BIS Working Papers No 1141

Is high debt constraining monetary policy? Evidence from inflation expectations

by Luis Brandao-Marques, Marco Casiraghi, Gaston Gelos, Olamide Harrison and Gunes Kamber

Monetary and Economic Department

November 2023

JEL classification: E31, E41, E52, E62.

Keywords: inflation expectations, monetary policy, fiscal dominance, debt.

群内每日免费分享5份+最新资料

300T网盘资源+40万份行业报告

为您的创业、职场、商业、投资、亲子、网赚、艺术、健身、心理、个人成长 …… 全面赋能！

关注公众号获取更多资料

添加微信，备
注“入群”
立刻免费领取
200套知识地图+最新研报

致终身学习者社群

行业报告/思维导图/电子书/资讯情报

收钱文案、增长黑客、产品运营、品牌企划、营销战略、办公软件、会计财务、广告设计、摄影修图、视频剪辑、直播带货、电商运营、投资理财、汽车房产、餐饮烹饪、职场经验、演讲口才、风水命理、心理思维、恋爱情趣、美妆护肤、健身瘦身、格斗搏击、漫画手绘、声乐训练、自媒体打造、效率软件工具、游戏影音……

扫码先加好友，以备不时之需

BIS Working Papers are written by members of the Monetary and Economic Department of the Bank for International Settlements, and from time to time by other economists, and are published by the Bank. The papers are on subjects of topical interest and are technical in character. The views expressed in them are those of their authors and not necessarily the views of the BIS.

This publication is available on the BIS website (www.bis.org).

Is high debt constraining monetary policy? Evidence from inflation expectations

Luis Brandao-Marques*, Marco Casiraghi† , Gaston Gelos‡ , Olamide Harrison*, and Gunes Kamber*1

Abstract

This paper examines whether high public debt levels pose a challenge to containing inflation. It does so by assessing the impact of public debt surprises on inflation expectations advanced- and Emerging Market Economies. It finds that debt surprises raise long-term inflation expectations in Emerging Market Economies in a persistent way, but not in advanced economies. The effects are stronger when initial debt levels are already high, when inflation levels are initially high, and when debt dollarization is significant. By contrast, debt surprises have only modest effects in countries with inflation targeting regimes. Increased debt levels may complicate the fight against inflation in Emerging Market Economies with high and dollarized debt levels, and weaker monetary policy frameworks.

JEL classification: E31, E41, E52, E62.

Keywords: inflation expectations, monetary policy, fiscal dominance, debt.

Introduction

The COVID-19 crisis propelled sovereign debt levels around the world to new heights. Between 2019 and 2022, in advanced economies, debt levels rose from 103.9 to 112.5 percent of GDP, and in Emerging Market Economies, they grew from 55.1 to 64.6 percent of GDP. Sovereign debt levels are expected to remain high in the coming years (IMF 2022), and demands on fiscal policy remain substantial. At the same time, higher levels of interest rates could increase debt services considerably.

Could the rise in government debt levels pose a problem for monetary policy? This question is of particular relevance as central banks are struggling with the major challenge posed by inflationary pressures. In principle, various mechanisms by which high sovereign debt levels can complicate monetary policymaking are conceivable. Central banks may be (seen as) hesitant to raise interest rates as much as needed to achieve their inflation objective, out of concern for debt sustainability. They may also be concerned about the impact of rising rates on their own net income if they remunerate excess reserves, and on their balance sheets as they proceed to actively shrink them. In an extreme form of fiscal dominance, people may believe that the central bank could try to inflate away part of the debt or resort to outright monetization of future deficits, which in turn would lead inflation expectations to shoot up. The fiscal theory of the price level has reformulated these notions, arguing that price levels adjust so that the real value of government debt equals the present value of taxes less spending (Cochrane 2023).

Such concerns are, in principle, relevant for both advanced- and Emerging Market Economies. In many Emerging Market Economies, these issues used to be very much at the forefront until the early 2000s. For example, Celasun, Gelos, and Prati (2004) found that fiscal variables were an important determinant of inflation expectations in major Emerging Market Economies. Fiscal consolidation, the adoption of more credible macroeconomic and monetary frameworks, and the granting of autonomy to central banks lessened these concerns in these countries over the following years. However, central bank credibility remains less established in some Emerging Market Economies, which as a group are generally still more vulnerable to external shocks. Although sovereign foreign currency exposures have fallen, currency mismatches in the private sector remain widespread (BIS 2019), posing vulnerabilities that ultimately can threaten the fiscal position of these countries. Because of this historical background and the structural specificities of Emerging Market Economies, in this study we differentiate between advanced- and Emerging Market Economies.

