make banks withdraw the expansion impulse, PBC had to pay
interest rates to the banks excessive reserve account. Today it pays only
for the required reserves. If the ratio of excessive reserves fell below 5
percent, the banks would be confronted with a liquidity risk. Therefore,
excessive reserves led the PBC to loosen the annual loan quota. While
in depression, the excessive reserves were a safe haven. Banks could
increase them to gain a stable income without any risk, reducing their
willingness to make loans.
The PBC for sometime paid a higher premium for banks excessive
reserves than the interest rate of deposits. Within a closed and insular
cycle, the PBC only had recourse to three policy tools: nominal interest
rates, loan planning, and required and excessive reserve accounts. Only
the excessive reserve actually acted as a tool for daily and seasonal
tuning. This was abolished in December 1998, however, because of banks
decreasing desire to make loans. The PBC thus lost a key policy tool.
OMO is touted as the most likely substitute.
TA BL E 6
Interest R ates for R equired and Excessive R eserve
(percent)
Required Reserve Excessive Reserve

21 Aug. 1990 6.84 6.84
21 Apr. 1991 6.12 6.12
15 May 1993 7.56 7.56
11 Jul. 1993 9.18 9.18
1 May 1996 8.82 8.82
23 Aug. 1996 8.28 7.92
23 Oct. 1997 7.56 7.02
25 Mar. 1998 5.22 5.22
1 Jul. 1998 3.51 3.51
7 Dec. 1998 3.24
10 Jun. 1999 2.07
Note: The account of excessive reserve was abolished in 1998.

Source: Peoples Bank of China Quarterly Statistical Bulletin 1999, p. 37.
The annual planning of loans was abolished at the same time.

Previously, the quota was planned every year and strictly carried out.
Now, central Government modies the supply scale of loans by control-
ling fundamental currency (M1 and M2), and all commercial banks must
stick to the management of ratio of capital to current liabilities. This
reform built an initiative form of modern banking and the PBC acquired
a primary working platform for macro-control.
Interest rates. The PRC has a xed interest rate system, with the rate
regulated by the PBC. The deposit interest rate is xed and cannot be
oated. The PBC has lowered nominal interest rates six times since 1996
according to economic circumstances.
TA BL E 7
Interest R ates for M ajor Deposits
(percent)
Demand Three- Six- One- Two- Three- Five-

Deposit month month year year year year
Time Time Time Time Time Time
Deposit Deposit Deposit Deposit Deposit Deposit
21 Apr. 1991 1.80 3.24 5.40 7.56 7.92 8.28 9.00

15 May 1993 2.16 4.86 7.20 9.18 9.90 10.80 12.06
11 Jul. 1993 3.15 6.66 9.00 10.98 11.70 12.24 13.86
1 May 1996 2.97 4.86 7.20 9.18 9.90 10.80 12.06
23 Aug. 1996 1.98 3.33 5.40 7.47 7.92 8.28 9.00
23 Oct. 1997 1.71 2.88 4.14 5.67 5.94 6.21 6.66
25 Mar. 1998 1.71 2.88 4.14 5.22 5.58 6.21 6.66
1 Jul. 1998 1.44 2.79 3.96 4.77 4.86 4.95 5.22
7 Dec. 1998 1.44 2.79 3.33 3.78 3.96 4.14 4.50
10 Jun. 1999 0.99 1.98 2.16 2.25 2.43 2.70 2.88
Sources: China Statistical Yearbook 1997 and 1999, Peoples Bank of China Quarterly
Statistical Bulletin 19944, Volume XVI.
The authorities regulate the interest rates of loans too. Since No-
vember 1998, nancial institutions have had the right to oat up to 20
percent, according to the clients risk, which was raised to 30 percent in
September 1999. Only short-term loans can be oated.
TA BL E 8
R egulated Interest R ates of L oans
(percent)
Three-
Six- One- year or Five-year More than
month year Less or Less Five-year
1 May 1996 9.72 10.98 13.14 14.94 15.12
23 Aug. 1996 9.18 10.08 10.98 11.70 12.42
23 Oct. 1997 7.65 8.64 9.36 9.90 10.53
25 Mar. 1998 7.02 7.92 9.00 9.72 10.35
1 Jul. 1998 6.57 6.93 7.11 7.65 8.01
7 Dec. 1998 6.12 6.39 6.66 7.20 7.56
10 Jun. 1999 5.58 5.85 5.94 6.03 6.21
Source: Peoples Bank of China Quarterly Bulletin 19994, Volume XVI.

Nominal interest rates before 1997 discriminated heavily against

individuals engaged in nonstate-owned industrial and commercial busi-
ness, with loans for them all carrying the united premium for risk: oating
over 20 percent, no matter whether the risk was high or low. While the
nonstate-owned economy has created more than 56 percent of the whole
output value of the country, the percentage of bank loans they can ob-
tain is less than 10 percent. SOES, on the other hand, receive continuous,
special nancial support through loans, even though their ratio of out-
put value is becoming smaller and smaller the (Figure 4). This has caused
many policy problems. When the PRC fell into deation after the Asian
nancial crisis, the Government wanted to stimulate the economy through
a positive scal policy supported by debt issuance. But the monetary
system was not able to coordinate loan structure well, because the tradi-
tional bank system is designed to support SOEs rather than to stimulate
private investment. The PRCs bank system must nd ways to reform
these mechanisms to adapt to a true market economy.
The positive di¬erential between deposits and loans since the start
of the reform of the SOB system in 1994 is another important phenom-
enon.2 High savings have led to bank capital supply exceeding demand.
FIGUR E 4
Com parison of SOEs and Non-SOEs
100 100
80 80
Percent
Percent
60 60
40 40
20 20
0 0
’89 ’90 ’92 ’93 ’94 ’95 ’97 ’89 ’90 ’92 ’93 ’94 ’95 ’97
Year Year
SOEs Private Enterprises Collective Enterprise
Sources: Wu Jian, Savings, Investment and Economic Growth, Economic Research

Journal, No.11, November 1999.
2. Di¬erential between deposits and loans refers to balance of deposits

minus balance of loans.
Since the Asian crisis, the PRC has strengthened management of

the ratio of capital to current liabilities to decrease nancial risk, and
the proportion of new bad loans is strictly limited. With SOEs earning
less and less, SOBs have lost their interest in making loans to them.
Rapidly growing individual savings are mainly deposited in SOBs, and
used for the SOEs. Since the crisis, private savers have been keen to add
to their deposits, because of the risks inherent in the period of depres-
sion and reform. This has created an increasing burden on the bank
system, meaning banks must earn more prot from loans to meet the
repayment of deposits.
B. M onetary Policy and Growth of the Debt M ark et
The development of the debt market in the PRC is closely related

to reform in the nancial sector and implementation of monetary policy.
Apart from raising capital for the Government and covering government
decit, the debt market also supports the adoption of indirect instru-
ments of monetary policy.
The PBC began to act as the Central Bank of the PRC in 1984.
Until 1994, its main monetary policy tool was credit planning. The PBC
could loosen or tighten money supply, by increasing or decreasing the
scope of credit through the adjustment of the credit plan. When nan-
cial sector reform started in 1994, however, three signicant changes in
the nancial system impaired the e¹ciency of credit planning.
The rst and most important was the diversication of nancial
institutions. New commercial banks, nonbank nancial institutions such
as trust and investment corporations (TICs) and credit cooperatives sprang
up. Without a well-developed capital market, the TICs had no choice
but to conduct banking business. This meant that, apart from the origi-
nal four large state-owned banks, there were now many new participants.
As the credit plan was aimed only at the four SOBs, it could no longer
fully control the credit capital in the economy. The PBC could not
tighten credit by decreasing the credit plan because this encouraged the
other banks and TICs to increase credit.
The second change was the birth and growth of the nancial mar-
ket. After two periods of boom (19861988 and 19921993), the issuing
and trading market for T-bonds and corporation stocks developed rap-
idly. The interbank market has become the main channel of short-term fund
nancing among nancial institutions through repurchase agreements
(repo). Poor regulation of interbank and TIC practices, as well as a lack
of nancial contracts and transparent information, exacerbated the disor-
dered interbank borrowing, which was out of the PBCs control.
Meanwhile, specialized SOBs were converted into commercial banks.

Since direct control methods such as credit planning were not suitable
for the market-oriented activities of commercial banks, some indirect
methods, including oating interest rate and OMO, were introduced.
The Government realized that the establishment and growth of the
debt market were not only important for raising capital itself, but also
for implementing monetary policy because it: (i) supports the establish-
ment of a benchmark that would assist in the liberalization of interest
rates, an indispensable though indirect tool for channeling monetary
policy, and (ii) provides a new capital source to supersede the SOB
capital.
Unfortunately, since the PRC abolished credit planning at the be-
ginning of 1998, the conduct of monetary policy has been hampered by
the lack of e¬ective indirect monetary policy instruments. OMO requires
a mature debt market, and is the rst choice to replace credit quota
control, but it is still at the early stages of development. Meanwhile, the
interbank market remains small and illiquid. Finding an e¬ective chan-
nel for monetary policy is thus an urgent task for the Government.
The experiment with OMO in 1996 had little e¬ect because the
scale of short-term billswhich have always been less popular with the
Chinese than midterm onesis too small to a¬ect the demand and sup-
ply of public capital. The current scale is so small that the PBC can
easily buy all of them in the interbank bond market, which remains
underdeveloped. The traders in this market are commercial banks and
insurance companies, which are capital suppliers rather than demanders.
They hold onto bonds because of their high nominal interest, and trans-
actions are sparse. Without enough capital demand, OMO has little e¬ect.
Also, the nominal interest rate of deposits and loans is regulated by the
central Government and does not oat in response to the demand and
supply of public capital. Loans from banks are the main source for capi-
tal users in the PRC. OMO has no inuence on loan interest rates, meaning
it cannot a¬ect benchmark interest rates either.
Another problem is that the interest rates of short-term bills are
higher than for other kinds of bond. The current issue period is too
long, and frequent buying and selling means the Government is not able
to prepare for frequent, short-term bond issuance at present.
However, as the debt market grows, OMO will play a bigger role in
the implementation of monetary policy.
TA BL E 9
Interest R ates of Short-Term Bonds and Others (1994 and 1996)
Category Maturity Nominal Interest

