Czech Republic: Selected Background Studies
Author: International Monetary Fund
Publication Date:
14 Sep 1995
eISBN:
9781451809992
Language:
English
Keywords:
ISCR; CR; rate; VAT; wage; growth target; bank portfolio; volume of CNB; banks resident; CNB's operation; Monetary base; Open market operations; Capital inflows; Sterilization; Demand for money
Download PDF (920.1 KB)

This paper examines a few selected aspects of adjustment and restructuring in the Czech enterprise sector. It examines financial performance of the Czech enterprise sector, drawing on data reported by nonfinancial enterprises to the Czech Statistical Office (CSO) in 1993. The findings reveal that significant progress has been made in the nonfinancial aspects of adjustment, but a large part of the enterprise sector has been slow to adopt measures to achieve sustainable financial viability.

Abstract

This paper examines a few selected aspects of adjustment and restructuring in the Czech enterprise sector. It examines financial performance of the Czech enterprise sector, drawing on data reported by nonfinancial enterprises to the Czech Statistical Office (CSO) in 1993. The findings reveal that significant progress has been made in the nonfinancial aspects of adjustment, but a large part of the enterprise sector has been slow to adopt measures to achieve sustainable financial viability.

IV. Challenges to Monetary Policy 1/

1. Introduction

With no budgetary recourse to central bank credit and minimal quasi-fiscal operations, the Czech National Bank (CNB) has been relatively free to concentrate on its central task of providing a stable macroeconomic environment in the transition to a market economy. Monetary policy has been largely successful: after adjusting for the impact of the introduction of the VAT, underlying inflation has been held down to around 10 percent for the past three years, while foreign exchange reserves have grown dramatically, even in the face of major economic shocks, including the collapse of vital trade links with the CMEA and the split of the Czech and Slovak Federation. At the same time, direct controls over credit and interest rates have given way to flexible prices, market-based instruments of monetary control, and the development of a supervisory framework within which some 59 domestic and foreign banks now compete.

Most recently, however, strains have emerged, two of which merit special mention. The first has been the task of maintaining monetary control in the face of large capital inflows, which have entailed difficult decisions about monetary sterilization, interest rate policy, and the need to guard against prudential risks to banks. The sheer size of the inflows has severely burdened monetary policy. In 1994, for example, the capital account surplus was virtually equal to the starting stock of base money. A second problem is the effect on the financial system of rising bad loans, which are now estimated at over one third of total bank credit.

2. Targets and instruments of monetary policy

a. Nominal anchors and intermediate targets

Since the former Czech and Slovak Federal Republic’s successful stabilization program in 1991, the fixed exchange rate has served as a basic nominal anchor for monetary policy in the Czech Republic. While the exchange rate anchor was invaluable in stabilizing expectations in the wake of big-bang price liberalization, over time the implied endogeneity of the money supply became problematical as mounting balance of payments surpluses raised money growth and threatened the disinflation effort. Rather than abandon the peg, the CNB began emphasizing broad money (M2) as an independent target (i.e., not merely as an indicative figure consistent with expected balance of payments and fiscal developments)--thus, in effect, establishing it as a second nominal anchor in addition to the fixed exchange rate. Ever since, the CNB has been struggling with the inherent tension in fixing two policy anchors, with the result that it has had only limited success in achieving its announced targets for M2.

In the uncertain environment of economic restructuring, the targeting of M2 has been fraught with difficulties. Structural upheavals in the economy--the split of the federation, the emergence of an imperfectly accounted services sector, and new statistical methods--render the data series too short and unreliable for proper estimation of money demand. In practice, M2 targets have been set based on a simple velocity equation, i.e., by adding up targeted inflation and real GDP growth and then making an allowance for lower velocity. Indeed, M2 velocity has been on a declining trend for the last four years, but even a one standard error band around the velocity trend line is quite wide, roughly equivalent to 4 percent of broad money (Chart 4.1). 1/

CHART 4.1
CHART 4.1

CZECH REPUBLIC: VELOCITY OF MONEY AND INTEREST RATES

Citation: IMF Staff Country Reports 1995, 085; 10.5089/9781451809992.002.A004

Source: Czech National Bank.1/ Prague interbank offered rate.

Partly because of this uncertainty about underlying money demand, the CNB has tended to accommodate missed monetary targets. Thus, both in 1993 and 1994, the original M2 target was revised upwards at mid-year in the face of heavier-than-expected capital inflows, while the final outturn proved to be higher still (Table 4.1). In turn, the following year’s targets were built on actual outcomes, thus giving rise to some further “base drift.”