Several other studies have empirically investigated the relationship between fiscal variables and inflation and inflation expectations in economies with inflationtargeting central banks (Catão & Terrones 2003; de Mendonça & Machado 2013; Celasun, Gelos, and Prati 2004; Coibion, Gorodnichenko, and Weber 2021). These papers generally find that tightening fiscal balances are helpful in reducing inflation expectations and observed inflation, with stronger effects in economies with high and persistent inflation. More recently, Cevik and Miryugin (2023) investigate the impact of fiscal shocks on inflation in a cross-country study, finding that both headline and core measures of inflation increase in response to expansionary shifts in the fiscal policy stance. Similarly, Banerjee and others (2023a), using data for advanced economies, find that the inflationary effect of fiscal deficits depends on the prevailing fiscal-monetary policy regime; for Emerging Market Economies Banerjee and others (2023b) report that the exchange-rate channel plays an important role. This paper takes a somewhat different approach.

Specifically, we investigate how sovereign debt may affect inflation dynamics by examining its impact on inflation expectations.2 For identification purposes, we focus on debt surprises. We trace out the response of inflation expectations to an unexpected increase in sovereign debt using local projections. Since we are interested in the effects of debt and debt sustainability, to avoid capturing demand-side effects related to unexpected fiscal expansions, we focus on long-run expectations (5 years ahead), under the assumption that any inflationary effects of such demand pressures are shorter-lived. To further minimize demand effects (such as initial price pressures leading to a de-anchoring of inflation expectations unrelated to longer-term debt concerns), we also condition on existing debt levels.

We find that in response to debt surprises, inflation expectations in Emerging Market Economies rise significantly, whereas they do not in advanced economies. Specifically, for Emerging Market Economies, a 10-percentage-point increase in the government-debt-to-GDP ratio triggers a statistically significant 20-basis point increase in long-term inflation expectations within the first year after the shock, which subsequently reaches a peak of 70-basis point within the second year. The effect is much larger in Emerging Market Economies that do not feature inflation targeting regimes, because these inflation-targeting regimes are typically associated with sound monetary frameworks and autonomous central banks that help contain expectations of fiscal dominance. These results are broadly consistent with earlier ones of Catão and Terrones (2005) who find a strong positive association between deficits and inflation among high-inflation and developing countries, but not among low-inflation advanced economies.

However, in line with the notion that debt sustainability concerns are behind our results, the effect on inflation expectations is not present for Emerging Market Economies with low sovereign debt levels. These findings are in line with those by Kwon, McFarlane, and Robinson (2009), who report evidence consistent with the notion that the risk of a debt-inflation trap is significant in highly indebted economies.

Theoretical mechanisms

We are primarily interested in the extent to which government debt and concerns regarding solvency of the sovereign might affect monetary policy and the ability of a central bank to achieve its price stability objective. A large theoretical literature establishes direct links between government debt and the price level. The starting point is fiscal dominance: that a central bank tightens its monetary policy stance in response to a rise in inflation by less than it would otherwise because of the level of government debt.

To the extent that the government debt is in local currency, high government debt may lead central banks to limit increases in the policy rate out of concern for the government’s solvency. Economic agents, in turn, anticipate this reaction, raising their inflation expectations. In the extreme, firms and households may fear that the central bank may resort to outright debt monetization. Implicit in these mechanisms is that some portion of government debt is not backed by the government’s current and future primary surpluses. These arguments were well established by Sargent and Wallace (1981) and in subsequent forward-looking models of inflation (Aiyagari and Gertler 1985; Calvo 1988; Bohn 1988, among others). Although Sargent and Wallace (1981) never discuss the fiscal stance or debt sustainability, the key point is that the fiscal path is predetermined and needs to generate seigniorage. A corollary of these arguments is that a tighter monetary policy may deliver lower inflation now (and in extreme cases not even that) but surely higher inflation in the future if fiscal policy dominates monetary policy. This is because (sufficiently) forward-looking agents anticipate the need for looser monetary policy later to inflate away the public debt.

A similar mechanism, albeit with important theoretical departures, is developed in the fiscal theory of the price level or FTPL (Leeper 1991, Sims 1994, Woodford 1994, Cochrane 2005, and Benigno 2020). The central mechanism in the FTPL is that prices adjust so that the real value of government debt equals the present value of taxes less spending (Cochrane 2023). More recently, Bianchi, Faccini, and Melosi (2022) developed a theoretical framework where monetary-policy fiscal interaction is more acute when fiscal imbalances are large and fiscal credibility wanes, making it harder for the monetary authority to stabilize inflation around its desired target. In their framework, a fiscally-driven rise in trend inflation does not necessarily stem from a lack of credibility of the central bank, but rather from the incompatibility between the objectives of the central bank and the expected behavior of the fiscal authority in credibly stabilizing its debt.