(years) Rate (%)
1996
1st Issue Book-Entry T-Bond 1 12.10
2nd Issue Book-Entry T-Bond 6/12 10.53
3rd Issue Book-Entry T-Bond 3/12 9.90
4th Issue Book-Entry T-Bond 1 12.00
1st Issue Book T-Bill in Bearer Form 3 14.50
Voucher Form T-Bill 5 13.06
2nd Issue Book T-Bill in Bearer Form 3 10.96
Special Purchaser Bond 5 8.80
1994
6-Months T-Notes 6/12 9.80
1-Year T-Bill 1 11.98
Special Purchaser Bond 5 15.86
Source: Peoples Bank of China Quarterly Statistical Bulletin, 19993, Volume XV.
III. Overview of the Bond M ark et
Issuance of domestic debt was resumed in the early 1980s, shortly

after the launch of the Governments economic reform program. The
motivation was the need to nance the recently emerged decit.
In the PRC, there is only government debt, which is divided into
domestic and foreign. Domestic debts comprises Treasury bonds (T-bonds),
Enterprise bonds (E-bonds) and Financial Institutional bonds (F-bonds).
Only central Government can issue T-bonds. Only the Central Bank and
policy banks can issue F-bonds, and only strictly selected enterprises (almost
all are SOEs and guaranteed by local or central government) can issue E-
bonds. Other entities have no right to issue debt instruments publicly.
The turnover of debt always exceeds that of stockson average
by 190 percent. Debt issuance has become the most important means of
raising capital aside from bank loans, with amount of domestic capital
raised by debts compared with annual augmentation of loans having
increased to more than one-third of the amount from loans in 1997.
However, secondary market trading remains thin for various rea-
sons, including fragmentation of the market and the preference of the
countrys individual-dominated investor base to hold to maturity.
TA BL E 10
Debt and Bank L oans
(yuan billion)
Domestic Bank Ratio in

Year Debt Loan Percent
(A) (B) (A/B)
1993 61.72 633.54 9.74
1994 129.93 721.66 18.00
1995 181.17 933.98 19.40
1996 317.23 1068.33 29.69
1997 409.85 1071.25 38.26
Sources: China Securities and Futures Statistical Yearbook 1999; and China Statistical
Yearbook 1999.
A . Size and Potential of the Bond M ark et and Fiscal R isk
There is a strong relationship between the debt balance and amount

issued. The interest rate on deposits and loans was high for a long time,
with the interest rate of T-bonds about 2 percent higher than the former,
making repayment each year a heavy burden. From 1989 to 1993, T-
bonds were issued with an interest rate of 10 percent to 15.86 percent,
directly leading to the huge sum balance from 1993 to 1998.
To decide whether the volume of debt is suitable, it is necessary to
judge whether the economy can support larger debt issuance, and gauge
the size of the scal risk arising from such issuance. Liability ratio (ratio
of debts balance to GDP) and the ratio of debt balance to deposit bal-
ance is used to measure the former.
The liability ratio could be 12.66 percent in 1998 if the special T-
bond issued that year is taken into account. With the foreign borrowing
balance of Y146.04 billion (88.13 percent of this long-term debt), the
liability ratio would be 24.47 percent. However, compared with western
countries, this index is not high.
Meanwhile, the growth rate of savings in the PRC in recent years
has been startling. Until the end of 1996, most individual nancial as-
sets were saved in banks due to the lack of direct nance channels.
Bonds accounted for only 5 percent, leaving plenty of room for expansion.
The market is thus clearly able to support a larger bond volume.
The key problem is to e¹ciently transfer the huge sum of deposits in
the banking system into government capital channels.
The ratio of total debt issuance to scal expenditure increased from
6.82 percent to 30.66 percent from 1982 to 1998, while the ratio of
FIGUR E 5
Structure of R esidents Financial A ssets
Stocks
Bonds 4%
5%
Cash
14%
Insurance
1%
Deposits
76%
FIGUR E 6
R atio of Debt Issuance to Fiscal Expenditure
40
35
30
25
Percent
20
15
10
0
’82 ’83 ’84 ’85 ’86 ’87 ’88 ’89 ’90 ’91 ’92 ’93
Year
DIFE-a DIFE-b DIFE-c

domestic debt issuance to scal expenditure increased from 3.56 percent

to 35.27 percent. Usually, such a ratio in western countries is about 20
percent. As the PRC is a developing country and the demand for capital
is large, the above two indices are not incompatible. However, the fact
that central Government assumes all the costs of government debt can-
not be ignored, the ratio of domestic debt issuance to central government
expenditures must also be considered. This ratio reached 121.86 percent
in 1998, highlighting the central Governments scal risk burden.
This does not mean the domestic debt volume is too large, but
rather that the foundation of government income is too fragile. Taxes on
SOEsthe backbone of the tax basehave been cut, and subsidies to
them raised.
On top of this, recipient departments use the debt incurred without
the obligation of repayment. Since only central government can issue
bonds, all the costs of debts (including repayment and others), and thus
the scal risk, are concentrated on central Government, which itself never
invests.
If repayment is considered as pressure on the whole government,
the ratio of repayment to total scal revenue would be 23.78 percent in
1998.3 In contrast, in most western countries, the alarm-line ratio is no
higher than 10 percent. The ratio of repayment to central government
revenue was 48 percent in 1998. This implies that central government
revenue (not including debt incurred) is about two times as much as
repayment and proves that the PRC has had to borrow to repay.
The Governments aim, ideally, is to raise more and more construc-
tional capital by means of revenue from programs carried out on the
basis of debt-incurred capital. This has never been realized, however.
The main part of debt incurred is used as repayment, and the rest ows
to consumption items.
Since 1990, the ratio of repayment to debt incurred was very high,
showing the ine¹ciency of the channel supposed to carry the peoples
capital to the Government. Except for 35.48 percent in 1994, when the
3. Some economists suggest the scal revenue should be adjusted because

real government income includes budgetary, extrabudgetary and extrasupervision
income. Zhang Haijun calculated this gure as 5.4 percent, 6.7 percent, 9.8 per-
cent, 11.6 percent, 12.8 percent, and 14.5 percent from 1993 to 1998. His standard
is that adjusted total government revenue is about 20 percent of GDP. A little
adjustment may be necessary, but the degree is not too high because most in-
comes are controlled by local governments and their departments without any
central supervision. This income is not useful for supporting any greater issuance
of debt.
FIGUR E 7
R atio of R epaym ent to Dom estic Debt Incurred
140
120
100
80
Percent
60
40
20
0
’86 ’87 ’88 ’89 ’90 ’91 ’92 ’93 ’94 ’95 ’96 ’97 ’98
Year
amount of issuance erupted, this ratio has been continuously over 50

percent, and even reached 121.7 percent in 1990.
In a nutshell, increased government bond issuance following the
Asian crisis has put more pressure on government nance, with central
government scal risk growing sharply. Without strong scal revenue,
the Government has been forced to depend on continuous debt issuance,
resulting in a dilemma. To stimulate economic recovery, it has had to
issue more and more debt, but high issuing costs have made repayment
pressure a heavy burden.
The e¹cacy of bond policy has thus been discounted. While the
positive nancial policy supported by bond issuance can motivate some
investment from SOEs, it cannot a¬ect total aggregate demand level in a
continuous and e¹cient way. Without drawing out popular investment,
the multiplier e¬ect is discounted. After the crisis, SOBs had no incen-
tive to make loans because of the risk, while residents deposits were
owing into banks at high speed and the consumption market was forced
into depression.
B. Types of Securities
T-bonds are the most issued kind of security, followed by F-bonds

and then E-bonds. Treasury bonds include T-bills, Special National bonds,
Construction bonds, Fiscal bonds, and Ination-Proof bonds. Enterprise
bonds include Central Enterprise bonds, Short-Term paper and Local

Enterprise bonds. Financial Institutional bonds include Investment Fund
bonds, Trust Income securities, Certicate of Deposits, and Finance bonds.
State investment bonds were securities issued by six state-owned
investment companies to support state key construction. They had a
small issue amount, and were stopped in 1992. From an issuers view-
point, they are similar to State Construction bonds. MOF also issues
Special Purchase bonds, Price Index bonds, Special bonds, State Key
Construction bonds and Fiscal bonds.
State Key Construction bonds and State Construction bonds are
designed to raise funds for budgetary production projects. Fiscal bonds
are issued through private placement to specialized banks, comprehen-
sive banks, and other nancial institutions. Price index bonds are domestic
government bonds with index-linked principal subsidy. Special bonds
were issued to domestic entities with sound economic records, a pension
fund and unemployment insurance fund for state employees, although
issuance ceased in 1992. Since then, Special Purchase bonds have been
issued through private placement to pension funds and the unemploy-
ment funds of employees in domestic enterprises.
TA BL E 11
Issuing Sum m ary of Dom estic Securities by Category
(billion yuan)
State
Policy Other State Investment Total
Year T-Bond Financial Financial E-Bond Investment Company Domestic
Bond Bond Bond Bond Bond
1986 6.251 3.0 10.00 19.251

1987 11.687 6.0 3.00 3.00 23.687
1988 18.877 6.5 7.54 9.00 41.918
1989 22.391 6.1 7.53 2.25 38.236
1990 19.723 6.4 12.64 0.62 39.415
1991 28.125 6.7 24.99 9.5 0.23 69.541
1992 46.078 5.5 68.37 6.0 0.80 126.750
1993 38.131 23.58 61.715
1994 113.755 16.17 129.930
1995 151.086 30.08 181.170
1996 184.777 104.10 1.5 26.89 317.230
1997 241.179 139.95 3.2 25.52 409.850
1998 380.877 195.02 14.79 590.689

Every category of government debt is divided into di¬erent classes.