Table 4.1.

M2 Growth Targets and Outcomes

(Percent change)

article image
Source: Czech National Bank.

Adjusted for special factors at end-year (see section 3 below).

Aside from the uncertainty of money demand, three other factors have constrained monetary policy: (i) problems in filtering out the “noise” in monetary data arising from large swings in bank float (see below); (ii) the emergence of offsetting capital inflows following open market operations to mop up liquidity, and (iii) the possibility of sterilization costs forcing the central bank into a loss-making position, which could ultimately compromise its independence.

b. Operating targets and instruments

Up until July 1994, the CNB used the monetary base (currency plus banks’ reserves) as the operating target in its quest to reach quarterly intermediate targets for M2. Once a week, the CNB’s operating desk combined net sales of CNB bills, government securities, and auctioned refinance credits to achieve the desired level of base money. From time to time, changes in reserve requirements were also used, as were adjustments to the discount and Lombard rates (the latter, however, are mostly of symbolic/signaling value, since the volume of CNB rediscounting is relatively small). After an initial period of volatility in 1993, the money multiplier has been on a generally stable, if declining, trend--thus, permitting the CNB to make more informed judgements about the desired path of reserve money (Chart 4.2). 1/

CHART 4.2
CHART 4.2

CZECH REPUBLIC: MONEY MULTIPLIER

Citation: IMF Staff Country Reports 1995, 085; 10.5089/9781451809992.002.A004

Source: Czech National Bank.1/ M2A/average base money. M2A is defined as M2 plus interbank clearing account float, which eliminates the effects of swings in the latter.

Starting in August 1994, the CNB switched to total bank reserves as its operating target and began daily--instead of weekly--intervention in the money market, principally through reverse purchase operations and outright sales of CNB and government securities. Daily open market operations are carried out with a high degree of sophistication, and have been supplemented with other control techniques when necessary (such as the transfer of National Property Fund (NPF) deposits from commercial banks into the central bank). The operating procedure was modified at the start of 1995 when the CNB was confronted with two problems. First, the reserves target resulted in significantly higher money market rates that began attracting a high volume of short-term capital inflows. 2/ Second, temporary volatility in the level and composition of deposits at the end of the year made it difficult to forecast the demand for bank reserves. To mitigate the first problem, the operating desk began working with an explicit upper limit on interbank interest rates, which was to override the reserves objective. To address the second problem, the CNB switched to targeting a relatively small amount of excess reserves, the idea being that the system should run with minimum slack. In practice, when the excess reserves objective began to conflict with the interest rate objective, the latter nearly always won out. Since April 1995, the CNB has resumed using total bank reserves as the principal operating target--though the upper bound on the interbank interest rate has been retained.

3. Monetary developments in 1994 and early 1995

After adjusting for special and transitory factors, broad money grew by about 21 percent in 1994, well ahead of the original and mid-year revised targets (Chart 4.3). 1/ The bulk of this expansion was concentrated in time and savings deposits, so that the narrower aggregate Ml increased by a more moderate 18 percent. Credit to enterprises and households grew by about 16 percent, with the bulk of the expansion consisting of short- and medium-term loans to the business sector. The “true” pace of credit expansion was higher than suggested by the above figure, since direct enterprise borrowing from banks resident abroad--a substitute for domestic bank credit--also increased markedly (see Chapter V).

CHART 4.3
CHART 4.3

CZECH REPUBLIC: GROWTH OF MONEY AND CREDIT

(In Percent)

Citation: IMF Staff Country Reports 1995, 085; 10.5089/9781451809992.002.A004

Source: Czech National Bank.1/ Adjusted for swings in customer float (i.e. deposits that, pending clearance, are temporarily reclassified in other items net), and other special factors noted in section 3 of the text.