A different, but related mechanism developed more recently in Arellano and others (2020) argues that sovereign default risk on external defaultable government debt affects inflation expectations. In their model, by assumption, inflation is high (and consumption low) whenever the sovereign defaults because productivity is low. Specifically, they assume that total factor productivity depends on the government’s credit standing. Taking expectations, when default risk is high, expected inflation is high. Thus, any shock that increases the sovereign’s default risk – for instance, a productivity or a terms of trade shock – raises (near-term) inflation expectations. In this paper, we identify shocks to debt that are plausibly orthogonal to shocks that move inflation and output and find that these shocks are relevant for the conduct of monetary policy as summarized by dislocations of long-term expectations in Emerging Market Economies. Furthermore, we establish the presence of nonlinearities by demonstrating higher sensitivity of inflation expectations to debt surprises in high-debt and high-inflation Emerging Market Economies.

Data and empirical methods

Econometric approach

Our goal of assessing how debt sustainability concerns might impinge on monetary policy and its ability to anchor inflation expectations faces various empirical challenges, mostly related to endogeneity. First, in the relationship between debt levels and inflation, causality probably flows both ways. For example, policymakers may tighten fiscal policy in response to a rise in inflation expectations, which could result in a negative correlation of changes in debt and inflation expectations (Celasun, Gelos, and Prati 2004). On the other hand, fiscal policy actions which affect debt may also directly influence inflation expectations through aggregate demand effects, which we would like to exclude. Second, the co-movement between debt levels and inflation could be spurious and driven instead by an omitted variable. Particularly in Emerging Market Economies, adverse GDP shocks stemming, say, from a sudden stop of capital flows as a result of changes in global risk aversion often come along with exchange rate depreciations that induce a rise in inflation expectations. In such situations, we would observe a simultaneous increase in the debt-to-GDP ratio and inflation expectations that is not (necessarily) driven by the mechanisms we are trying to capture. Moreover, information shocks that manifest as positive or negative interest rate shocks – reflecting a central bank communicating a surprisingly positive or negative economic outlook – could cause higher/lower nominal debt along with higher or lower inflation expectations (Nunes and others 2022).In addition, other shocks to inflation expectations that have an impact on the nominal interest rate bill (hence, on debt sustainability) could introduce a positive (negative) bias for local (foreign) currency denominated instruments (Celasun and others 2004).

Moreover, since we are interested in mechanisms related to the government’s intertemporal budget constraint, as discussed earlier, to the extent possible, we would like to exclude short-term inflationary demand effects stemming from fiscal expansions.

To address these concerns, we pursue a threefold strategy. First, we use forecast errors from various vintages of International Monetary Fund (IMF) World Economic Outlook (WEO) publications to identify unanticipated exogenous shocks (surprises) to government debt, which both mitigate endogeneity concerns. Moreover, debt surprises may capture changes in debt that do not necessarily move one-for-one with the budget deficit and are less likely to be driven by changes in government expenditures (Singh and others 2005). For example, such debt surprises may include the recognition of hidden contingent liabilities. Second, we focus on long-term inflation expectations (ie 5-year ahead expectations), which are less likely to be affected by short-term aggregate demand effects (assuming reasonable estimates of aggregate price stickiness). Third, to further reduce the influence of aggregate demand effects and simultaneity, we also condition our estimations on existing debt levels; if debt surprises affect long-term inflation expectations more the higher preexisting debt levels are, this is unlikely to stem from such demand effects or other shocks.

The debt shocks are defined as the difference between the realized annual growth rate of the debt to GDP ratio and the corresponding IMF forecast, which is published in October. In other words, the shock is unanticipated change in the debtto-GDP ratio that is observed in the last quarter of each year. Specifically, the debt

where the lowercase debt variables denote debt-to-GDP ratios.

As argued in the literature using this approach (eg Blanchard and Perotti 2002; Auerbach and Gorodnichenko 2012, 2013; Abiad and others 2016; Furceri and Li 2017), debt levels projected in the October WEO publications are likely to reflect all available information at the time of publication. Any difference between realized nominal debt levels and the October WEO projections are thus likely to reflect unanticipated policy changes or debt revaluations.

One concern with the government-debt forecast errors is that they may not represent true surprises, because they may be forecastable. To verify the sensitivity of our results to such issues, we consider two approaches used in the literature. First, we follow the method used in the October 2017 WEO and Magud and Pienknagura (2022) by regressing debt forecast errors on forecast errors of inflation and real GDP growth, and using the residuals from this regression as our debt shock. The other approach, which follows Auerbach and Gorodnichenko (2013), consists of regressing the debt forecast errors on lagged macroeconomic variables, such as the output gap, debt levels, and the primary balance. The residuals from this regression are then used as the debt shock. In both approaches, we include a set of country fixed effects. The results and conclusions discussed below are quantitatively and qualitatively robust to these alternative shocks.

As previously discussed, government debt shocks can affect inflation expectations through multiple channels. The impact on long-term expectations will depend on the monetary policy regime, the degree of central bank independence and other institutional features, and the anchoring of inflation expectations. Thus, analyzing the sensitivity of long-term inflation expectations to unexpected shifts in government debt helps shed light on whether non-aggregate demand mechanisms are at work.