T-bonds, for instance, can be: (i) Book-Entry T-bonds, which are paperless
T-bonds with the advantages of a convenient process and lower cost and
risk, (ii) T-bonds in bearer form, which are paper face T-bonds and can
be traded through exchange or OTC, (iii) Voucher-Form T-bonds, which
are issued voucher as certicate for owner, and (iv) Specially Allocated
T-bonds, also known as Special Purchase bonds, which mainly target
pension funds and unemployment insurance funds.
A consequence of the lack of homogeneity in government debt issues
has been that, even after trading became permitted in many debt types,
markets have remained thin. The Government is aware of this, and there has
been a trend towards reducing their variety and placing more emphasis on
T-bonds. This will undoubtedly assist liquidity in the secondary market.
Treasury Bonds. T-bonds are designed to raise potential disposable

capital for state-directed construction to meet scal decit (especially
since 1994, when overdrawing from the Central Bank was prohibited by
law), and to adjust the macroeconomy. (This function has become im-
perative since the Asian crisis.) They are also used to improve the structure
of social capital to boost the development of nancial markets.
The main forms of T-bonds are the Voucher and Book-Entry T-
bonds. The uctuation of nominal interest rate of T-bonds stays in line
with that of SOB deposits, although always about 2 percent higher.
T-bonds are issued by the central Government and so have no risk.
Because of the scarcity of liquidity, investors (mostly individuals) con-
sider T-bonds as an alternative form of saving. However, the liquidity of
T-bonds is lower than that of deposits. Therefore, to remain attractive to
investors, a higher premium is necessary.
Historical Overview. From 1981 to 1985, the purpose of issuing debt

was to meet scal decit. The central Government was cautious, keeping
the volume of issuance under Y10 billion.
From 1986 to 1990, the issuance amount increased gradually, but
this could not meet the requirements of scal decit and debt repay-
ment, so the Government had to depend on foreign loans and overdrafts
from the Central Bank.
From 1991 to 1993, repayment of the debt itself became an impor-
tant factor in the expansion of the aggregate volume, and the high interest
rate of the T-bond exacerbated the repayment burden. From 1994 to
1997, the amount issued exceeded Y100 billion, and the growth rate
each year was startling, due to the 1994 Budget Law banning overdrafts
from the Central Bank, and the growing need to cover repayments.
TA BL E 12
Issuance History of T-Bonds
(billion yuan)
Year Repayment Amount Balance

Issued
1982 4.383
1983 4.158
1984 4.253
1985 6.061 23.70
1986 0.798 6.251 29.30
1987 2.318 6.307 39.10
1988 2.844 9.217 55.80
1989 1.930 5.607 76.90
1990 11.375 9.346 89.00
1991 15.669 19.930 117.02
1992 34.242 39.564 140.13
1993 224.300 314.780 1801.10
1994 36.496 102.857 247.78
1995 77.946 150.186 314.81
1996 122.317 184.777 380.35
1997 182.040 241.202 506.65
1998 234.860 380.870 735.13
By 1996, repayment was increasing by more than 50 percent each

year and the Government began to control the expansion of debt vol-
ume. As a result, the net augmentation (annual issuance value minus
decit redemption and repayment) of bond capital was only Y920 mil-
lion in 1997. Fluctuations in this index throughout the 1990s reect the
Governments policy dilemmas in the face of repayment pressure. In 1998,
the index stood at Y53.787 billion.
After the 1997 Asian crisis, the Government used T-bond issuance
as a tool to stimulate the depressed economy, issuing Y380.87 billion in
1998. Meanwhile, to enhance the capital adequacy ratio of the four
state-owned commercial banks, they were issued with Y270 billion of
Special T-bonds.
The ratio of the amount issued compared with GDP reached 4.8
percent in 1998, up from 0.83 percent in 1982, while the ratio of bal-
ance to GDP stood at 9.26 percent, up from 2.64 percent in 1985. T-bond
issuance has clearly become an important economic indicator and has
begun to inuence the investment behavior of the Chinese people and
enterprises.
FIGUR E 8
Net Augm entation of Bond Capital
600
500
400
300
100 Million RM
200
100
–100 ’85 ’86 ’87 ’88 ’89 ’90 ’91 ’92 ’93 ’94 ’95 ’96 ’98
–200
–300
Year
Debt M aturity. The bulk of the government debt is medium-term debt

(three to ve years). Issue periods are very long (sale of debt maybe take
several months), and it was not until 1994 that two short-term debts (six
months) were rst issued. These two paperless issues took more than one
week to place. Secondly, because of long intervals to redemption and
high rates of ination, people are reluctant to accept long-term debt.
This unevenness of maturity has caused the bunching of repayment,
increasing the pressure of payment of principal and interest and restrict-
ing investors diversied needs on maturity staple. In addition, such
short-term debt cannot satisfy the needs of long-term capital for infra-
structure projects.
The Government needs to address these problems by increasing
issuance of long-term debt, while maintaining the rapid growth trend of
short-term debt. Long-term bonds have already been increased, reaching
40 percent of the market in 1998, and 58 percent between January and
September in 1999.
Enterprise Bonds. The issuance of E-bonds has been on the decline

since 1993. Currently, both issuing volume and trading turnover are
very small compared with other securities.
There are three kinds of E-bonds: Local E-bonds, Central E-bonds,
and Enterprise Short-Term bonds. Strict procedures regulate the issue of
each kind of E-bond, and issuing them is so di¹cult that it has limited
the enthusiasm of enterprises.
TA BL E 13
Term Structure of T-Bonds
(billion yuan)
Year 01-Year 25-Year 610-Year Total

Amount % Amount % Amount % Amount %
19811984 3.532 20 14.128 80 17.600 100
19851987 23.999 100 23.999 100
19881990 60.991 100 60.991 100
19911993 109.490 100 109.490 100
19941996 846.51 19 315.953 71 45.03 10 445.634 100
19971999 573.700 64 325.80 36 899.500 100
Source: Peoples Bank of China Quarterly Statistical Bulletin 19994, Volume XVI.
FIGUR E 9
Scale of E-Bond Issuance
100
80
RMB Billion
60
40
20
0
’86 ’87 ’88 ’89 ’90 ’91 ’92 ’93 ’94 ’95 ’96 ’97 ’98
Year
Amount Issued Amount Repaid Balance
Central E-bonds were issued for the rst time in 1992. These have
been important since 1995, always accounting for more than 30 percent
of total E-bonds. Local E-bonds are issued without any termination, and
give local government, which does not have the power to issue T-bonds,
a means of raising capital for its enterprises. Enterprise Short-Term bonds
have been issued since 1988. These bonds are issued by an enterprise to
meet its seasonal short capital needs, with the principal and interest
paid back in three, six or nine months. These bonds accounted for 90
percent of E-bond issuance from 1993 to 1996, as short-term capital
needs were large while the economy was booming. However, since the
Asian crisis, this need has fallen. Internal bonds were issued from 1988
to 1992. House Construction bonds and Local Investment Company bonds
were issued as an experiment with a maturity of three years in 1992.
FIGUR E 10
Status of E-Bonds in the Overall Debt M ark et
(percentage of total bonds)
60
50
40
Percent
30
20
10
0
’86 ’87 ’88 ’89 ’90 ’91 ’92 ’93 ’94 ’95 ’96 ’97 ’98
Year
E-Bond Issuance E-Bond Balance
The main maturity of central and local E-bonds is three or ve

years, or occasionally eight, while for Enterprise Short-Term bonds it is
three to nine months.
The nominal interest rate of E-bonds may not exceed that of de-
posits by more than 40 percent. Among E-bonds of the same maturity,
no matter whether the credit rating is high or low, there is little di¬er-
ence in nominal interest rates. As the authorities select each enterprise,
and most of the credit risk is transferred to the Government, no company
has the incentive to make itself stand out from the others. Moreover, the
interest rate is decided not only by the enterprises manager, but by the
censoring and authorizing departments.
The cost of E-bond issuance is not high, compared with raising
money by stock issuance and bank loans. As E-bonds are midterm bonds,
the nominal interest rate is much lower than loans taken out in the same
year. However, the key problem is that not every qualied enterprise
gets the chance to issue, as issuance is restricted by complicated selection
procedures.
TA BL E 14
Explicit Cost of E-Bond Issuance
Items Cost
Issuance Fee 1.5 percent of the total raised capital
Credit Rating and Auditing Fee 0.1 percent of total raised capital
Nominal Interest Rate Cannot exceed nominal interest rate of
deposits by 40 percent
Another reason for the decline of E-bonds is that SOEs generally

have a high liability ratio, which makes issuance of debt with higher
interest rates seem unsuitable. Also, the Government has a policy of
guaranteeing T-bond issuance, which is very important for government
revenue, so it does not want E-bonds to interfere with T-bond issuance.
Historical Overview. From 1983, when E-bonds rst appeared, until

1987, the reason for issuance was to raise funds from employees and
society by enterprises. It was not regulated by law. From 1988 to 1992,
the Government encouraged SOEs to issue debt to supplement the scar-
city of working capital.
From 1993 to the present day, issuance levels have been low and
declining. The ratio of E-bond issuance to total domestic debt issuance
is now down to 2.5 percent, while the ratio of E-bond balance to total
bond balance is 4.36 percent. Compared with the same indices in 1992,
(57 percent and 36.57 percent, respectively), it is clear that E-bonds are
on the decline.
Unlike stocks, enterprises that issue E-bonds must repay the pre-
mium and principal in the future, and the high interest rate is a heavy
burden. Those which have the right to issue stocks need not worry about
repayment, however, so interest has switched to stock issuance. Cur-
rently, some loss-making local enterprises are refusing to repay the principal
and premium, and this has unsurprisingly killed o¬ investor interest in
E-bonds.
Financial Institution Bonds. With the reform of the economy, the capi-
tal market has begun to take shape. Since 1985, the Government has
allowed commercial banks and nonbank nancial institutions to issue F-
bonds. The aims were to: (i) raise mid- to long-term capital to recapitalize
SOBs; (ii) change the traditional capital structure; (iii) improve risk con-
trol in the banking system; and (for Policy F-bonds) (iv) form a capital
source to support policy banks.
TA BL E 15
Share of E-Bonds by Classication to Total
(percent)
Enterprise House Local