Monetary expansion was particularly brisk during the first half of 1994, and seasonally adjusted M2 velocity fell well below the band around the longer-term trend. Although monthly inflation was rising only moderately at that point, the CNB nevertheless decided that the underlying momentum in money growth was excessive, and that it was necessary to pull back somewhat, even if to a less ambitious rate of M2 growth than originally targeted. Accordingly, the stance of monetary policy was progressively tightened during the second half of the year. An increase in the minimum reserve requirement was announced in July (raising the average reserve requirement by about 1 percentage point). This was followed by increases in the discount and Lombard rates in October, each of which was raised by 0.5 percentage points to 8.5 percent and 11 percent, respectively. In November, agreement was reached with the NPF to transfer to the CNB the bulk of its deposits with commercial banks (Kc 16 billion, equivalent to roughly 10 percent of reserve money, was moved to the CNB during the period December 1994-February 1995). 2/ The upshot of these measures was to significantly tighten liquidity conditions during the second half of the year, as reflected in rising money market and bank loan rates. Correspondingly, the growth of seasonally adjusted M2 fell from around 25 percent per annum in the first half of the year to an annualized rate of about 18 percent in the second half.

The tightening of monetary policy was not without problems or complications. Starting in late 1994, the CNB began observing that the widening of the differential between domestic and foreign interest rates (Chart 4.4) was prompting large short-term capital inflows especially via foreign banks’ branches. 1/ However, it was not until the first quarter of 1995 that short-term inflows began approaching levels that might be described as a “flood.” For the first time, banks became the major conduit for capital inflows: net foreign borrowing by resident banks was close to US$1 billion in the first quarter of 1995, and accounted for more than half of total reserve accumulation of the central bank during that period (Table 4.2). Aside from the sheer scale of the inflow, the authorities were concerned by the fact that most of the inflow consisted of short-term--and hence potentially volatile--borrowing by foreign banks’ branches. Although the period under consideration is short and the data limited, preliminary estimates suggest that during the period September 1994-April 1995, up to two thirds of CNB sterilization operations may have been “offset” by such short-term capital flows responding to local liquidity conditions (see section 4 below).

CHART 4.4
CHART 4.4

CZECH REPUBLIC: SELECTED INTEREST RATES

(In Percent)

Citation: IMF Staff Country Reports 1995, 085; 10.5089/9781451809992.002.A004

Source: Czech National Bank.1/ Difference between one-month PRIBOR and a weighted average of one-month interbank rates in U.S. dollars and Deutsche marks.
Table 4.2.

Commercial Banks’ Foreign Exchange Position, 1994-95 Q1

article image
Source: Data provided by the Czech National Bank.

In this rather difficult environment, broad money growth began to pick up. After adjusting for special factors, the seasonally adjusted growth of M2 during the first quarter of 1995 was close to 24 percent--well ahead of the 14-17 percent target range for the year as a whole. Interest rates softened from the peak levels observed in December, and estimated velocity declined towards the bottom of the velocity band. Credit expansion remained at around 16 percent (annual rate) through end-March, but picked up markedly in the following two months.

With broader macroeconomic indicators--inflation and surveys of expected inflation, the trade balance, and consumption demand--pointing to a greater risk of overheating, the CNB announced in mid-June a series of measures aimed at reducing the force of the short-term offset to sterilization operations and curtailing the growth in money and credit. These included: (i) a 1 percentage point increase in the discount and Lombard rates (to 9.5 and 12 percent, respectively); (ii) an increase of about 1 percent in the average reserve requirement via a unified rate of 8.5 percent, effective August 1 (previously, the requirement was 3 percent on time deposits and 12 percent on sight deposits); and (iii) the introduction of a limit on short-term foreign borrowing by banks, also effective August 1, requiring each bank to ensure that its net short-term liabilities to nonresidents, in all currencies, do not exceed the smaller of 30 percent of claims on nonresidents or Kc 500 million. The limit applies to on-balance-sheet assets and liabilities, as only these affect the monetary base. The following section sheds some additional light on the motivations for instituting capital controls, which arguably is one of the most significant and controversial steps taken by the CNB in recent years.

4. The offset to monetary sterilization

Until mid-1994, the CNB was largely successful in sterilizing inflows; such hesitation as existed with regard to sterilization activities was based, apart from uncertainty over money demand, on only a generalized recognition of the limits of such policies over time. However, starting in the last quarter of 1994, the limits to sterilization became less and less abstract as the CNB’s operations desk began noticing that its efforts to tighten liquidity conditions were unravelling at a rapid rate. For example, during the first quarter of 1995, the CNB found that of the Kc 14 billion increase in the stock of prime securities (Treasury, NPF, and CNB bills) sold to sterilize external inflows, roughly Kc 12 billion had been acquired by nonresidents and foreign banks’ branches.

Despite the many empirical limitations--notably, the short period of observation and the lack of data on key variables--more formal estimates of the “offset coefficient” are presented in this section. The exercise is best considered work in progress and the results preliminary at this stage.