We use panel regressions and the local projections method of Jordà (2005) to estimate the responses of long-term inflation expectations to the identified government debt shocks over different horizons. As noted by Auerbach and Gorodnichenko (2012), the local projections approach easily accommodates nonlinearities in impulse response functions and cross-country correlation in the error term.

The baseline specification we estimate is given by:

$$\tag{1}$$

where ??ሺ??ሻ௜,௧ା௟௘ denotes the measure of inflation expectations for horizon H for country I at time t+l, ??௜,௧ି௦ௗ ,diu t s − are debt surprises, ??௜௧ᇱ X is a vector of controls (to be introduced later as robustness), h i and h are country and time fixed effects, and ??௜,௧ା௛ is a random disturbance. We include lags j of the debt shock to reduce concerns that the dynamics of the impact are due to properties of the shock (ie that the shock is persistent and to some extent predictable)3 as opposed to the operating mechanisms.4 The coefficients, ??଴௛, of the response of inflation expectations to debt surprises specification are estimated using the within estimator for each projection period, with standard errors coming from Driscoll and Kraay’s (1998) heteroskedasticity and autocorrelation-consistent estimator.

Summary Statistics

The sample includes 41 emerging market and developing economies, 7 low-income countries, and 34 advanced economies over the period 2000 – 2020.5 We group emerging market- and developing economies together. We use five-year-ahead inflation expectations from surveys of professional forecasters reported by Consensus Economics.6 We use inflation expectations data at a semi-annual frequency because for most of our sample, in the Consensus Economics forecasts, long-term inflation forecasts are only collected at a semi-annual frequency (in April and October). Our choice of expectations from professional forecasters is driven solely by their consistent availability across countries of inflation expectations data.7 Data on government debt, GDP, output gap, and consumer price inflation are from the IMF’s WEO dataset, which covers all economies from 1995 to 2019. In constructing our debt shock, the forecast for government debt and GDP for each country in a given year is taken from the October vintage of WEO forecasts of that year. The realized values of these variables are then taken from the following year’s October WEO database.

We follow Flores and others (2023) in cleaning the data to minimize measurement errors and inconsistencies across data vintages. The debt forecast data in WEO vintages are available for a small number of advanced economies before 2002, which imposes the starting year of the sample for our empirical analysis (Graph 1). Graph 2 summarizes the debt shocks for advanced- and Emerging Market Economies by presenting the interquartile range, the mean, and median forecast errors by group of economies. The mean- and median debt-to-GDP forecast errors are similar across the two groups. The median forecast error is 0.15 for Emerging Market Economies, and 0.05 for advanced economies, and the mean forecast error is 0.65 for Emerging Market Economies and 0.59 for advanced economies; they are not statistically different from each other (Table 1).

Table 1

SummaryStatistics
	(Emerging)	(Advanced)		(Difference)
	mean	s.d.	mean s.d.	b	t
Inf. Expectations, 5-yr	3.91	1.92 2.04	0.44	1.87***	(17.92)
Debt Surprise	0.65	4.49	0.59 4.14	0.06	(0.19)
Debt to GDP	37.91	18.27	57.10 44.62	-19.19***	(-7.59)
CPI, y-o-y	5.62	7.33	1.69 1.71	3.93***	(9.56)
Observations	352		426	778

5 Table A.1 in Annex I contains the list of economies included in our sample.

Graph 2

Data availability

Notes: The chart shows the number of economies for which debt to GDP shocks are available for the second WEO vintage of the year, usually published in October. The grey bars denote the count for advanced economies while the blue bars denote the count for Emerging Market Economies

The positively skewed distribution suggests a bias towards optimism in IMF forecasts (IMF 2021), although this bias is found to be more pronounced, the longer the forecast horizon (Flores and others 2023). Graph 3 plots the evolution of the size of forecast errors over time. Errors tend to become large and positive during recessions, as for example during the global financial crisis (GFC).

Debt-to-GDP surprises on impact

Notes: The Chart shows the interquartile range (shaded region), mean (blue dot), and median (blue line in the shaded box) for the debt-to-GDP shock at the beginning of the projection period for the entire sample.

Advanced economies

Notes: Charts show the interquartile range (shaded region), mean (blue dot), and median (blue line in the shaded box) for the debt-to-GDP shock for Emerging Market Economies (left) and advanced economies (right).

Results

Main results – Emerging Market- vs. Advanced Economies

The results indicate that debt surprises do affect inflation expectations in Emerging Market Economies. Graph 4 and Tables 2, 3 and 4 present the results from our baseline specification, displaying the impact of a debt shock on inflation expectations for the following three years. When we use the full sample, our results suggest that a surprise 10-percent rise in government debt to GDP leads to a significant increase in long-term inflation expectations after about two years. However, this result seems to be driven mainly by Emerging Market Economies. When estimate our baseline specification for emerging market- and advanced economies separately, the impact of a debt shock on inflation expectations significantly differs. For advanced economies, the response of inflation expectations to a debt shock is zero across all horizons. For Emerging Market Economies, however, an unanticipated increase in the government debt to GDP ratio is associated with a rise in long-term inflation expectations. Specifically, a 10 percent unanticipated rise in the government debt to GDP ratio results in a statistically significant 20-basis point hike in long-term inflation expectations within the first year after the shock, and reaching a peak of 70-basis point within the second year.

