Selected Issues

Abstract

Selected Issues

Thailand: Bringing Inflation Back to Target1

A. Context

1. Thailand’s headline inflation was below target in 2015–16. The low headline inflation reflected to a large extent external drivers, in particular the sharp decline in oil prices in recent years. Yet, core inflation also declined and is projected to remain subdued.2 The causes of weak inflation dynamics warrant closer investigation.

A01ufig1

Headline and Core CPI Inflation

(Y/y percent change)

Citation: IMF Staff Country Reports 2017, 137; 10.5089/9781484301845.002.A001

Sources: CEIC Data Co. Ltd.; and IMF staff calculations.

2. Monetary policy rates have been kept on hold since early 2015. In early 2015, the Bank of Thailand (BOT) cut the policy rate by 50 basis points to a near historical low of 1.5 percent, citing the need to foster the recovery and maintain well-anchored inflation expectations.3 The BOT interpreted the cause of negative inflation in 2015 to be largely attributable to supply side factors, which would subsidy as base effects from oil gradually disappeared. The BOT has kept the policy rate constant since April 2015 to preserve room for maneuver amid volatile global conditions and to safeguard financial stability in a low-interest-rate environment. Several other central banks in the region also facing below-target inflation rates eased monetary policy during 2016, though they still have higher policy rates than Thailand.

A01ufig2

Selected Countries: Growth and Inflation, 2016

(Y/y percent change)

Citation: IMF Staff Country Reports 2017, 137; 10.5089/9781484301845.002.A001

Sources: IMF World Economic Outlook; and IMF staff calculations.

Main Drivers of Inflation

3. The optimal policy response to a protracted period of low inflation depends on the causes of disinflation. This paper estimates a hybrid New Keynesian Phillips Curve with time varying parameters to gauge the quantitative role of (long-term) inflation trends, economic slack (output gap) and import price inflation in shaping inflation dynamics:4

πt=θtπte+(1θt)πt1MA4+αt(yty*)+μtπtIM+ɛt,(1)

where:

4. The analysis reveals some important changes in Thailand inflation dynamics. The impact of lower import prices was a major factor behind the decline in headline inflation in 2015, with low oil prices being the largest contributor to inflation dynamics. Yet, the estimation also reveals a weaker role of the forward-looking “expectations channel.” Particularly, the forward-looking parameter θt, but also the level of long-term inflation trend πte, have been on a declining trend over the last few years.5 This led to a lower degree of forward-looking inflation dynamics and made Thai headline inflation more vulnerable to the oil price decline and the weaker cyclical position, contributing to a protracted period of very low inflation rates.

Thailand: Monitoring Medium- to Long-Term Inflation Expectations

Long-term inflation expectations are a crucial input in modern monetary policymaking. The consistency of private sector’s inflation expectations at medium-to-long horizons with the central bank’s target provides a direct assessment of the credibility of monetary policy. Moreover, in an environment of very low inflation, stable long-term inflation expectations are crucial to bring inflation back to target and avoid entrenched low inflation.

Surveys of inflation expectations and expectations extracted from financial instruments are standard indicators monitored by many central banks. Surveys are a traditional source of information on long-term expectations and are available several times per year for many countries. With the issuance of inflation-linked bonds (ILBs) in many advanced but also emerging economies, the so-called “break-even inflation rate” (BEIR)―the yield spread between comparable conventional bonds and ILBs―has also become a key indicator of inflation expectations. BEIRs provide more timely information on investors’ expectations than survey-based expectations. Yet, in addition to the expected inflation, BEIRs may incorporate other factors, notably inflation risk and liquidity risk premia. Hence, BEIRs are better interpreted as the overall inflation compensation requested by investors to hold nominal assets, rather than a pure measure of expected inflation.

