Chapter

11 Background Paper: A Reexamination of Indicators of Monetary and Financial Conditions

Author(s):
Tamim Bayoumi, Guy Meredith, and Bijan Aghevli
Published Date:
June 1998
Share
  • ShareShare
Show Summary Details
Author(s)
Gabrielle Lipworth and Guy Meredith 

To assess quantitatively the impact of changes in financial conditions on aggregate demand in recent years, this chapter discusses extensions to the staff’s past analysis of the monetary conditions index (MCI) for Japan. This work has been motivated in part by criticisms that the MCI—which focuses on the combined impact of real interest rates and the real exchange rate on aggregate demand—does not incorporate a sufficiently broad range of financial indicators: in particular, it omits the effects of changes in equity prices and fiscal policy. To include these factors, a broader measure of financial conditions has been constructed that combines the effects of changes in interest rates, exchange rates, equity prices, and fiscal variables. The staff refers to this construct as the financial conditions index (FCI).

The second section of this chapter examines how the MCI/FCI concepts can be used to assess the impact of financial shocks. The third section expands the MCI to include the impact on demand of equity prices and fiscal policy, deriving the FCI. The fourth section describes how the picture provided by the FCI differs from that of the original MCI in the 1980s and early 1990s, and discusses the implications of fiscal tightening in 1997 for financial conditions. The last section offers a conclusion.

What Is the MCI?

Before discussing empirical extensions to the staff’s MCI, this section addresses issues that have arisen in the past regarding the use and interpretation of this type of composite financial indicator.

The monetary conditions index, or MCI, was originally created at the Bank of Canada as a weighted sum of short-term interest rates and the exchange rate, where the weights reflected the estimated impact of these variables on aggregate demand.1 The concept has raised the following questions: How does an indicator of monetary conditions differ from one of monetary policies? What information can usefully be inferred from the index? And, how can such information effectively be used? While the MCI remains controversial, the following definition has been increasingly accepted: the net effect on aggregate demand of movements in financial variables that are affected both by monetary policy actions and also other market forces.2 In this sense, the MCI should be regarded as an indicator of financial conditions—as are real interest rates and real exchange rates separately—rather than as an indicator of the stance of monetary policy, or as an instrument or target of policies. In expanding the index to include equity prices and fiscal policies, the FCI is intended to indicate the net impact on aggregate demand of movements in a broader range of financial variables. Again, such variables can be affected both by financial policies as well as market forces.

While information about conditions in individual markets may be lost in the process of aggregating various indicators, the primary benefit of such composite indices is that the signal provided by the underlying variables is made clearer and more accessible. Thus, as in other cases in which information is combined, for example, in constructing real interest rates or effective exchange rates, the processing cost to users of filtering the information is reduced. Clearly, different users will apply different criteria in making choices about aggregation. In addition, the economic circumstances and the shocks the economy is experiencing will dictate the appropriateness of using aggregate indicators, and how they should be interpreted.3 Thus, the staff’s indicators are neither intended to be definitive nor intended to be used mechanically in all circumstances. Rather, they are regarded as potentially interesting and useful ways of aggregating information to obtain a clearer signal of the combined impact of financial variables on demand.

The following examples describe various shocks that an economy might experience, and illustrate how the information provided by the MCI/FCI can be interpreted.

  • Shock to risk premiums in exchange markets: Suppose that the risk premium on a currency increases. First, the exchange rate will depreciate, although the decline will probably be (at least partially) countered by higher domestic interest rates. The MCI would indicate the net effect of changes in exchange rates and interest rates on aggregate demand. If the MCI were to remain unchanged, then the estimated net impact on demand via interest rates and exchange rates would be roughly offsetting.

  • Change in monetary policy: Consider a tightening in monetary policy aimed at reducing aggregate demand—interest rates would rise and the exchange rate would typically appreciate, both of which would cause aggregate demand to contract. The MCI would indicate the overall effect on activity, which would be useful in judging the appropriate magnitude of the change in policy.

  • Autonomous demand shock: Suppose that demand falls autonomously; for instance, because firms decide that their existing capital stock is too high at current levels of interest rates. Activity would fall and there would be deflationary pressures, even if interest rates remained unchanged. In these circumstances, interest rates and the exchange rate would likely fall, indicating an easing in monetary conditions. The MCI would then indicate that the decline in demand and activity was autonomous, in the sense that it had not been caused by a tightening in financial conditions; it would also suggest the magnitude of the easing in monetary conditions that would be required to offset the original shock to demand.