Central Local Short-term Internal Construction Investment Total
Year E-Bond E-Bond Bond Bond Bond Company Bond
1990 39.0 39.7 21.3 100

1992 10.8 37.8 33.4 16.3 0.9 0.6 100
1994 23.8 76.2 100
1997 32.0 37.1 25.8 100
Source: China Securities and Futures Statistical Yearbook, 1999.
F-bonds have been very important since 1996, and especially since
the Asian crisis, when people were reluctant to invest and consume.
Household deposits increased to more than Y6 billion, but the growth
rate of xed capital investments decreased continuously. This meant the
main task for the Government was to stimulate domestic investment and
consumption.
More and more F-bonds are now held by commercial banks and
other nonbank nancial institutions, increasing the capacity for future
OMO, as these institutions will be the means for currency creation. The
Central Bank can repurchase F-bonds in the interbank bond market to
control expansion or contraction of money supply.
F-bonds account for as much as 34.94 percent of total bond issu-
ance and 37.32 percent of total bond balance. In 1998, the amount of
F-bonds issued was about Y348.7 billion, almost equaling the amount of
T-bonds issued. Policy F-bonds are the most important type of F-bond
(98 percent in 1997) and play an important role in channeling huge
capital from commercial banks to policy banks, which carry out public
nance investments to support positive scal policy. In addition, Special
F-bonds are issued to meet liabilities arising from securities repurchase.
From 1985 to 1994, F-bonds were mainly issued by four large
specialized SOBs: the Industry and Commercial Bank of China, the
Agricultural Bank of China, the China Construction Bank, and the Bank
of China.
After the 1994 bank reform the National Development Bank (NDB),
Agricultural Development Bank, and China Import and Export Bank started
to issue Policy F-bonds to commercial banks and other nonbank nan-
cial institutions. This guarantees the political separation of commercial
banks from policy banks. Policy banks use the capital to support infra-
structure construction, develop basic and mainstay industries, release
bottlenecks and adjust industry and local economic structure. The main
issuer is NDB, which accounts for more than 95 percent of issuance
volume. Around 90 percent of NDBs loans are raised by bond issuance.
TA BL E 16
F-Bond Issuance
(billion yuan)
Year Total of Bond Policy Financial Other Financial Total of F-Bond

Bond Bond
Amount Balance Amount Balance Amount Balance Amount Balance
Issued Issued Issued Issued
1986 19.3 40.2 3 2.5 3.0 2.5

1987 23.7 56.3 6 5.5 6.0 5.5
1988 41.9 87.4 6.5 8.0 6.5 8.0
1989 38.2 112.9 6.1 7.1 6.1 7.1
1990 39.4 131.9 6.4 8.5 6.4 8.5
1991 69.5 175.4 6.7 11.8 6.7 11.8
1992 126.8 255.9 5.5 14.3 5.5 14.3
1993 61.7 275.9 10.9 10.9
1994 129.9 335.5 9.5 9.5
1995 181.2 430.1 161.3 9.5 170.8
1996 317.3 611.4 104.1 240.0 1.5 11.0 105.6 251
1997 409.9 972.5 140.0 348.7 3.2 14.2 143.2 362.9
1998 195.0 511.7
Note: Some sections of the table are incomplete due to di¹culty in nding access to
su¹cient statistics on F-Bonds.
Sources: China Statistical Yearbook 1999; China Securities and Futures Statistical Year-
book 1999.
Maturities of F-bonds are from one to ve years, with interest rates
a little higher than nominal savings interest rates. Principal and interest
are repaid only when F-bonds mature, so creditors cannot withdraw prin-
cipal before maturity. F-bonds can be traded in the interbank bond market,
although turnover is limited as investors tend to hold until maturity.
Following changes in the economic situation since 1989, the inter-
est rate has been adjusted continuously. F-bonds have therefore adopted
a oating coupon rate pegged to the deposit interest rate. The maturity
of F-bonds is partly determined by that of special loans. Generally, these
are required to match each other, and are normally one to ve years.
Trading of F-bonds is strictly supervised. It is prohibited to trade
them publicly, and they can only be repurchased in the interbank bond
market. The main participants are commercial banks and insurance com-
panies. The Central Banks OMO is another probable means of trading.
To improve the secondary market for F-bonds, the interbank bond

market should be opened to a wider range of investors.
C. Investor Base
The major target of debt sales are individuals, unlike in developed

debt markets, where the major investors are institutional investors, such
as banks, insurance companies, and mutual funds. At present, individual
investors hold more than 60 percent of government debt.
The Government is aware of the drawbacks of this investor struc-
ture. Firstly, an individuals main purpose in investing in T-bills is to
chase higher yields for prot, which increases the cost of raising capital
by issuing Treasury bills. Secondly, If individuals hold the bulk of T-
bonds, it is di¹cult for the central Government to carry out OMO, as
they always hold T-bills until maturity without exchanging them in the
secondary marketunlike institutional investors, who hold bonds for
liquid management.
The Government therefore adopted a series of measures to reduce
the proportion of debt issued to households. The Y13 billion paperless
issue in 1994 was the rst attempt to explicitly target wholesale buyers
(such as banks, insurance companies, etc.) via voluntary T-bill sales.
Prior to 1993, the PBC did not permit banks to hold government securi-
ties, except those issued on a mandatory basis. Likewise, pension funds
and insurance companies have only been encouraged to hold T-bills
since late 1993.
Although the PBC recognizes that control of monetary and credit
conditions via OMO as a policy tool requires the relevant institutions to
hold a stock of T-bills, as well as a liquid secondary market, the
Governments e¬orts have not had the desired results. Individuals still
maintain their position at the top of the investors league, and banks
have been forbidden to enter the stock exchange for T-bond trade since
1997, to prevent securities risks owing into the nancial system. The
proportion of institutional investors has therefore decreased once again.
However, it is inevitable that individuals will be superseded in the fu-
ture by other sources of funds.
IV. Bond M ark et Infrastructure
A . Issuance M ethods and Procedures
Methods used to issue government debt over the past two decades
have moved away from administrative placement towards more market-
oriented methods. However, aws in institutional arrangements have lim-

ited the degree of success.
In the rst phase (19811990), debt was placed administratively,
quotas were assigned by the MOF to the nancial department of the
provincial and local governments, and quotas distributed among pro-
duction units. These units in turn allocated bonds to workers and
employees, as an automatic deduction from wages. Although this was
compatible with circumstances at that time (high ination and lack of
investment consciousness, which curbed peoples desire to buy govern-
ment bonds), the hostile public response to semicompulsory placements
became intense. A new issuance method was badly needed.
In 1991, MOF adopted an underwriting syndicate to launch an
issue for the rst time on an experimental basis. The underwriters re-
ceived a commission of 0.15 percent of the underwritten amount, and
sold the bonds on a voluntary basis to clients. But in 1992, the volume
issued by underwriting syndicates was only a small proportion of total
issues, at 12.5 percent. Underwriting syndicates faced di¹culties in sell-
ing the desired quantity of government bonds due to competition from
the booming equity and property markets. The Government tried to im-
prove bond distribution by appointing 19 nancial institutions as primary
dealers (PDs) that were to be responsible for distributing government
bonds. After they fullled their placement, they would be entitled to
privileges such as priority in bringing equity o¬ers to market.
Auction methods have been introduced from 1995 to the present
day, a further step towards market-oriented placement. In 1995, seven
billion of the 10 billion T-bills were underwritten by 50 PDs, while the
rest were underwritten by the auction method. Afterwards the price auc-
tion method (Dutch auction) and yield auction method (American auction)
were adopted.
The form of placement has thus been a¬ected by temporal condi-
tions, such as high ination and lack of a market mechanism at the
beginning of the 1980s, which compelled the Government to adopt ad-
ministrative placement. Later, decreasing ination (that comparatively
increased the yield of government bonds) made them relatively attrac-
tive, so market-oriented methods were preferred. When ination became
serious again in 1993, and illegal fund-raising made issuance of govern-
ment debt di¹cult, the Government had to fall back on administrative
placement. There is clearly a long way to go before a fully market-
oriented issue method is achieved.
B. Issue Frequency
Primary issues for government debt have been o¬ered irregularly.

Without regular maintenance of a su¹cient volume of debt issue of any
maturity (but especially short-term), the development of a liquid second-
ary market is impossible. Consequently a short-term market yield curve
cannot be developed. Moreover, the absence of a liquid short-term mar-
ket presents problems for the use of indirect methods of monetary control.
In addition, since issues are made irregularly, there is little scope for
investors to synchronize the timing of sales with their cash ows. The
issue pattern puts considerable strain on investors liquidity and risk
management capabilities, limiting the number of investors and causing a
liquidity problemboth of which raise the Governments cost of funds.
Therefore, the Government has considered undertaking regular issues of
short-term debt to properly meet investors needs and help to form a
benchmark.
C. Secondary Trading System s
T-bonds. In the early 1980s, trading of T-bonds was banned and illegal
trading was rife. The ban was lifted in 1985, and the Government tried
to set up secondary T-bond market, with a discounting business intro-
duced as a rst step for the circulation of T-bonds.
The PBC formulated detailed rules on the discount T-bond busi-
ness, which stipulated that T-bond holders could transfer the bonds to
banks for discounted cash after holding them for two years. This is lim-
ited, however, because of high discounting rates.
The circulation system improved substantially in April 1988, with
the State Council approving cities with a sound base and experience in
nancial reform to develop OTC trading of T-bonds.
In 1992, trading accelerated with the setting up of the Shanghai
Stock Exchange (SSE) and several regional trading centers. More than
Y108.26 billion was traded that year. The ratio of trading volume of
bonds to stock was 84.4 percent. In 1993, T-bond derivatives were also
developed. Repo of T-bonds between securities rms and banks was al-
lowed in the interbank bond market. There was an experiment with T-bond
futures on the SSE, and some securities rms promoted T-bond trading
with T-bond portfolios.
By 1995, the ratio of turnover to balance had shot up to 1,799
percent. Most trading was not through the spot market but rather through
repo and T-bond futures. From 1996, T-bond trading turnover decreased,
however, as a result of the ceasing of T-bond futures trading.
TA BL E 17
Volum e of T-Bond Trading (19881998)
(billion yuan)
Turnover of Outstanding
T-Bonds T-Bonds Ratio in percent
Year (A) (B) (A/B)
1988 0.024 65.41 0.04
1989 0.021 68.71 0.03
1990 0.116 95.79 0.12
1991 0.370 106.07 0.35
1992 0.713 128.27 0.56
1993 8.717 154.07 5.66
1994 1,991.127 228.64 870.86
1995 5936.000 330.03 1,798.62
1996 1,803.778 436.14 413.58
1997 1,645.881 550.89 298.77
1998 2,160.079 725.69 297.66
At present, centralized trading is carried out on the Shenzhen and