For Thailand, both surveys and BEIRs can be used to monitor inflation expectations. The figure below reports the recent developments in survey measures (6 years to 10 years ahead from Consensus Economics) together with the BEIRs from the two ILBs issued in Thailand (with maturities 2021 and 2028, reflecting average expectations over the next 5 years and 12 years, respectively). As for many other countries experiencing sharp disinflation, the diverging patterns of survey and market-based inflation expectations is striking. Survey measures have barely changed and continue hovering above the mid-point inflation target of 2.5 percent. In contrast, BEIRs have declined significantly since 2013 and remain well below target. Although a higher liquidity premium in ILBs (compared to conventional Thai government bonds), and possibly a lower inflation risk premium, likely played a role in the decline of the BEIRs, they cannot account for the full difference between the two measures, thereby casting some doubts on the level of long-term inflation expectations measured by surveys.

The estimation of long-term inflation trends using econometric models has become increasingly common in major central banks since the GFC. The long-term inflation trend can be interpreted as the level of inflation that would prevail once temporary influences unwind. The rationale behind recent research efforts to estimate trend inflation is twofold. First, given the forward-looking orientation of modern monetary policymaking, policy decisions should be based on reliable indicators of long-term inflation expectations. While survey and financial indicators provide useful information, both have important shortcomings, which may have rendered them less reliable in periods of persistently low inflation, as low inflation periods have been less frequent than high inflation periods. Second, discrepancies among indicators call for regular assessment of their information content, and the estimation of trend inflation can be instrumental in this respect.1

A01ufig5

Inflation Expectations

(In percent)

Citation: IMF Staff Country Reports 2017, 137; 10.5089/9781484301845.002.A001

Sources: Consensus Forecasts; and IMF staff calculations.
A01ufig6

Inflation Expectations and Target

(In percent)

Citation: IMF Staff Country Reports 2017, 137; 10.5089/9781484301845.002.A001

Sources: Consensus Forecast; and IMF staff calculations.

Empirical estimates show that the (long-term) inflation trend in Thailand has decreased below the mid-point inflation target since early 2015. Moreover, trend estimates are significantly below survey measures of long-term inflation expectations. This evidence suggests there are some risks of de-anchoring of inflation expectations following the protracted period of below-target inflation. Looking ahead, these results warrant close monitoring, also given the increasing dispersion among survey point estimates for long-term expectations. The main section of this paper uses these trend inflation estimates in the econometric analysis.

1 Trend inflation estimates are from Garcia and Poon (forthcoming), based on the approach in Chan et. al. (2015).

B. Policy Response

Monetary Policy Stance

5. Nominal policy rates are low by historical standards but real interest rates have risen in recent years. Staff estimates of the natural real interest rate suggest it has been low, between 0.1 percent and 0.2 percent, in 2015–16.6 In contrast, various measures of effective real rates have ranged above that level. For instance, one measure of the effective real rate, calculated as the difference between the policy rate and actual inflation over the last two years, was in the range of 2.4 percent to 1.3 percent. Similarly, the policy rate minus expected inflation over 2016 (i.e., expected inflation at the beginning and at the end of 2016) was in the range of 0.3 percent to 1.3 percent. Assuming policy rates on hold, and considering the latest average inflation expectations by the BOT (1.2 percent) or Fund staff (1.0 percent), (ex-ante) real rates would remain above the estimated natural rates in 2017.

6. Model simulations suggest that in a passive scenario inflation would remain at or below the lower end of the tolerance band. Simulations are based on a quadratic central bank policy reaction function that considers gaps for inflation and for output, as well as interest rate smoothing, within a stylized new Keynesian model calibrated for the Thai economy.7 Under a passive scenario (red line in Figure 1 below), without monetary or fiscal stimulus, inflation and inflation expectations would remain subdued, with substantial risks of inflation falling back outside the tolerance band, once base effects from higher oil prices fully unwind. Keeping rates on hold over the forecast period would only partially counteract weak inflationary dynamics. Moreover, the small but negative output gap would persist and the cost of further adverse shocks would be large.