Interestingly, our results suggest that the impact of a debt shock on inflation expectations is delayed and builds over time. Several factors may explain this finding. First, data on public finances and national accounts become available with a lag, and given that the first survey on long-term inflation expectations occurs at the end of the first quarter of the year, the full extent of the forecast error on public debt may not be available to forecaster, and forecasters may be slow to incorporate this information fully in their new forecasts. Second, the delayed response of inflation expectations may be related to inflation expectations being more backward looking or adaptive, consistent with the evidence from inflation expectations surveys as in Coibion and Gorodnichenko (2012). For instance, recent work by Chen and others (2022) also finds that the cross-section average of individual forecasts of expectations in survey data tends to underreact to shocks initially but overreact in the mediumterm.

Graph 4

Response of 5-year inflation expectations to government debt shocks, baseline (Basis points, annual rate)

Notes: t=0 is the quarter of the shock. The graphs plot for the relevant horizon the 5-year ahead inflation expectations response to a 10 percent surprise in the debt-to-GDP ratio. The blue dots denote the inflation expectations response for Emerging Market Economies in our sample while the red dots denote the corresponding response for advanced economies. The whiskers represent 90 percent confidence intervals. The chart on the left shows the response for the full sample.

Another possible channel is that debt shocks today may signal future higher deficits, which would then support higher inflation expectations. For example, bad economic news may prompt fiscal policy to turn stimulative for several years. Such an explanation would be consistent with the estimated unconditional and conditional effects. While our study tries to control for contemporaneous demand shocks, it does not include at forward-looking effects, but an exploration of this issue suggests that this channel is not important. In our sample, debt surprises do not forecast persistently higher deficits.8

These results suggest that inflation expectations in Emerging Market Economies appear more sensitive to the fiscal position than in advanced economies. Furthermore, fiscal shocks increase the risk that long-term inflation expectations deanchor. The heightened sensitivity of inflation expectations to the fiscal stance could suggest that, on average, fiscal dominance – in which the fiscal authority’s solvency constraint determines inflation – is more likely to prevail in emerging market- than in advanced economies, possibly in part reflecting more limited fiscal capacity among the former. The result may also capture broader concerns that emerging market central bank independence is less secure than in advanced economies, given weaker institutional frameworks and protections (Unsal and others 2022).9 These mechanisms are explored further below, where we examine the role of inflation-targeting regimes and initial debt levels in shaping the impact of the debt surprises on inflation expectations.

Response of 5-year inflation expectations, baseline (full sample)

	(1)	(2)	(3)	(4)	(5)
	h = 4	h = 10	h = 16 0.0237*	h = 22	h = 28
Debt Surprise	0.0042	0.0114		0.0374*	0.0262*
No. of Obs.	(0.0042)	(0.0072) 672	(0.0126) 671	(0.0196) 607	(0.0153) 606
Adjusted R sq.	778 0.60	0.32	0.12	0.05	0.03

Notes: All regressions include 2-month and 8-month lags of 5-year ahead inflation expectations, a 12-month lag of the debt shock, and country and time fixed effects. Driscoll-Kraay standard errors in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01

Response of 5-year inflation expectations, baseline (Emerging Market Economies)
Table 3

	(1) h = 4	(2) h = 10	(3) h = 16	(4) h = 22	(5) h = 28	(6) h = 34
Debt Surprise	0.0010	0.0211**	0.0447**	0.0680**	0.0484**	0.0221*
	(0.0064)	(0.0105)	(0.0192)	(0.0300)	(0.0238)	(0.0114)
No. of Obs.	352	280	279	252	251	232
Adjusted R sq.	0.63	0.40	0.17	0.12	0.05	0.03

Response of 5-year inflation expectations, baseline (advanced economies)

	(1) h = 4	(2) h = 10	(3) h = 16	(4) h = 22	(5)	(6)
Debt Surprise	0.0042*	0.0008	0.0015	0.0013	h = 28 0.0002	h = 34 -0.0008
	(0.0025)	(0.0028)	(0.0021)	(0.0021)	(0.0033)	(0.0030)
No. of Obs.	426	392	392	355	355	331
Adjusted R sq.	0.32	0.11	-0.05	-0.08	-0.10	-0.13

State dependence

Given the focus of this paper on assessing the role of fiscal policy in shaping inflation expectations, we pay particular attention to whether measures of fiscal buffers and sustainability matter for the impact of fiscal shocks. We also explore the extent to

which the ex-ante level of inflation matters in shaping the response dynamics to the shock.