SSE, while decentralized trading is carried out through trading centers
as well Securities Trading Automated Quotation (STAQ). This is mainly
in the interbank bond market.
E-bonds. There are two types of secondary market for E-bondsstock

exchange and OTC. In PRC, OTC covers 90 percent4 of E-bond trading,
while stock exchange trading is becoming smaller and smaller.
Securities institutions play a key role in OTC deals. Due to the
lack of a fully liquid and unied secondary market, dealers usually
monopolize the price. Regional segmentation is the main problem in the
OTC market.
SSE accounts for about 80 percent of E-bond exchange turnover.
However, the absolute quantity is still so small that this was only Y3.1
billion in 1998, compared with a massive Y2.2 trillion total T-bond
trading volume.
The yield of E-bonds in the exchanges is as high as T-bonds, but
unlike T-bonds, they cannot act as a tool for liquidity management.
Thus institutional investors, especially commercial banks, do not like
them very much.
4. Many papers quote 90 percent, but we have been unable to locate the
source in any public publication.
TA BL E 18
Trading in Shanghai and Shenzhen Stock Exchanges
Listed Number Turnover Deals

Year (million yuan) (lot)
Shanghai Shenzhen Shanghai Shenzhen Shanghai Shenzhen
1993 12 8 7.82 3.20 162
1994 18 6 1.83 0.03 65 3
1995 12 1 61.06 23.04 4,442 1,846
1996 5 1 116.12 29.97 11,032 2,368
1997 5 2 1,550.32 257.78 49,205 2,145
1998 5 2 3,131.84 936.73 304,452 6,047
Securities institutions play the biggest role in OTC deals, but it

has not been possible to turn around the inactivity of the bonds. OTC is
mainly made up of institutions buying and selling for themselves, i.e.
using their own capital to buy E-bonds from individuals at a xed price
decided by dealers and then selling the bonds to other individuals or
institutions at a higher price. Due to the lack of a fully liquid secondary
market and regional splitting, individuals have no power to negotiate
the price, which sharply decreases the liquidity of E-bonds.
There were no new listed E-bonds in 1998, largely due to govern-
ment neglect. Since T-Bond issuance has become such an important means
of raising revenue for the Government, it does not want E-bond issuance
to disturb T-bond issuance. Interest among enterprises has also tailed o¬
because of the high costs and complicated procedures of going public,
which also requires enterprises to pay many fees to securities exchanges
every year. Individuals, meanwhile, do not like the inconvenience of
trading in stock exchanges.
TA BL E 19
Trading Scale of E-Bonds
Listing Transaction Transaction

Year Numbers Turnover Volume
(million yuan) (lots)
1993 20 47,903.97 31,550
1994 24 223,318.50 330
1995 13 631,971.40 238,400
1996 6 1,838,962.00 282,420
1997 7 1,127,512.00 2,281,200
1998 7 4,068.57 7,763,910

Another impediment was that commercial banks were forbidden to

invest in stock exchanges, as part of the Governments e¬ort to cut o¬
their capital access to stocks to prevent securities risks from owing
into the nancial system. This removed nancial risk, but also reduced
an important capital source in E-bond trading. Turnover in 1998 ac-
counted for only 0.09 percent of total securities turnover.
FIGUR E 11
Transaction Turnover of E-Bonds
2,000,000
1,600,000
RMB Million
1,200,000
800,000
400,000
0
1993 1994 1995 1996 1997 1998
Year
Individual investors, no matter whether they have T-bonds or E-

bonds, prefer to hold on to them until maturity. Until such time as an
improved secondary market can distinguish between T-bonds and E-bonds
in terms of correct risk, yield, and liquidity, E-bonds will not play an
important role in the securities market.
F-bonds. F-bond trading is divided between transaction on exchanges

and transaction on the interbank bond market. The latter is divided into
repo between commercial banks, and repo or purchase through PBCs
OMO.
Very few F-bonds are listed on exchanges. In the past two years no
new F-bonds have been listed, while in 1993 there were just 21, in 1994
there were nine, in 1995 four, and in 1996 two. The Government intends
to concentrate F-bond trading in the interbank market.
Policy F-bonds cannot be formally traded, but the Central Bank
has approved monetary market repo of them.
TA BL E 20
Turnover of Bonds by Type
Share of Share of
Turnover Turnover of E-Bond E-Bond to
Turnover Turnover of Financial Total Turnover T-Bond on
Year of Stocks of T-Bonds E-Bonds Bonds (%) Turnover (%)
1993 366,702 8,717 47,904 4.47 423,328 11.32 549.54

1994 812,763 1,991,127 223,318 0.91 3,027,209 7.38 11.22
1995 403,647 5,936,000 631,971 0.03 6,971,618 9.06 10.65
1996 2,133,216 1,803,788 183,962 0.05 5,775,966 31.84 101.95
1997 3,072,184 1,645,881 1,127,512 5,845,577 19.29 68.51
1998 2,354,425 2,160,079 4,069 4,518,573 0.09 0.19
The turnover of F-bonds is small compared with the absolute issu-

ing amount. F-bond buyers are commercial banks and other nancial
institutions. Once these buyers hold F-bonds, they prefer not to sell, as
they can be a tool for capital management.
Currently, only Policy F-bonds and Special F-bonds have custodial
accounts at China Government Securities Depository Trust & Clearing Co., Ltd.
Interbank market participants are the Central Bank, commercial banks,
and insurance companies, which are mainly capital suppliers (until secu-
rities companies were allowed to join this market for the rst time in
1999), rather than demanders. Therefore, there are no capital demanders
wanting to sell their bonds. A wider range of participants is needed.
Without the participation of capital demanders, OMO through repurchas-
ing Policy F-bonds will have little e¬ect.
D. Interbank Bond M ark et
Currently, trading on the stock exchanges is withering. All govern-

ment bonds are moving towards trading in the interbank bond market.
E-bonds, however, have not been accepted by the interbank bond market.
The interbank bond market is an important vehicle for transferring
funds between the banking system and nonbank nancial institutions,
which include a variety of trust and investment corporations, nancial
companies, and securities dealers. The principal instruments on the in-
terbank bond market are short-term loans. However, a growing repo
agreement market has been emerging since 1992.
Although loans on the market are intended to be strictly short term
for liquidity management purposes, short-term loans have tended to be-
come long term through roll-over agreements, which the PBC periodically
attempts to curtail. Nonbank nancial institutions participation in the

interbank bond market was virtually unrestrained prior to 1993, and was
identied as the principal source of credit leakages responsible for the
overheating of the economy that year.
This access was subsequently cut o¬, and these institutions could
no longer get funds from the interbank market through repo or borrow-
ings from the banking sector to invest in securities markets (including
the bond and equity markets). This resulted in a fall in the bond and
equity markets in late 1993 and early 1994.
In 1999, some nonbank institutions were allowed to reenter the
interbank bond market, and repo was resumed at the same time. However,
restrictions have been imposed on both the volume of their borrowing
and the term of loans (now restricted to a maximum maturity of seven days).
Although the interbank bond market is designed to improve the ow of
funds between the banking system and securities rms, securities rms no
longer appear to depend heavily on the interbank system for resources.
E. Futures M ark et
The futures market, based on T-bond trade, was mainly focused on

the Shanghai (from 1993) and Shenzhen (from 1994) Stock Exchanges.
Trading of T-bond futures reached a peak in 1994. As a result of illogi-
cal future varieties, ignorance of risk control, and excessive speculation,
however, the Government decided to halt this trade in May 1995.
FIGUR E 12
Com parison of Spot, Futures, and R epo M ark ets
100
80
60
Percent
40
20
0
’93 ’94 ’95 ’96 ’97 ’98
Year
T-Bond Repurchase T-Bond Futures T-Bond Spot

T-bond futures failed for three reasons. Firstly, the scale of the spot
market was too small to support future operations. Speculative capital
could manipulate the future market, which increased risks for hedgers.
Secondly, there were aws in the design of trading rules, such as limita-
tions on price uctuations and holding volume. The third problem was
the ine¹ciency of government regulations due to imperfections in the
relevant laws and regulations.
F. R epo M ark et
A repurchase (repo) agreement is the sale of a security with a com-

mitment by the seller to buy the security back from the purchaser at a
special price at a designated future date. Basically, it is a collateralized
loan, where the collateral is a security (especially a T-bond).
In the early 1990s, T-Bond repo sprang up with the establishment
of the STAQ system. The rst repo trading of T-bonds was accomplished
between two STAQ members in July 1997. Wuhan Trading Center intro-
duced repo in T-bonds the following year. Later, Shanghai and Shenzhen
Stock Exchanges, as well as several regional trading centers, started repo
of T-bonds as T-bond issuance escalated and the T-bond spot market
ourished.
Although the practice of T-bond repo is a recent occurrence, turn-
over has grown rapidly. The number of institutions dealing with T-bond
repo reached more than 3,000 in 1994, and overall trading turnover
exceeded Y1,554 billion in 1998. The ratio of T-bond repo to overall
trading turnover has been higher than 50 percent since 1996.
TA BL E 21
Share of T-Bond R epo Turnover
(billion yuan)
Year Total T-Bond T-Bond Proportion