Figure 1.
Figure 1.

Thailand: Monetary Policy Simulations: Active and Passive Scenarios

Citation: IMF Staff Country Reports 2017, 137; 10.5089/9781484301845.002.A001

Source: IMF Staff calculations.Note: Optimal monetary policy path based on a standard quadratic loss function.

7. Monetary policy easing, within a broader expansionary policy mix, should contribute to bringing inflation back to target. Results suggest that monetary easing, in which the policy rate is gradually cut over several quarters, combined with fiscal stimulus, can steer inflation back to target. In an active scenario (black line in panel chart below) that combines monetary easing and fiscal stimulus, inflation would return to the 2.5 percent mid-point target (with a slight overshooting) and the output gap would close. Furthermore, given the protracted period of below-target inflation, readiness to act and communicate in a decisive manner would be fundamental to prevent a de-anchoring of long-term inflation expectations (Alichi et al., 2015; and WEO, 2016).

Low Interest Rates and Financial Stability Risks

8. Financial stability concerns do not need to constrain monetary policy in the current environment. Financial stability risks remain contained. Monetary policy actions aimed at increasing inflation from its current low level will also have positive effects on financial stability. First, higher inflation will lower the real burden of highly indebted economic agents, particularly lower-income households. Second, restoring inflation may prevent a de-anchoring of inflation expectations, which, in turn, would lead to higher (long-term) real rates and debt service. Finally, the sooner the inflation rate returns to target, the lower the risk of low inflation becoming entrenched and the faster the exit from the low-interest-rate environment.

9. Further monetary policy stimulus and tighter regulatory and prudential measures can provide a better policy mix. Macroprudential tools designed to target specific pockets of fragility can be used if the temporarily lower interest rates trigger any risk to financial stability. As part of the ongoing strengthening of the financial stability framework, a broadening of the macroprudential toolkit is, in any case, beneficial. For example, loan-to-value ratios can be decreased to tame house price increases and regulations changed to close loopholes for regulatory arbitrage.8

The Role of Communication and Enhanced Transparency

10. The BOT has successfully enhanced the transparency of monetary policy over the last decade. For example, Thailand’s score in the Dincer-Eichengreen index of central bank transparency rose from 6 in 2000 to 10 by 2014. Yet, by the same metric, there is still some scope for further improvement towards the maximum index score of 15.

11. Higher central bank transparency is associated with strengthened forward-looking inflation dynamics, and therefore helps avoid low-inflation traps. Using estimation results from equation (1) for each ASEAN–5 country,9 Dany and Garcia (forthcoming) find a strong and positive relationship between the Dincer-Eichengreen index of central bank transparency and forward-looking inflation dynamics.

A01ufig7

ASEAN-5: Central Bank Transparency and Forward Looking Inflation Dynamics

(Coefficient of persistence)

Citation: IMF Staff Country Reports 2017, 137; 10.5089/9781484301845.002.A001

Sources: Dincer and Eichengreen (2014); and IMF staff calculations.
A01ufig8

MPC Votes in Favor and Against Policy Rate Changes

(Number of votes)

Citation: IMF Staff Country Reports 2017, 137; 10.5089/9781484301845.002.A001

Sources: Bank of Thailand; and Barclays Research.

12. The BOT’s communication is crucial to manage inflation expectations and dismiss the perception of constrained monetary policy. Further guidance in terms of envisaged actions to achieve the target over a given time horizon is fundamental to avoid a de-anchoring of long-term inflation expectations after a protracted period of below-target inflation. Assessments of past performance, discussion of the shocks impairing the achievement of the inflation target, and elaboration on the internal debate underlying the policy decisions―reflecting dissenting views and arguments in the Monetary Policy Committee (MPC) minutes―can be instrumental to guide private sector’s expectations and enhance the effectiveness of monetary policy.