Government debt level

Although our empirical strategy so far has sought to minimize the potential impact of short-term demand pressures stemming from changes in the fiscal stance, it is still possible that our estimates may be contaminated by such effects. For example, in environments with weak central bank credibility, a rise in inflation driven by a fiscal expansion may have persistent effects, de-anchoring inflation expectations through mechanisms not directly driven by debt concerns.

Therefore, to go one step further in identifying those effects we are after, we condition our estimation on initial debt levels. If debt surprises affect long-term inflation expectations more, the higher the initial debt stock is, this is a strong indication that the change in inflation expectations is driven by debt concerns. When present, it is likely that such effects are nonlinear.

We focus on the difference in the impact of the debt shock on Emerging Market Economies in the 10th and 90th percentiles of initial government debt levels. We refer to these as low- and high debt groups, respectively. The modified empirical specification now includes an interaction term between the debt shocks and the initial debt level:

$$\tag{1}$$

where i t, 1debt − is the government debt-to-GDP level at the end of the year before the shock. The coefficient of interest, 0 0 ( , 1)h h i t debt + − , depends on the debt level with a positive value for the marginal effect indicating that inflation expectations are more sensitive to debt shocks in Emerging Market Economies with higher debt levels.10

Our results show that the sensitivity of inflation expectations to debt surprises depends on the debt level. Graph 5 illustrates the state dependence by tracing out the response of inflation expectations for economies with government debt-to-GDP levels at the 10th and 90th percentiles of the Emerging Market Economies. The graph shows that for countries in the low-debt group, the impact of the unanticipated debt increase on inflation expectations is statistically indistinguishable from zero. By contrast, high-debt countries experience as much as a 100-basis point increase in long-term inflation expectations after two years in response to a 10-percent surprise rise in the debt-to-GDP ratio. Our interpretation is that debt sustainability concerns are more prevalent in Emerging Market Economies with high debt levels, and forecasters anticipate the effect this might have on the central bank’s ability to stabilize prices.

Response of 5-year inflation expectations, by initial debt level classification (Emerging Market Economies)
Table 5

	(1) h = 4	(2) h = 10	(3) h = 16	(4) h = 22	(5) h = 28	(6) h = 34
Debt Surprise	0.0334* (0.0178)	0.0364 (0.0267)	0.0165 (0.0412)	-0.0582 (0.0440)	-0.0552 (0.0425)	0.0207 (0.0168)
Debt to GDP	-0.0034 (0.0044)	0.0050 (0.0086)	0.0017 (0.0093)	-0.0097 (0.0071)	-0.0112 (0.0073)	-0.0179*** (0.0067)
Debt Surprise # Debt	-0.0006*	-0.0003 (0.0007)	0.0006 (0.0009)	0.0027** (0.0012)	0.0022** (0.0011)	-0.0000
No. of Obs.	(0.0004) 333	276	275	248	247	(0.0003) 228
Adjusted R sq.	0.65	0.40	0.19	0.24	0.11	0.04

Response of 5-Year Inflation Expectations to Government Debt Shocks, by Initial Debt Level Classification

(Basis points, annual rate)

Graph 5

Notes: t=0 is the year of the shock. The graphs plot for the relevant country groups the inflation expectations response to a 10 percent surprise in the debt-to-GDP ratio. The blue dots denote the inflation expectations response for economies with a debt level at the 10th percentile of the relevant country group sample, while the red dots denote the corresponding response for economies with a debt level at the 90th percentile of the relevant country group sample. The whiskers represent 90 percent confidence intervals.

Response of 5-year inflation expectations, by initial debt level (Advanced Economies) Table 6

	(1)	(2) h = 10	(3) h = 16	(4)	(5)	(6)
Debt Surprise	h = 4 0.0018	-0.0028	-0.0016	h = 22 0.0004	h = 28 0.0023	h = 34 0.0025
Debt to GDP	(0.0052) 0.0001	(0.0053) -0.0004	(0.0050) -0.0007	(0.0052) 0.0000	(0.0051) -0.0008	(0.0065) 0.0003
Debt Surprise # Debt	(0.0010) 0.0000	(0.0014) 0.0000	(0.0019) 0.0001	(0.0017) 0.0000	(0.0021) -0.0001	(0.0019) -0.0001
	(0.0001)	(0.0001)	(0.0001)	(0.0001)	(0.0001)	(0.0001)
No. of Obs.	369	340	340	305	305	284
Adjusted R sq.	0.29	0.07	-0.09	-0.12	-0.13	-0.15

Inflation level

We next explore the implications of pre-existing inflation levels for the sensitivity of inflation expectations to debt surprises. To this end, we divide the emerging market sample into initial inflation percentiles, and condition the debt shock on the preshock inflation level. The modified empirical specification is the same as (2), except that it includes the starting level of inflation instead of the previous year’s level of debt, both on its own and interacted with the debt shock. The marginal effect of the debt surprise is ??଴௛ + ??଴௛??௜,௧ିଵ, with a positive value indicating that inflation expectations are more sensitive to debt shocks in economies with higher inflation levels.