  • Fiscal tightening: Suppose fiscal policy is tightened to achieve medium-term consolidation objectives. While aggregate demand would fall as a direct result of the tightening, part of the impact may be offset by lower interest rates and/or a depreciation in the exchange rate. The movement in the FCI would provide a measure of the net impact on demand, which could be useful in assessing the appropriate stance of monetary policy.

The examples presented above suggest various ways in which the MCI/FCI measures can be employed: (1) to assess the net impact of financial shocks, either policy induced or otherwise, on aggregate demand; (2) to infer the nature of shocks to the economy—that is, to ascertain whether shocks are due to autonomous shifts in spending relationships or to financial factors; and (3) to help in assessing the changes in policies that would offset the impact on demand of financial shocks.

Extensions to the MCI

The staff’s previous analysis of the MCI for Japan has focused on the net impact of real interest rates and the real exchange rate on aggregate demand. Estimation and model simulation results have yielded the following rule of thumb: a 1 percentage point rise in the real interest rate reduces aggregate demand by 1 percent over time, as does an appreciation of the real effective exchange rate by 10 percent.4 In this section, the authors extend the MCI to include the effect on aggregate demand of equity prices and fiscal policy in order to construct the FCI.

Why include equity prices? Japanese equity prices have experienced large fluctuations in recent years. In the second half of the 1980s, prices rose sharply, with the Nikkei 225 index increasing threefold between early 1986 and the end of 1989. Over the same period, the total market value of all shares traded on organized exchanges increased to 1.5 times GDP from 0.6 times GDP. These sharp gains were followed by the dramatic collapse of equity prices starting in early 1990—the Nikkei 225 index fell by nearly two-thirds by mid-1992, and the market capitalization dropped back to 0.6 times GDP. Equity prices then fluctuated sharply, before dropping in mid-1995 to their lowest level since the prebubble period, as the yen rose to record levels. In conjunction with the sharp reversal in the yen during the second half of 1995, equity prices staged a sharp reversal, rising by mid-1996 to their highest levels since 1991. These equity price movements are likely to have had an important direct impact on economic growth, both through the wealth effect on consumption and through changes in capital costs on investment. In addition, changes in equity prices may also have had indirect effects through their impact on bank profitability and financial sector balance sheets.

What role has fiscal policy played since the early 1980s? Following a long period of consolidation beginning in the late 1970s, fiscal policy has been used actively since 1992 to offset weakness in private demand. Countercyclical measures have included the adoption of five major fiscal stimuli packages and the staggered implementation of a tax reform package. The core of the packages has been substantial increases in public investment; direct spending on public investment has been boosted by a cumulative ¥28 trillion (5½ percent of GDP). In addition, the February 1994 package included a major temporary cut in income taxes, amounting to ¥6 trillion (1¼ percent of GDP). These measures, combined with the cyclical weakness in revenues, caused the general government balance (excluding social security) to deteriorate from approximate balance in 1990 to a deficit of 6½ percent of GDP in FY 1995. Of the overall deterioration, two-thirds reflects a decline in the structural balance, while the remainder is due to the normal impact of the cyclical downturn on revenues. Thus, fiscal policy has been actively used to manage demand during the downturn.

To examine the impact on aggregate demand of interest rates, exchange rates, equity prices, and fiscal policy, simulations were performed on the Japanese block of the MULTIMOD model, including a 1 percentage point rise in the real interest rate; a 10 percent change in the real exchange rate; a 10 percent change in the market value of the capital stock; a 1 percentage point increase in government spending as a share of GDP; and a 1 percentage point decline in taxes as a ratio of GDP. The impact of these shocks on real GDP is summarized in Table 11.1. 5

Table 11.1.Effects on Real GDP of Changes in Financial Variables(Percent deviation from baseline)
Year 1Year 2
One percentage point increase in real interest rates-0.9-1.2
Ten percent increase in real exchange rates-0.7-0.9
Ten percent increase in market value of the capital stock1.62.9
One percentage point increase in government s pending/GDP1.21.1
One percentage point increase in net taxes/GDP-0.6-0.5

The effects of shocks to interest rates and the exchange rate on aggregate demand by the second year of the simulation are broadly in line with the staff’s original rules of thumb.6 The effect of a change in equity prices on aggregate demand is evaluated by permanently raising the market value of the capital stock by 10 percent, resulting in a 2.9 percent increase in real GDP by the second year of the shock. Taking into consideration that the market value of the capital stock in MULTIMOD includes both debt and equity, a debt-to-equity ratio of slightly over 4:1 would imply that a 20 percent increase in equity prices would raise aggregate demand by about 1 percent.7 The MULTIMOD simulations yield the following results regarding the fiscal policy variables: a 1 percentage point increase in direct government spending as a share of GDP results in an increase in real GDP of about 1 percent in the second year; while a 1 percent increase in net taxes (i.e., taxes less transfers) as a ratio to GDP reduces GDP by about ½ of 1 percent in the second year.