The standard procedure for calculating the offset coefficient involves estimating equation:

KAP=a1ΔNDA+a2CAB+a3ΔR*+a4ΔY+ui(1)

where KAP, NDA, CAB, R*, and Y are, respectively, the capital account balance, net domestic assets, the current account, the foreign interest rate, and nominal output; ui is the usual error term and the prefix Δ denotes the change in a variable. This equation is a reduced form derived from substituting demand functions for (non-money) domestic and foreign assets into an inverted money demand equation, with the ais’s being arithmetic combinations of various income and interest elasticities of asset demand functions. By adding CAB to both sides, noting that KAP + CAB = ΔNFA (net foreign assets), and setting a2 = a’2+1, the equation can also be expressed as:

ΔNFA=a1ΔNDA+a′2CAB+a3ΔR*+a4ΔY+ui(2)

When a1 is “high” (i.e., close to -1), the central bank’s domestic money operations are fully undone by capital flows in the reverse direction; monetary sterilization is both pointless and expensive for the central bank. When a1 is “less high,” e.g., -0.5, sterilization is feasible, but still relatively expensive (because mopping up of Kc 100 of external inflows requires Kc 200 of open market operations).

To estimate equations (1) or (2), fortnightly data--coinciding with the cycle for banks’ minimum required reserves--were used; where data at this frequency was not available, proxies were utilized. 1/ Moreover, given the difficulties in implementing the underlying economic model at a high frequency and over a short time period, alternative specifications and lag structures were experimented with.

Six alternative specifications, calculated using ordinary least squares (OLS), are reported in Table 4.3. 2/ The change in net domestic assets is, in all cases, a significant variable explaining movements in the central bank’s net foreign assets. Summing over the lags of ΔNDA gives a rough estimate of the offset coefficient, which averages around -0.65 across all the equations. Alternatively, if one adds up only the significant lagged coefficients, the offset averages -0.40 across the six specifications (see last two columns of Table 4.3). Reassuringly, this range of estimates for the offset coefficient is broadly in line with those calculated for other countries facing capital inflow surges in the late 1980s and early 1990s (see e.g., Schadler, et al. (1993)). The main implication of the above is that sterilization remains a feasible, if expensive, option for the central bank.

Table 4.3.

Summary of Regression Analysis of Offset Coefficient 1/

(Based on data for the peried September 1, 1994 - May 10, 1995)

article image
Source: Data provided by the Czech National Bank.

All regressions are estimated by the ordinary least squares (OLS) method. In equations 3 and 4, the trend component of the change in NFA was removed (by taking the residuals of a regression on a constant and time trend). The numbers in parentheses are t-statistics.

Sum of statistically significant (at least 15 percent confidence) A, B, and C.

KAP is change in net foreign assets less the trade balance, enterprise credits, and net clearing payments to Slovakia.

significant at 1 percent level.

significant at 5 percent level.

significant at 10 percent level.

significant at 15 percent level.

5. The bad loan problem

For the past several years, the problem of “bad loans”--i.e., credits not regularly serviced and hence potentially unrecoverable--has been a major drag on the development of the Czech banking sector. 3/ The recorded volume of bad loans rose in 1994 from about 23 to 38 percent of outstanding credit. Although some of this rise may be due to further degradation of bank portfolios, the bulk--up to two thirds--probably reflects the introduction of tighter accounting norms in September 1994. The new loan classification system utilizes both objective and subjective criteria: 1/

  • Standard loans: timely servicing and information;

  • Watch loans: payments or information on borrower overdue by 30-90 days or loan restructured during the past 6-36 months;

  • Substandard loans: payments or information on borrower overdue by 90-180 days or loans restructuring more than 6 months ago;

  • Doubtful loans: payments or information on borrower overdue by 180-360 days or full repayment assessed to be improbable;

  • Loss loans: payments or information on borrower overdue by more than 360 days or full repayment assessed to be improbable.

Banks are expected to create reserves on a quarterly basis to cover a risk-weighted average of the above loans, less applicable collateral (the risk weights for the categories noted above are, respectively, zero, 5, 20, 50, and 100 percent). It is still too early to judge whether most banks were indeed fully provisioned by end-1994 as per norms, because auditors’ reports on the subject had not been completed for most banks as of mid-June. For the largest banks, the level of provisions is generally thought to be adequate; even so, provisioning in the banking system as a whole continues in 1995 at about the same rate observed the year before.