Initial inflation levels do indeed matter. We evaluate the effects of debt shocks for low (10th percentile) and high (90th percentile) inflation levels. Long-term inflation expectations increase by about 40 basis points after two years for countries with initial inflation levels at the 90th percentile of the emerging market sample, rising to 50 basis points after three years (Graph 6 and Table 7). By contrast, the impact of the debt shock is statistically indistinguishable from zero for economies with initially low inflation levels. The differences between the responses under low and inflation is statistically significant at the 90 percent confidence level. When inflation is already high, the willingness and the ability of the central bank (or of the fiscal authority) to tighten sufficiently in response to debt surprises might be more limited.

Response of 5-year inflation expectations to government debt shocks, by initial inflation level classification

(Basis points, annual rate)

Graph 6

Notes: t=0 is the year of the shock. The graphs plot for the relevant country groups the inflation expectations response to a 10 percent surprise in the debt-to-GDP ratio. The blue dots denote the inflation expectations response for economies with an initial inflation at the 10th percentile of the relevant country group sample, while the red dots denote the corresponding response for economies with an initial inflation level at the 90th percentile of the relevant country group sample. The whiskers represent 90 percent confidence intervals.

The profile of the impulse response for high inflation economies to the debt surprise suggests that inflation expectations remain dislodged for at least three years after the debt shock. This stands in contrast to the results for the interaction between the debt shock and the initial debt level – in that case, medium-term inflation expectations return to the pre-shock level after roughly three years. The difference between the two profiles might be related to the persistence of inflation in economies with chronically high inflation (Dornbusch 1982; Rodriguez 1982; Buiter & Grafe 2001). In other words, past inflation might be driving current inflation and inflation expectations for these types of economies. When the debt shock hits, the contemporaneous effect of the shock is compounded by the impact of lagged inflation, and this causes a more persistent dislocation of medium-term expectations.

Response of 5-year inflation expectations, by initial inflation level (Emerging Market Economies)
Table 7

	(1) h = 4	(2) h = 10	(3) h = 16	(4) h = 22	(5) h = 28	(6) h = 34
Debt Surprise	0.0027 (0.0081)	0.0027 (0.0109)	0.0003 (0.0104)	-0.0116 (0.0082)	-0.0090 (0.0081)	-0.0240 (0.0224)
CPI, y-0-y	0.0162*	0.0429** (0.0209)	0.0580** (0.0277)	0.0320 (0.0346)	0.0358 (0.0405)	0.0151 (0.0349)
Debt Surprise # Inflation	(0.0094) -0.0005	0.0003	0.0018**	0.0045***	0.0035***	0.0076**
No. of Obs.	(0.0004) 334	(0.0007) 268	(0.0007) 267	(0.0006) 240	(0.0008) 239	(0.0036) 220
Adjusted R sq.	0.68	0.50	0.39	0.46	0.31	0.13

Debt dollarization

So far, we have explored the relationship between inflation expectations and debt surprises irrespective of the currency composition of government debt. How does this composition affect the sensitivity of medium-term expectations to debt surprises? On the one hand, concerns about debt monetization and its inflationary impact are only directly present for the local currency share of government debt (Panizza and Tadei 2020, and Sunder-Plassmann 2020). This direct channel should imply a larger sensitivity of inflation expectations for a smaller foreign currency debt share. On the other hand, a large foreign currency debt share increases the vulnerability of a sovereign to external shocks, reducing debt sustainability. And large adverse shocks are typically accompanied by sizeable depreciations, which increase the real value of the foreign-currency-denominated government debt and cause inflationary pressures (Banerjee and others 2023b). Moreover, when the government is confronting a fiscal crisis, a smaller share of local-currency debt implies that a higher inflation level is needed to achieve the same overall reduction in real debt value.11

To explore these mechanisms further, we condition on the initial foreign currency debt share prior to the debt surprise. We also control for the size of debt and its interaction with the debt surprise shock because the size of debt and its currency composition could be correlated: sovereigns that cannot issue much debt due to concerns about price stability might be forced to issue mostly foreign currency debt. If the magnitude of the response to the debt surprise is higher for economies with higher debt, this would introduce a negative bias for the interaction term of debt surprises and the share of foreign currency debt. Accordingly, the regression equation is modified yielding:

$$\tag{3}$$

The marginal effect of the debt surprise – ??଴௛ + ??଴௛????????௜,௧ିଵ + ??ଶ௛??????ℎ??????௜,௧ିଵ + ??ସ௛??????ℎ??????௜,௧ିଵ × ????????௜,௧ିଵ – depends not only on the currency composition of government debt, ??????ℎ??????௜,௧ିଵ , but also on the government debt level.