From the MULTIMOD simulations, the following rule relating changes in interest rates, exchange rates, equity prices, and fiscal policy to aggregate demand was derived, where the effects of each variable are based broadly on the simulation results discussed above:

where R is the real interest rate measured in percentage points; REER is the logarithm of the real effective exchange rate (multiplied by 100) defined in terms of deviations from trend;8NIK is the logarithm of the Nikkei stock price index as a ratio to GDP (multiplied by 100); TAX is net taxes less transfers as a ratio to GDP; and G is direct government spending as a ratio to potential GDP. The weighted sum on the left-hand side can be interpreted as a financial conditions index, or FCI, expressed in terms of the percentage effect on aggregate demand of changes in interest rates, the real exchange rate, equity prices, taxes, and government spending.9

Interpreting Movements in Financial Conditions

The historical movements in the original MCI and the FCI are compared in the upper panel of Figure 11.1, while the lower panel shows the movements in the underlying components.10 The FCI provides a somewhat different picture of changes in financial conditions in recent years than does the MCI.

Figure 11.1.Indicators of Monetary and Financial Conditions1

(Index, 1980-95 average = 100)

Sources: Nikkei Telecom; and IMF staff estimates.

1 An increase indicates a tightening in monetary conditions.

The surge in equity prices in the second half of the 1980s—although partly offset by fiscal contraction—contributed to considerably easier financial conditions during the bubble period than the MCI indicated. Indeed, the MCI itself did not suggest that financial conditions were extraordinarily easy during the bubble, although the decline in the yen in 1989-90 did contribute to easier conditions in the latter stages. Looking at these indicators together suggests that unusually easy financial conditions did not play a major role in stimulating demand in the early stages of the bubble, as fiscal policy was generally contractionary (except in 1986), real interest rates were only modestly below levels earlier in the decade, and the yen rose sharply in 1986-87.11 The implication is that a more important role in the early stages of the bubble was played by other factors, possibly including financial liberalization and increasing investment following the sharp decline in world oil prices in 1985.12

The inclusion of equity prices also provides a different picture of the economic downturn: the drop in equity prices starting in early 1990 led to a sharp tightening in financial conditions as measured by the FCI, which was consistent with the subsequent contraction in aggregate demand. As the yen surged from mid-1992 to mid-1993, the MCI shows a sharp tightening in financial conditions. In contrast, the FCI suggests that the rise in the yen was largely offset by fiscal stimuli, as well as a modest recovery in equity prices.

The second surge in the yen in the first half of 1995 was accompanied by a sharp drop in equity prices. In addition, falling consumer prices led to a rise in real short-term interest rates. Together, these factors caused both the MCI and the FCI to tighten significantly by mid-1995. Starting in the second half of 1995, however, all of the components of the FCI moved in the direction of easier financial conditions: interest rates were cut to historic lows, the yen overvaluation was reversed, equity prices rose sharply, and stimulative fiscal actions were taken. As a result, both the MCI and the FCI show a sharp swing from relatively tight conditions to relatively easy conditions by early 1996, with the movement in the FCI being even sharper than that in the MCI.

Conclusions

What can be inferred from the MCI/FCI about developments in Japan in the late 1980s and early 1990s?

  • While both the asset-price bubble and the subsequent downturn do not appear to have been primarily policy induced, as suggested by the stability of real interest rates and the steady consolidation trend in the fiscal stance, changes in financial conditions owing to swings in equity prices reinforced underlying cyclical forces during 1986-92.

  • Reflecting the direct and indirect effects of yen appreciation, the MCI/FCI tightened sharply from late 1994 to mid-1995. Thus, despite stimulative monetary and fiscal policies, both the MCI and FCI indicated tight financial conditions by mid-1995.

  • By early 1996, financial conditions were easier than at any time since the beginning of the downturn, as the effects of the weakening of the yen and the drop in real interest rates since mid-1995 were reinforced by rising equity prices and the stimulative impact of the September 1995 fiscal package. Indeed, relative to the period of peak tightness in mid-1995, financial conditions were almost 6 percentage points easier in terms of aggregate demand.