13. Lower interest rates will not exhaust the scope for monetary policy actions. Recent measures by other central banks since the beginning of the GFC have expanded the modern monetary policy toolkit near or at the zero lower bound on interest rates (Table 1). Should further adverse shocks materialize, for example, forward guidance regarding envisaged policy actions can be considered by the BOT.

Table 1.

Overview of Unconventional Monetary Policy Measures Since the GFC

article image

References

  • Bank of England, 2011, “Instruments of macroprudential policy,” Discussion Paper, (London: Bank of England)

  • Chan, J., Clark, T., and Koop, G., 2015, “A New Model of Inflation, Trend Inflation, and Long-Run Inflation Expectations,” Federal Reserve Bank of Cleveland Working Paper No. 15–20 (Cleveland, OH: Federal Reserve Bank of Cleveland).

    • Search Google Scholar
    • Export Citation
  • Dany, G., and Garcia, J. A., forthcoming, “Inflation Dynamics in ASEAN–5 Countries,” IMF Working Paper.

  • Dincer, N., and Eichengreen, B., 2014, “Central Bank Transparency and Independence: Updates and New Measures,” International Journal of Central Banking, March, pp. 189253.

    • Search Google Scholar
    • Export Citation
  • Garcia, J. A., and Poon, A., forthcoming, “Long-Term Inflation Trends in ASEAN–5 Countries,” IMF Working Paper.

  • International Monetary Fund, 2016, “Global Disinflation in an Era of Constrained Monetary Policy,” in World Economic Outlook, October, (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Lubik, T., and Matthes, C., 2015, “Calculating the Natural Rate of Interest: A Comparison of Two Alternative Approaches,” Economic Brief No. 15–10, (Richmond, VA: Federal Reserve Bank of Richmond).

    • Search Google Scholar
    • Export Citation
1

Prepared by J. A. Garcia (lead), with contributions from G. Dany, D. Laxton, A. Poon, U. Rawat and F. Zhang.

2

Energy price inflation declined from 1.7 percent in 2014 to -15 percent in 2015, while core inflation declined from 1.6 percent to 1.1 percent over the same period. Since 2015, energy price inflation has recovered given higher commodity prices, though core inflation remained on a downward trend, reaching 0.6 percent by March 2017. Headline inflation in Thailand is more sensitive to changes in global oil prices than in other countries, as the tax structure on retail petrol prices provides a relatively smaller cushion.

3

After the onset of the Global Financial Crisis (GFC), the policy rate was cut to 1.25 percent between April 2009 and June 2010.

4

The model follows a standard specification and is estimated over the period 1995: Q1–2016: Q3. For further model details and a discussion of alternative Phillips curve specifications and robustness checks see Dany and Garcia (forthcoming) and WEO (2016).

5

This finding is robust to using survey (long-term) inflation expectations (i.e., Consensus Economics).

6

Based on the TVP-VAR estimation following Lubik and Matthes (2015).

7

See for example Clinton and others (2015). The model does not incorporate a financial sector module or possible asymmetric monetary policy transmission at low interest rates. At the same time, it also does not consider the reinforcing effect of improved communication advocated by staff.

8

See for example Bank of England (2011).

9

ASEAN–5 covers Indonesia, Malaysia, Philippines, Thailand, and Singapore.

Appendix I. Demographic Transition and the Current Account

1. Thailand is undergoing a major demographic transition. Key factors include:

2. Demographic factors have important implications for the economy. A decline in the share of a country’s working population in the total population reduces potential economic growth.1 With declining working-age population, the “demographic dividend” that Thailand reaped in the past decades will turn into a “demographic drag” on its real GDP growth, in the order of 0.5-1 percentage point per year in the coming two decades, per staff’s projections. Population aging is also putting increasing pressure on public finances. Staff projects that public spending in pensions and health would increase at least 3 percent of GDP in Thailand in the coming two decades, given demographic trends. Improving replacement ratios to more adequate levels would entail even higher fiscal costs.