The bad loans problem has not, so far, posed systemic risks to the banking system. Since a spate of incidents in early 1994, when three small banks were shut down, there have been no major crises. Nevertheless, the condition of several smaller, privately-owned banks, remains fragile--prompting not only heightened surveillance by the CNB, but also an informal moratorium on the licensing of new banks; the only exception to the latter in the past two years has been the licensing of the state-owned Czech Export Bank in May 1995.

The impact of bad loans, however, extends well beyond the question of banking system stability, and is closely related to the macroeconomic problem of capital inflows. The high incidence of bad loans and associated provisioning has been a major factor pushing up intermediation costs in the Czech economy. As such, it has encouraged newer and more creditworthy enterprises to seek financing abroad. These links are taken up further in Chapter V.

Reference

Schadler, Susan, Maria Carkovic, Adam Bennett and Robert Kahn, Recent Experiences with Surges in Capital Inflows, Occasional Paper 108 (Washington: International Monetary Fund, December 1993).

  • Search Google Scholar
  • Export Citation
1/

Prepared by Ranjit Teja.

1/

Although the velocity trend line is an imperfect guide to predicting money demand, deviations from the trend line provide some useful insights into liquidity conditions and interest rate movements. For example, periods when seasonally adjusted velocity is close to or below the bottom of the band coincide with generally lower interest rates in the money market, and vice versa.

1/

The fall in the multiplier was essentially driven by a sustained rise in the currency deposit ratio, reflecting a return to more normal levels after the trauma of the currency split in 1993; however, since mid-1994, the multiplier has been flat.

2/

Short-term flows had not, until late 1994, been a major force in the capital inflows phenomenon.

1/

To eliminate the distortion of underlying trends, it is necessary to adjust for swings in customer float, which consist of deposits that--pending clearance--get temporarily reclassified in “other items net” as nondeposit liabilities to clients. A second important adjustment involves correcting for a large operation by a foreign bank at the end of 1994 that temporarily inflated both sides of its balance sheet. These adjustments are incorporated in Charts 4.1-4.3.

2/

A further Kc 6 billion was transferred in May 1995.

1/

Some of these inflows were initially related to tax arbitrage resulting from tax treaties that eliminated the withholding tax on NPF bills for nonresidents. This difference allowed nonresidents to make substantial returns as the bills approached maturity. While a tax play may serve as an initial impulse, it inevitably involves the repatriation of proceeds at completion, and by itself cannot be a source of continued net inflows.

1/

The current account was proxied by the monthly trade balance--converted to a fortnightly basis using polynomial smoothing--plus net clearing account payments to the Slovak Republic; in addition, direct financial credits from abroad--which are mostly long term and require prior CNB authorization--were also added as a means of further isolating the component of the capital account capable of quickly reacting to money market conditions. Output was proxied using monthly data on household income (converted to a fortnightly frequency as above).

2/

OLS estimates of the offset coefficient are known to be biased due to a “simultaneity problem” arising from central bank sterilization. To illustrate, suppose an increase in output raises money demand and draws capital from abroad, and that the central bank tries to sterilize the inflow by lowering NDA. Although the ΔNFA is due to higher output, OLS attributes some of the rise in NFA to lower NDA, thus exaggerating the size of the offset coefficient. Unfortunately, it was not possible to satisfactorily implement an instrumental variables procedure to overcome this problem.

3/

For a description of the origins and handling of early problems, see Czech Republic--1994 Recent Economic Developments (SM/94/194), Annex III. In addition, some further analysis on the sectoral composition of bad loans can be found in Chapter I.

1/

These criteria are applied by borrower and not by loans--so that even if one loan of a particular borrower enters the problem loan categories, all loans to that borrower are classified in the same way.

Same Series

Other IMF Content

Other Publishers

Cover IMF Staff Country Reports
Czech Republic: Selected Background Studies
Author: International Monetary Fund
Volume/Issue:
Volume 1995: Issue 085
Publisher:
International Monetary Fund
ISBN:
9781451809992
ISSN:
1934-7685
Pages:
124
DOI:
https://doi.org/10.5089/9781451809992.002

CZECH REPUBLIC: VELOCITY OF MONEY AND INTEREST RATES

CZECH REPUBLIC: MONEY MULTIPLIER

CZECH REPUBLIC: GROWTH OF MONEY AND CREDIT

(In Percent)

CZECH REPUBLIC: SELECTED INTEREST RATES

(In Percent)