T-Bond Repo Spot of Repo (%)
1993 8,717 42 6,102 0.49
1994 1,991,127 7,578 46,837 0.38
1995 5,936,000 124,852 77,520 2.10
1996 1,803,788 1,300,864 502,924 72.12
1997 1,645,881 1,287,606 358,274 78.23
1998 2,160,079 1,554,084 605,995 71.95
The repo market is used by dealing rms (investment banking rms

and banks acting as dealers) to nance their positions and cover short
positions. It has evolved into one of the largest sectors of the money
market in developed countries. Financial and nonnancial rms partici-
pate in the market as both sellers and buyers, depending on their
circumstances. In the PRC, commercial banks are typical net buyers of
securities (i.e. net providers of funds), while securities rms and bank
trust departments are typically net buyers of fund (i.e. providers of secu-
rities). Securities rms and banks are thus the two basic categories of
participants in the repo market.
During the period of rapid growth of T-bond repo, some serious
problems were revealed. For instance, many disqualied institutions were
taking part in the repo market, and illegal speculation was prevalent.
There were other problems, too. Repo is basically a short-term
nancing tool (term range of T-bond repo in the US is from seven days
to one month). In the PRC, however, this could be several years because
of illegal operationscreating a dangerous debt-chain of overdue repos.
Repo therefore became a major source of nancial risk, and the Govern-
ment decreed that the term should not be longer than one year.
Another problem is repo collateral. In the PRC, repo turned into an
illegal means of nancing for securities rms. Dealers could raise funds
without delivering the securities to the customer, or hold the securities
in a segregated customer account, making the risk of default very high.
The Government therefore banned T-bond repo without su¹cient Trea-
sury securities as collateral, and it was made illegal for securities rms
to make repo arrangements by embezzling clients T-bonds. In 1995, a
central deposition and clearing company was set up to help overcome
the institutional drawbacks of T-bond repo.
Repo rates should be lower than the cost of bank nancing, which
is unsecured borrowing. In the PRC, however, the lack of availability of
collateral, long terms of repos, and poor quality of repo market partici-
pants made the repo rate very high. The main function of repo became a
means of nancing which bypassed the Governments credit plan control.
Another problem was the interrelationship between the repo and
stock markets. Stock index and repo rates move in opposite directions
in a developed market. In the PRC, however, the stock index and repo
rates move in the same pattern. This highlighted the fact that some
security rms were using repo to raise funds to invest in the stock mar-
ket. The unregulated repo market fostered speculation on the nancial
market, including the repo and stock markets.
G. Other A spects
1. Benchmark Yield Curve
Bond market issues should provide a benchmark rate, which can

help establish a yield curve. A yield curve has only been discernible
since 1996, when there were seven kinds of marketable Treasury bonds
on the two stock exchanges, from which a yield curve can be drawn.
However, the concept of a benchmark issue in relation to new debt
issue is not really operative in the PRC today due to: (i) the adminis-
tered setting of yields in the primary market (the operative benchmark is
the interest rate on bank deposits, which is administratively determined),
and (ii) poorly determined pricing of credit risk, making it hard to dis-
tinguish between the credit risk of a government-owned company and
actual government risk. With continuing economic reform, exibility in
rates and tough budgetary constraint of SOEs is needed, which would
make benchmark issues play a signicant role in future debt pricing.
Developing the role of benchmark issuance in dening the second-
ary market yield curve requires: (i) a greater range of maturites; (ii) a
more regular issuance calendar, and (iii) a greater focus on developing a
wholesale market to enhance liquidity.
FIGUR E 13
Y ield Curve A ccording to 1996 Data
12
10
8
Percent
0
0.0329 0.252 1.1644 2.189 2.6 6.8329 9.452
Year to Maturity
Source: Xiao Yu, China Treasury Bond MarketDevelopment, Comparison, and

Prospects, p. 161, Social Science Publishing House 1999.
2. Tax Treatment
All the interest from E-bonds incurs 20 percent income tax while
T-bonds are income tax-exempt.
3. Credit Rating
Credit rating started in the PRC in 1987. There are about 50 com-
panies, but most belong either to the Central Banks subo¹ces or to
SOBs, and only about 20 companies have independent legal qualica-
tions. Nowadays, only nine credit rating institutions can engage in the
rating of E-bonds. Most of them are local, except Dagong Global Co.,
Ltd. and Chenxin Securities Rating Co., Ltd. Most of the business is
carried out by the latter.
Credit rating is still very new, and investors do not care about it
very much. In fact, the ratings are only shown to the authorities which
make decisions on the procedure of issuance. In an environment of strict
state control, investors know that in reality the bonds are government-
guaranteed, because local governments generally have very close relations
with the issuing enterprises.
Normally, E-bonds that are issued publicly have the grade above A
or A-. While strict state-control reduces the risk of repayment failure, it
results in confusion between corporate and government credit.
Reform of SOEs is therefore crucial to help reduce this confusion
and to allow the E-bonds rated grade to reect the true standard of the
enterprises themselves.
The credit rating system of F-bonds must also be improved. There
is a very real di¬erence in levels of risk among commercial banks and
levels of e¹ciency of capital use in policy banks, but until recently
rating for these bonds was nonexistent. Without credit rating, all issuers
are treated the same, creating confusion between the real and govern-
ment risk. Without a pricing function in the F-bond market, the formation
of a ourishing F-bond market will be hampered.
4. Clearing and Settlement
It is important that the Government should develop a central regis-

tration and clearing company as a unied center for registering, depositing,
and clearing.
The China Government Securities Depository Trust & Clearing Co.,
Ltd. (CGSDTC), set up in 1994, is a good rst step. The CGSDTC is
currently building an OTC system of E-bonds. Since 1997, most E-bonds
have been issued in book-entry form. Once book-entry E-bonds are issued,
CGSDTC can build a united depository account for them. Investors, no
matter whether individuals or institutions, can use one depository account
to trade in di¬erent local markets, avoiding the waste of resources caused
by settlement and dealing in an uncombined OTC market. The state
wants to use the united depository, registration, and clearing system to
bring about united administration of E-bond dealing. It also wants to
integrate the uncombined local OTC into a united E-bond market, using
the interbank call money system to build the quotation system for E-bonds.
If the necessary conditions are fullled, authorities will allow E-
bonds deposited in CGSDTC to enter the interbank debt market. Therefore,
T-bonds, E-bonds and F-bonds can make up a united debt market using
the same infrastructure, which will nally bridge the gap between the
capital and monetary markets.
V. R egulatory Fram ework
The regulatory framework has improved greatly in the past decade.

However, the substantial role and sometimes ad hoc intervention of the
Government, especially in primary issues, remains a problem. There are
di¹culties of ensuring compliance and enforcement, and the regulatory
framework is fragmented.
Overall authority for the government bond market in the PRC rests
with a variety of agencies. Planning and preparation of the bonds is
done by the MOF, while the State Economy and Trade Committee shares
responsibility with the MOF for developing an issuance plan. The draft
plan is put to the Peoples Representative Congress every August for
approval. MOF also conducts regulation and supervision of the securi-
ties debt market, while the China Securities Regulatory Commission (CSRC)
is in charge of regulating the primary and secondary market. PBC has
some role in the government debt market, especially when debt policy
has a close relationship with monetary issues.
Regulation of debt securities is less dened and more fragmented
than supervision of the equity markets. According to current regulations,
MOF is in charge of the issuing of T-bonds, while the PBC is in charge
of overseeing the approval of bonds issued by nancial institutions and
the securities of investment funds. The State Planning Commission (SPC)
is in charge of the inspection and approval of state investment bonds
and bonds issued by state investment companies. The PBC and the SPC
share responsibility for the approval of Central Enterprise bonds, and
provincial or municipal governments are in charge of the approval of
Regional Enterprise bonds.
This division of responsibilities only covers regulations in the pri-

mary market, and reects the great variety of government debt issued in
the PRC. The PBC is responsible for bond trading activities only to the
extent that it approves securities trading centers. Monitoring activities
on the two o¹cially recognized stock exchanges is done by the CSRC.
Monitoring the actual trading of bonds, especially government bonds,
has been a gray area, however. The PBC, while responsible for the trad-
ing of government securities, has not had the capacity for regular
monitoring.
A . R egulations on E-Bond Issuance
The Regulation of E-Bond Administration, promulgated in 1993,

is the most important law relating to E-bond issuance, and requires that
E-bonds are issued by legally registered PRC citizens, and that the issue
is made public.
The SPC, PBC, MOF, and the CSRC decide the issue plan by coun-
tersignature annually, and then confer the planned scale to the relevant
provinces, which carry out the plan.
In addition, the decided nominal interest rate of E-bonds cannot
exceed that of deposits by more than 40 percent, and E-bond issuance
must not have any impact on T-bond issuance. E-bond issuance is ar-
ranged after the end of T-bond issuance.
The issued E-bonds must be underwritten by securities institutions.
E-bonds can be transferred, mortgaged, and inherited, but only by the
ratied securities institutions.
Moreover, the Companies Act of China requires that the accumu-
lated bond value may not exceed 40 percent of the companys net capital,
which must be more than Y30 million for a joint stock limited company
and more than Y60 million for a limited liability company.
The enterprises average disposable prot of the past three years
must be enough to meet one years repayment, and the purpose of the
raised capital must t the state policy of industrial adjustment.
A dm inistrative Authorizing Procedure. To gain issuance permission,