Appendix
Multipliers of the EPA’s World Economic Model(Percent deviation from baseline)
Year 1Year 2
One percentage point reduction in official discount rate-0.3-1.1
Ten percent depreciation in yen against major currencies-0.6-1.6
One percent increase in real public investment1.21.4
Source: Ko Kawasaki, “Development of the ERI Compact Model,” Discussion Paper 64 (Tokyo: Economic Research Institute; Economic Planning Agency), 1996. Table 11.4.
Source: Ko Kawasaki, “Development of the ERI Compact Model,” Discussion Paper 64 (Tokyo: Economic Research Institute; Economic Planning Agency), 1996. Table 11.4.
References

    BaumgartnerUlrich and GuyMeredith1995Saving Behavior and the Asset Price “Bubble” in Japan: Analytical Studies IMF Occasional Paper 124 (Washington: International Monetary Fund).

    FreedmanCharles1994“The Use of Indicators and of the Monetary Conditions Index in Canada,” in Frameworks for Monetary Stability: Policy Issues and Country Experiencesed. by Tomás J.T.Baliño and CarloCottarelli (Washington: International Monetary Fund) pp. 45876.

    International Monetary Fund1996World Economic Outlook (WashingtonInternational Monetary Fund).

    KawasakiKemichi1996“Development of the ERI Compact Model,”Discussion Paper 64 (Tokyo: Economic Research Institute; Economic Planning Agency).

    MassonPaulStevenSymanskyRichardHaas and MichaelDooley1988“MULTIMOD: A Multi-Region Econometric Model” Staff Studies for the World Economic Outlook (Washington: International Monetary Fund) pp. 50104.

    MassonPaulStevenSymanskyGuyMeredith1990“MULTIMOD Mark II: A Revised and Extended Model” IMF Occasional Paper 71 (Washington: International Monetary Fund).

    MeredithGuy1992“Discretionary Monetary Policy Versus Rules: The Japanese Experience During 1986-91,”IMF Working Paper 92/63 (Washington: International Monetary Fund).

Note: This chapter was initially published in 1996 as part of Japan—Selected Issues, IMF Staff Country Report 96/114 (Washington: International Monetary Fund).

Freedman (1994) outlines the derivation and use of the MCI in the Canadian context.

The MCI is being used as a tool of analysis, for instance in the May 1996 World Economic Outlook.

These caveats are not, of course, unique to the MCI/FCI. For instance, information contained in real interest rates should be interpreted with care, as different measures of real interest rates matter in different circumstances, and changes in nominal interest rates will not necessarily have the same effect on activity (or reflect the same shocks) as changes in expected inflation.

See, for instance, Meredith (1992).

In performing the MULTIMOD simulations, the GNP deflator, the short-term nominal interest rate, and the nominal exchange rate were held exogenous to identify the partial effects of each shock.

As MULTIMOD is a forward-looking model, the effect of the shocks depends to some extent on their expected duration. The shocks described here were maintained for two years. In the event, results were similar to those obtained using time-series estimation of the effects of these variables on future activity, as discussed below.

This debt-to-equity ratio is typical of prebubble values in Japan, after cross-shareholdings of firms are excluded from the outstanding value of equity. A debt-to-equity ratio of just over 4:1 would mean that a 10 percent increase in the market value of the capital stock would result from a rise in equity prices of slightly over 50 percent. The simulation results then imply that a 10 percent rise in equity prices would result in a ½ of 1 percent increase in real GDP; a 20 percent increase in equity prices, in turn, would generate a rise in GDP of 1 percent.

The upward trend in the real exchange rate used here, based on unit labor costs in manufacturing, is estimated at about 2 percent a year.

The appendix table shows the effects on GDP of changes in interest rates, the exchange rate, and government spending generated by the Japanese Economic Planning Agency’s World Economic Model. By the second year of the simulation, they are broadly similar to the MULTIMOD results, at least in terms of relative magnitude, although the individual effects are typically somewhat larger in absolute size than in MULTIMOD.

For presentational purposes, these indices have been normalized such that a 10-unit movement in the index implies a 1 percent change in aggregate demand; they have then been rebased such that their average level during 1980-95 equals 100.

Of course, even though monetary and fiscal policies may not have played a major role in generating the bubble via their impact on easier financial conditions, in retrospect, it is generally agreed that tighter policies would have been appropriate to offset autonomous forces of demand.

The factors that may have contributed to the bubble are analyzed in Chapters VI and VII of Baumgartner and Meredith (1995).

    Other Resources Citing This Publication