3. Demographic factors are also a key determinant of the current account. Economic theory expects a lower current account for countries with a higher share of economically inactive dependent population. The life-cycle theory (Modigliani, 1970) of consumption and saving predicts that young households borrow, middle-age households save for retirement, and households in retirement dissave. Therefore, countries with a relatively young or old population are more likely to consume more than what they produce, resulting in a current account deficit.

4. Interactions among demographic variables are important. This can be illustrated in a simple OLG model (Diamond, 1965). Two effects are at work. The first is the composition effect—a higher dependency ratio lowers savings. However, this effect is counteracted by the life-cycle effect; that is, higher survival risk and lower income in the old age increase saving. Therefore, for a given aging speed, a higher dependency ratio implies higher survival risk, attenuating composition effect and leading to a higher current account. For a given dependency ratio, a higher aging speed also implies higher survival risk, reinforcing the life-cycle effect and increasing the current account.

5. Empirical evidence indicates large, economically significant effects of demographic factors on aggregate savings. For East Asian economies, most macro data analysis, including Kelley and Schmidt (1995) and Higgins and Williamson (1996 and 1997), confirmed the Coale and Hoover (1958) hypothesis that the impressive rise in Asian savings rates can be explained by the equally impressive change in dependency burdens (Williamson, 2001). For example, Higgins and Williamson (1996 and 1997) found the increase in East Asia’s saving rates due to the demographic transition from early 1970s to early 1990s was 13.6 percentage points of GDP, fully accounting for the rise in the savings rate in the region.

6. Within the EBA framework, demographic factors are expected to increase the current account norm for Thailand in coming years. The EBA model incorporates several demographic factors (Box 1). Reflecting a major demographic transition, demographic factors are expected to increase Thailand’s current account norm by 0.8 percent of GDP over 2020–30, compared with the Asian average of 0.2 percent of GDP. This is the third largest contribution of demographic factors to the norm among Asian countries, after Japan and Korea (IMF, 2017). Addressing the macroeconomic and social implications of rapid population aging is thus an urgent priority.

Demographic Variables in the EBA CA Model

The EBA CA model incorporates several demographic variables.1 Population growth (capturing shares of young population), old-age dependency ratio (capturing shares of old population), and aging speed (a proxy for future old-age dependency ratio) are used in the regression. In theory, countries with higher shares of dependent population generally have lower savings and current account balances. Therefore, both higher population growth and higher dependency ratios are linked with lower current account balances. Aging speed also affects national savings through interactions with the current dependency ratio.

Two interaction terms are included. These are the interaction of the dependency ratio with the aging speed ratio (relative to its world average), and the interaction of aging speed with the dependency ratio (relative to its world average). First, a positive coefficient on the interaction of the dependency ratio with the aging speed ratio (relative to its world average) would suggest that, given the same dependency ratio, a country with a faster aging speed is expected to have longer life expectancy and higher savings. Second, a positive coefficient on the interaction of aging speed with dependency ratio (relative to its world average) would indicate that, given the increase in aging speed, a higher dependency ratio would further increase national savings.

article image
1 The relevant metric is demographic factors in Thailand against the world average, as the assessment of the current account is multilaterally consistent.

References

  • Bloom, D., and Finlay, J., 2009, “Demographic Change and Economic Growth in Asia,” Asian Economic Policy Review, Vol. 4 (1), pp. 4564.

    • Search Google Scholar
    • Export Citation
  • Coale, A. J., and Hoover, E. M., 1958, Population Growth and Economic Development in Low-Income Countries (New Jersey: Princeton University Press).

    • Search Google Scholar
    • Export Citation
  • Demirguc-Kunt, A., 2005, “Financial Sector Development as an Essential Determinant for Achieving the MDGs: Increasing Private Credit Shown to Reduce Income Inequality,” (Washington: World Bank).