enterprises must get the recommendation of the local government and
the services in charge of the industry. The procedure is very compli-
cated, the quota very small, the issuing period very long, and the latent
transaction costs very high. If an enterprise does not have good rela-
tions with a whole string of governmental authorities, E-bond issuance
is nothing more than a dream. Potential issuers of E-bonds must gain
both quota and issuance approval:
Quota Approval. (i) Enterprises submit the projects that need capital
to the provincial supervisory agency, applying for formal ratication
and recommendation; (ii) the supervisory agency of one sector submits
the application to the Provincial Planning Commission and the Provin-
cial Sub-O¹ce of the PBC to declare for the quota; (iii) the Provincial
Planning Commission and the Provincial Sub-O¹ce of the PBC unify
the local plan for submission to the State Planning Commission and the
headquarters of the PBC, respectively; (iv) the State Planning Commis-
sion and the headquarters of the PBC integrate and censor each application.
They then compile the total plan of annual issuance to apply for ap-
proval at the State Council meeting, and (v) after the quota has been
approved, the State Planning Commission and the headquarters of the
PBC distribute the quota to the Provincial Planning Commission and
the Provincial Sub-O¹ce of the PBC. The quota then returns to the
supervisory agency of one sector and is distributed to the enterprises.
Issuance approval. If and when the quota is won, the enterprises must
get approval from the SPC and PBC (for central enterprises), and from
the Provincial Planning Commission and Provincial Sub-O¹ce of the
Central Bank (for local enterprises).
After getting the countersignature, the enterprises must submit the
application to CSRC. After CSRCs censorship, CSRC will make known
the issuance arrangements.
B. R egulations on F-Bond Issuance
Issuers must apply for approval from PBC, reporting their issuing
plans in detail. The Central Bank decides the scale of total F-bond issu-
ance, and allocates a quota to each issuer according to the credit capital
balance and demand for special loans. If provincial institutions plan to
issue F-bonds, they must report to the Provincial Sub-O¹ce of the Cen-
tral Bank.
The overall issuance plan is determined by the Central Bank, based
on credit programs and the amount of previous bond payment, so issuers
cannot break the approved quota. Issuing banks are required to report
their issue plans for each branch in detail and in good time.
After getting approval from the PBC, issuers advertise an issuing
notice, laying out in detail the objective of the issue, issue volume,
style, term, coupon rate, target, and place of sale. They then sign under-
writing contracts with their branches or underwriters.
Within a xed period, underwriters send raised capital to the issu-
ers, who pay an underwriting fee to the underwriters accounts. The raised
money must be used for repaying the bonds at maturity and making
particular loans. The amount of particular loans must be within the lim-
its of the approved quota.
There is clearly too much government regulation of issuance. The
fact that F-bond issuance has to coordinate with state macroeconomic
policy means issuers do not have much power to select date, interest
rate, etc.
Issuing and trading is conned to the Central Bank, commercial
banks, and other nonbank nancial institutions. This is too narrow and
insular, and has diminished interest among demanders and investors, making
trading inactive.
In addition, the PRC has an exacting nancial audit system. Any
organization that issues or buys bonds without PBC approval will be
ned or forced to stop business.
VI. M ajor Policy Issues and R ecom m endations
The current most important impediment to the development of the

bond market is the weakness of the scal system, exposed after the Asian
crisis.
Weak government nance is a key problem, and scal risk makes
it hard to increase bond issuance. This problem originates ultimately not
in debt issue, but the government income system.
A . R ecom m endations
Supervision of scal capital use should be improved to enhance

the e¹ciency of debt-incurred capital. To control scal risk, T-bond issu-
ance and repayment should be included in the annual government budget.
Enlarging F-bond issuance instead of T-bond issuance would also
reduce the large scal risk concentrated on central Government, making
it easier to gather capital from the banking system through policy banks.
It will be an important means of supporting the demands of current
positive scal policy, as well as preparing tools for future OMO, since F-
bond issuance will increase the amount of bonds held by commercial
banks and nonbank nancial institutions, feasibility of future money
creation will be enhanced.
OMO is urgently needed to ll the vacuum left by the nullication
of the excessive reserve requirement in 1999. Only OMO can substitute
this as a daily and seasonal economic tuning tool.
To liberalize interest rates, a benchmark must be built, potentially
by increasing long-term bond issuance. The PRC does not have a fu-
tures bond market to form this pricing mechanism, and there is no hedg-
ing power to balance the yield which will be used as benchmark. Thus
the futures bond market should be opened once again whenever possible.
The Government must impose cost-benet constraints on each user
of bond capital. The aggregate budget plan must be broken down into
the specic users revenue-cost audit. The NPC should have special power
to supervise this process.
M ark et Integration. Currently, trading on the stock exchanges is with-

ering. All government bonds are moving towards trading in the interbank
bond market. E-bonds, however, have not been accepted by the inter-
bank bond market.
CGSDTC should strengthen its functions for future unication of
the government bond market. Some T-bonds and nearly all E-bonds are
still not registered by it. This has resulted in regional E-bond separa-
tion. Moreover, CGSDTC currently does not have the power to supervise
trading in the interbank bond market, resulting in a long reaction inter-
val. A more rational arrangement is needed.
A unied bond market means a market open to all investors. How-
ever, current bond trading in the PRC has many qualication-based
limitations. Many would-be investors cannot go into the bond market
for capital management. The interbank market should be more open. If
the country is aiming for a unied bond market, the current entities in
the interbank bond market are too small, and the structure of investors
too simple.
The number of bonds listed on stock exchanges, no matter whether
T-bonds or F-bonds, should be decreased, with the eventual aim of abol-
ishing exchange trading. The Government should focus instead on the
development of the interbank bond market. Building too many markets
will hamper the realization of a unied bond market and waste resources.
STAQ should be updated to adapt to the needs of the future uni-
ed bond market. Current regional fragmentation is a result of
underdeveloped infrastructure, and STAQ and a system of settlement and
clearing are only a start. To develop a modern bond market, there is a
need for connected issuing, trading, settling, and clearing throughout
the nation. Such infrastructure will help overcome regional segmenta-
tion and enhance e¹cient resource allocation.
B. T-Bond M ark et
However, with rapid economic growth, there is potential for in-

creasing the scale of government debt. Infrastructure limitations of the
government debt market and drawbacks in the macroeconomic control

system will make it very di¹cult to realize this potential in the near future.
Illogical term structure and debt management makes government
debt very expensive as a means of raising capital. Meanwhile, the debt
service ratio is becoming higher and higher, leaving government debt
issuance locked into a vicious cycle where a large portion of debt rev-
enue is used to meet repayment for principal and interest. The relation
between bond issuance and government revenue and expenditure shows
very clear scal risks.
Fiscal risk exists mainly in central government nance, which un-
dertakes all of the costs of T-bond issuance, including issuance fees and
future repayment. Expensive issue costs force the central Government to
issue more debt, otherwise repayment would use up more than half of
the annual revenue. Annual expenditure is now less than debt issuing
volume, which has created a dilemma for the central Government. Its
policy performance is extremely dependent on debt issuance, but the
expense of debt issue drives an increase in issuance, which aggravates
this dependence still more.
The potential benets of the bond market are not being fully real-
ized. Most of what is issued is used for repayment, and without a liquid
secondary market many functions, such as OMO, benchmark, and risk
management cannot be fullled either. The current annual government
budget does not include the budget for bond issuance and repayment,
which leads the Government to issue debt without restraint, while the
NPC controls the annual scale of issuance without controlling the an-
nual balance scale. This makes the Government neglect the impact of
balance, which is the main source of repayment and an important tool to
promote bond augmentation.
The e¹ciency of capital arising from debt issuance is very low.
NPC and MOF control the total amount of annual issuance, but they do
not get involved in concrete procedures to supervise how it is used.
Limited participation of wholesale and institutional purchasers of bonds
leads to market distortion.
Interest R ates. Financial reform will allow a greater role for interest
rates as a pricing mechanism and the transformation of banks into enti-
ties which lend on the basis of risk evaluation and credit worthiness of
clients rather than credit plan and guanxi (relationship). State enter-
prises must also be reformed so that they face a real environment of
risks and return.
If the Government intends to undertake monetary control through
debt sales, it needs to be certain that the volume of debt sold will not
cause interest rates to rise to levels which might damage both the real
economy and the developing debt market. This is increasingly impor-
tant at present, with an escalating volume of Treasury debt outstanding.
Debt issues must therefore better coordinate with monetary policy and
spending requirements.
Supervision of Debt Capital Use. The NPC should step in to oversee

how debt capital is used, as well as changing annual issuance scale
control into balance scale control. The government budget should in-
clude revenue and repayment of government bonds. The practice of
dividing debt capital issuance and use should be eliminated. It would
be better to distribute a future repayments plan to local governments
based on the amount of capital they use. Furthermore, calling for bid
and price tendered should be introduced into the programs that use debt
capital.
Encourage Investors. The Government should reduce the variety of

issues and stabilize the issuing interval, and institutional investors should
be encouraged. To maintain the enthusiasm of individuals, various bond
funds should be built up as soon as possible.
The most important goal is not to augment T-bond issuance, but
rather to enhance the e¹ciency of bond issuance and capital use. Many
market-oriented issuance methods such as auction and a primary dealer
system should be encouraged.
The range of maturity should be extended to satisfy a wider range
of investors. Increasing the volume of short-term o¬erings would assist
nancial institutions with liquidity management, and encourage the de-
velopment of OMO by the Central Bank. Meanwhile, longer maturity
bonds would meet the growing need for nancial infrastructure invest-
ment by central and local governments.
Pricing. Liberalizing the interest rate in areas such as short-term and

money market rates would help to better dene a short-term yield curve
and thus help the pricing of new government debt issue.
The yield at issue should be related to current secondary market
yields rather than deposit rates. To form a benchmark, a futures market
is necessary for the price discovery mechanism of the bond market.
Secondary M ark et. Many of the problems in the secondary market

stem from underlying primary market problems, but could be alleviated
by spacing issues more evenly throughout the year on a preannounced
schedule, and primarily targeting wholesale investors.
Other measures in the secondary market include improving the regu-

latory framework. The present jurisdictional overlap between PBC and
CSRC must be addressed, and consolidated under a single authority.
The underlying laws need to be claried.
Clearing and Settlem ent. An e¹cient clearance, settlement, and de-

posit system should be established. The setting up of the Central Registration
and Clearance Company has been an encouraging step.
Central Bank Participation. The Central Bank should be encouraged to

act as a main participant of the bond market. MOF can commission the
Central Bank to underwrite short-term bonds, which it could hold. The tim-
ing of sales to banks and nonbank nancial institutions could be decided
according to the needs of monetary policy, thus ensuring that the Central
Bank has enough power to inuence money supply and benchmark rates.
As PRC liberalizes and reforms its nancial markets and monetary
policy instruments, it should rely more on OMO. However, this requires
the Central Bank to be able to buy and sell government bonds, so the
Central Bank should be encouraged to act as a main bond market participant.
C. E-Bond M ark et
In terms of developing the E-bond market, the Government should

consider the following:
Credit R ating. The Government should cultivate independent credit

rating institutions, with good reputations and prestige to cover most of
the E-bond credit rating business, and the close relationship between
SOEs and the Government should be broken down, making the enter-
prises more independent.
The Government can encourage enterprises with high ROA to issue
E-bonds, providing a good example for future issues. Investors interest
in E-bonds would be revived because of the higher quality. Meanwhile,
those enterprises with a high asset-liability ratio should be prevented
from issuing E-bonds. The Government should not regard E-bond issu-
ance as simply another means of supporting enterprises which lack current
capital. E-bond issuance can overcome this scarcity in a short time, but
it will hamper the bonds long-term development.
Sim plify Issuance. Changing the current complicated censoring and

approval procedures is very important. If conditions permit, this could
be carried out by intermediate social institutions.
Interest R ates. The Government should get rid of the limitation which
states that nominal interest rate of E-bonds may not be higher than that
of deposits by more than 40 percent. Under current low interest rate
conditions, E-bond issue will otherwise become uncompetitive. Higher
premia are needed to hedge higher credit risk.
R educe R egional Segm entation. The Government could soon allow E-

bonds to be traded in the inter-bank bond market. The central trust
depository and clearing system can be opened to E-bonds, helping to
unify the currently-split regional market and enhancing the activity of
the E-bond market.
D. F-Bond M ark et
Rebuilding the F-bond market is a means of connecting policy-

oriented and market-oriented practices. The market must therefore be
meticulously designed.
Credit R ating. The credit rating system of F-bonds must be improved.

There is a very real di¬erence in levels of risk among commercial banks
and levels of e¹ciency of capital use in policy banks, but until re-
cently, rating has only been applied to E-bonds. Without credit rating,
all issuers are treated the same, creating confusion between the real and
government risk. Without a pricing function in the F-bond market, the
formation of a ourishing F-bond market will be hampered.
W iden Investor Base. A wider range of investors should be introduced

to the primary market to invest in F-bonds. If investors are limited to the
PBC and commercial banks, F-Bond issuance will be dependent upon
the wishes of commercial banks and the capital source will remain nar-
row. Under current economic conditions, the primary F-bond market could
be opened to individuals, allowing them to discount F-bonds in divi-
sions of commercial banks. This would form a new and innite capital
source for the Governments positive scal policy.
Increase issuance. With current high scal risks, MOF should slow
down T-bond issuance and instead increase the issuance of F-bonds,
making policy banks the main issuers of government bonds. Positive
scal policy would thus be more dependent on F-Bond issuance, trans-
ferring some of the scal risk to policy banks. Moreover, it would decrease
issue costs through the simpler issue system between policy banks and
commercial banks.
Currently, there is only one means of F-bond general trading, through

repo. Although liquidity is not very important in the current environ-
ment of deation, spot trading between commercial banks would enhance
commercial banks desire for F-bond subscription.
Interest R ates. The nominal interest rate of F-bonds should be decided

not by the standard of deposit premium, but by the monthly weighted
average interest rate in the interbank money market. The repo yield of
F-bonds in the secondary market can be decided on the basis of real
interest rates.
R eferences
China Securities and Futures Statistical Yearbook 1999, Edited by CSRC, China
Finance and Economic Publishing House 1999.
The Peoples Bank of China Quarterly Statistical Bulletin, 1999-4, Volume XVI,
Edited by Statistical Department of the PBC.
Task Team of Comprehensive Department of State Development and Planning
Commission, An Empirical Analysis of Macro Income Distribution in the
1990s, Economic Research Journal, No. 11, November 1999.
Wu Jian, Savings, Investment and Economic Growth, Economic Research Jour-
nal, No.11, November 1999.
Xiao Yu, China Treasury Bond Market-Development, Comparison, and Pros-
pects, Social Science Publishing House, 1999, p.109.
China Statistical Yearbook 1999 and 1997, Compiled by National Bureau of Sta-
tistics, China Statistics Press 1999.
Prospects to Year 2000 of China Economy, Edited by National Information Cen-
ter, China Plan Press 1999.
Xiong Daowei and Zhang Qiaoyun, Government Security Market in China,
Southwestern University of Finance and Economic Press 1998.
Report About China Finance Development, Edited by the CPPS of SUFE, Shang-
hai University of Finance and Economic (SUFE) Press 1999.
China Finance Report, Edited by MOF, China Finance and Economic Publishing
House 1999.
Xiao Yu, China Treasury Bond MarketDevelopment, Comparison, and Pros-
pects, Social Science Publishing House 1999.
Y. Dong, Treasury Bond Market: Fiscal Policy and Economic Growth, Eco-
nomic Science Press 1999.
Chinas Mounting Debt, The Economist, May 29 to June 4, 1999, p.88.
Y. Dong, Treasury Bond, Financial Innovation, and Liberalization of Interest
Rate, Xuelin Publishing House 1998.
Frank J. Fabozzi, Bond Markets, Analysis, and Strategies, Third Edition, Prentice
Hall International Editions 1996.

5 People S Republic of China: Zhang Mingli and Hui Liu

Uploaded by

Copyright:

Available Formats

5 People S Republic of China: Zhang Mingli and Hui Liu

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

5 People S Republic of China: Zhang Mingli and Hui Liu

Uploaded by

Copyright:

Available Formats

5

Peoples R epublic of China

I. Fiscal Policy and M anagem ent

In the Peoples Republic of China (PRC), scal policy has three

(i) Channel social resources to the departments that the Government

Expenditure and revenue have increased steadily due to rapid eco-

is falling as a result of fast economic development and a governmental

Year Fiscal Revenue Fiscal Expenditure

A . Governm ent R evenue and Expenditure

Government revenue in PRC is classied as follows: (i) budget

Central Government Central Government

Source: China Statistical Yearbook 1999.

earned by SOEs. Statistics show that more than 60 percent of govern-

Fiscal Extra-budgetary Extra-supervision

Source: China Statistical Yearbook 1999.

A key problem is the e¹cient use of government expenses. Gross

B. Current Problem s and Future Task s

Compared with the healthy scal systems of modern developed

Current Prices Constant Prices Balance

A major problem in scal revenue management is the fact that

1985 20.05 20.41 0.044 0.51 0.146   0.280

Note: The gures do not include revenue from debt issuance.

The Central Banks seigniorage income (the accretion of surplus of

1. Income tax in 1996 was Y690.982 billion.

by the Central Bank itself, most is transferred to state-owned banks (SOBs).

D. Financing of Fiscal Decits

In terms of debt, the proportion of domestic debt is getting bigger

1985 20.04 11.28 4.08 1.91 1.71 1.06

Source: China Statistical Yearbook 1999.

Year Fiscal Deficit Domestic Debt Issued

Fiscal policy does not require the annual government budget to

II. M onetary Policy and M anagem ent

A . M onetary Policy Tools

The Peoples Bank of China (PBC), which began to act as the

Required Reserve Excessive Reserve

Note: The account of excessive reserve was abolished in 1998.

The annual planning of loans was abolished at the same time.

Demand Three- Six- One- Two- Three- Five-

21 Apr. 1991 1.80 3.24 5.40 7.56 7.92 8.28 9.00

Source: Peoples Bank of China Quarterly Bulletin 19994, Volume XVI.

Nominal interest rates before 1997 discriminated heavily against

SOEs Private Enterprises Collective Enterprise

Sources: Wu Jian, Savings, Investment and Economic Growth, Economic Research

2. Di¬erential between deposits and loans refers to balance of deposits

Since the Asian crisis, the PRC has strengthened management of

B. M onetary Policy and Growth of the Debt M ark et

The development of the debt market in the PRC is closely related

Meanwhile, specialized SOBs were converted into commercial banks.

Category Maturity Nominal Interest

III. Overview of the Bond M ark et

Issuance of domestic debt was resumed in the early 1980s, shortly

Domestic Bank Ratio in

A . Size and Potential of the Bond M ark et and Fiscal R isk

There is a strong relationship between the debt balance and amount

DIFE-a DIFE-b DIFE-c

domestic debt issuance to scal expenditure increased from 3.56 percent

3. Some economists suggest the scal revenue should be adjusted because

amount of issuance erupted, this ratio has been continuously over 50

Peoples R epublic of China

In the Peoples Republic of China (PRC), scal policy has three

Government revenue in PRC is classied as follows: (i) budget

Compared with the healthy scal systems of modern developed

A major problem in scal revenue management is the fact that

1985 20.05 20.41 0.044 0.51 0.146 0.280

Note: The gures do not include revenue from debt issuance.

The Central Banks seigniorage income (the accretion of surplus of

D. Financing of Fiscal Decits

The Peoples Bank of China (PBC), which began to act as the

Source: Peoples Bank of China Quarterly Bulletin 19994, Volume XVI.

domestic debt issuance to scal expenditure increased from 3.56 percent

3. Some economists suggest the scal revenue should be adjusted because

1986 6.251 3.0 10.00 19.251

Year 01-Year 25-Year 610-Year Total

The main maturity of central and local E-bonds is three or ve

Historical Overview. From 1983, when E-bonds rst appeared, until

1990 39.0 39.7 21.3 100

1986 19.3 40.2 3 2.5 3.0 2.5

oriented methods. However, aws in institutional arrangements have lim-

E-bonds. There are two types of secondary market for E-bondsstock

attempts to curtail. Nonbank nancial institutions participation in the

The repo market is used by dealing rms (investment banking rms

Source: Xiao Yu, China Treasury Bond MarketDevelopment, Comparison, and

Supervision of scal capital use should be improved to enhance