    • Search Google Scholar
    • Export Citation
  • Diamond, P. A., 1965, “National Debt in a Neoclassical Growth Model,” The American Economic Review, Vol. 55, Iss. 5 (Dec.), pp. 112650.

    • Search Google Scholar
    • Export Citation
  • Gyimah-Brempong, K., and Traynor, T. L., 1999, “Political Instability, Investment and Economic Growth in Sub-Saharan Africa,” Journal of African Economies, Vol. 8, Iss. 1 (Mar.), pp. 5286.

    • Search Google Scholar
    • Export Citation
  • Higgins, M., and Williamson, J. G., 1996, “Asian Demography and Foreign Capital Dependence,” NBER Working Paper No. 5560 (Cambridge, Massachusetts: National Bureau of Economic Research).

    • Search Google Scholar
    • Export Citation
  • Higgins, M., and Williamson, J. G., 1997, “Age Structure Dynamics in Asia and Dependence on Foreign Capital,”. Population and Development Review, Vol. 23, No. 2 (Jun), pp. 26193.

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, 2017, Regional Economic Outlook: Asia and Pacific, May 2017.

  • Julio, B., and Yook, Y., 2012, “Political Uncertainty and Corporate Investment Cycles,” The Journal of Finance, Vol. 67:1, pp. 4583.

    • Search Google Scholar
    • Export Citation
  • Kelley, A. C., and Schmidt, R. M., 1995, “Aggregate Population and Economic Growth Correlations: The Role of the Components of Demographic Change,” Demography (Dec.), 32 (4): 54355.

    • Search Google Scholar
    • Export Citation
  • Klyuev, V., 2015, “Structural Transformation — How Does Thailand Compare?,” IMF Working Paper No. 15/51 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Knodel, J., Pothisiri, W., Prachuabmoh, V., and Teerawichitchainan, B., 2015, “The Situation of Thailand’s Older Population: An update based on the 2014 Survey of Older Persons in Thailand,” Research Report 15-847 (Ann Arbor, MI: University of Michigan).

    • Search Google Scholar
    • Export Citation
  • Mason, A., 2001, Population and Economic Growth in East Asia. Population Change and Economic Development in East Asia: Challenges Met, Opportunities Seized, pp. 130 (Stanford: Stanford University Press).

    • Search Google Scholar
    • Export Citation
  • Office of the National Economic and Social Development Board (NESDB), 2016, NESDB Economic Report (Nov.).

  • Sen, K., 2014, “The Political Determinants of Economic Growth: An Analytical Review,” in The Politics of Inclusive Development: Interrogating the Evidence, Bukenya, B., Hickey, S., and Sen, K., ed. (Oxford Scholarship Online).

    • Search Google Scholar
    • Export Citation
  • Svensson, J., 1998, “Investment, property rights and political instability: Theory and evidence,” European Economic Review, Vol. 42 (Jul.), pp. 131741.

    • Search Google Scholar
    • Export Citation
  • Williamson, J. G., 2003, “Demographic Change, Economic Growth, and Inequality,” in Population Matters: Demographic Change, Economic Growth, and Poverty in the Developing World, Birdsall, N., Kelley, A., and Sinding, S., ed. (Oxford Scholarship Online).

    • Search Google Scholar
    • Export Citation
1

Prepared by Yiqun Wu.

2

Cash-constrained rural households, with higher marginal propensity to consume, are more likely to be affected by negative shocks.

3

The rolling regression uses the inverse of the ICRG index, which measures higher risk/weaker environment. The rolling coefficient suggests that higher risk translated to weaker private investment (and a higher current account) in the most recent years. The change in the coefficient is large and statistically significant.

1

Bloom and Finlay (2009) suggest that demographic factors contributed significantly to economic growth in East Asian countries from 1965 to 2005. The working-age population is generally more productive and saves more, increasing domestic resources for investment. There is also evidence that population aging may have a negative impact on total factor productivity growth.

Thailand: Selected Issues
Author: International Monetary Fund. Asia and Pacific Dept