The results show a positive and statistically significant effect of the interaction between foreign currency debt levels, debt levels, and debt surprises, albeit only after 22 months. Evaluating the effects at the tail ends of the foreign currency debt share distribution contrasts the sensitivity of medium-term inflation expectations to debt surprises for higher vs lower foreign currency debt share Emerging Market Economies. Graph 7 shows the effects over time of a positive debt surprise shock for economies with low (10th percentile) and high (90th percentile) foreign currency debt shares, with government debt evaluated at its average across economies and periods in the estimation sample. The results suggest that the higher vulnerability to external shocks from fx debt, combined with a higher required inflation rate in case of a lower share of domestic debt outweigh the mitigating effects from lower concerns about the ability of the central bank to monetize the debt (see also Table 8).

Response of 5-Year Inflation Expectations to Government Debt Shocks, Emerging Market Economies – Initial FX Share of Government Debt

(Basis points, annual rate)

Notes: t=0 is the year of the shock. The graphs plot for Emerging Market Economies the inflation expectations response to a 10 percent surprise in the debt-to-GDP ratio. The blue dots denote the inflation expectations response for economies with an initial FX share of government debt beneath the 10th percentile while the red dots denote the corresponding response for economies with an initial FX share of government debt above the 90th percentile. The whiskers represent 90 percent confidence intervals. The responses are evaluated at the average government debt across economies and periods in the estimation sample.

Response of 5-year inflation expectations, by initial FX debt share classification (Emerging Market Economies)
Table 8

	(1) h = 4	(2) h = 10	(3) h = 16	(4) h = 22	(5) h = 28	(6) h = 34
Debt Surprise	0.0039	0.0136 (0.0150)	0.0039 (0.0164)	-0.0234 (0.0162)	-0.0208	-0.0011
FX Debt Share	(0.0092) 0.0078***	0.0105***	0.0163***	0.0168***	(0.0128) 0.0193***	(0.0153) 0.0202***
Debt Surprise # FX Debt Share	(0.0023) -0.0002	(0.0037) 0.0001	(0.0049) 0.0009	(0.0056) 0.0021***	(0.0061) 0.0017***	(0.0068) 0.0008**
	(0.0003)	(0.0004)	(0.0006)	(0.0008)	(0.0005)	(0.0004)
No. of Obs.	321	261	260	236	235	217
Adjusted R sq.	0.65	0.43	0.24	0.26	0.18	0.07

Inflation targeting

To explore the role of monetary policy frameworks and credibility, we divide the emerging market economy sample based on whether the central bank operates an inflation targeting (IT) regime according to the IMF’s Annual Report on Exchange

Arrangements and Exchange Restrictions, and re-estimate specification (1) for each subsample.

Economies with non-IT pursuing central banks display a higher sensitivity of long-term inflation expectations to a debt surprise – but the difference vis-à-vis IT economies is not significant. Nevertheless, the profile of the impulse response for non-IT Emerging Market Economies matches the delayed onset of de-anchored expectations shown in the baseline results for Emerging Market Economies (Graph 8). These results suggest that inflation-targeting emerging market central banks are better able to anchor medium-term expectations in response to fiscal shocks than their non-IT counterparts, possibly because there is more fiscal dominance in non-IT regimes.

Response of inflation expectations to government debt shocks, by inflation targeting regime classification

(Basis points, annual rate)

Graph 8

Advanced Economies

Notes: t=0 is the year of the shock. The graphs plot for the relevant country groups the response of 5-year inflation expectations to a 10 percent surprise in the debt-to-GDP ratio. The blue dots denote the inflation expectations response for economies with inflation-targeting central banks in the relevant country group sample while the red dots denote the corresponding response for economies with non-inflation targeting central banks. The shaded region represents 90 percent confidence intervals.

These results are not directly supportive of the empirical findings by Bianchi, Faccini and Melosi (2022) who also argue that unfunded debt shocks cause higher expected inflation via fiscal-monetary coordination. One of the restrictions for the identification of unfunded debt shocks in their approach is that the debt-to-GDP ratio declines in the years following the shock because monetary policy accommodates the fiscal expansion. Thus, drops in debt-to-GDP ratios would be associated with higher expected inflation, with greater sensitivity for tighter fiscal-monetary coordination (in the form of a larger central bank response to unfunded fiscal shocks). In contrast, our results above suggest higher debt-to-GDP ratios are associated with higher expected inflation.

Finally, we perform the same exercise for advanced economy economies and find that there is no clear distinction between the two regime classifications. At the end of the projection horizon, the sensitivity of inflation expectations to debt surprises appears to be greater for inflation-targeting advanced economy central banks.

(Basis points, annual rate)

Robustness

To further reduce the possibility that the estimated effects on expectations are driven by fiscal demand stimuli, we control for the fiscal stance. Specifically, we include the change in the primary balance as a percentage of GDP and the change in the output gap as additional controls in the regression specifications that condition on state variables. These are jointly meant to capture the fiscal impulse, which might have implications for aggregate demand. Our results remain quantitatively and qualitatively robust to introducing these controls (Graph 9).

Response of Inflation Expectations to Government Debt Surprises, Emerging Market Economies – Conditional on the Fiscal Stance

Graph 9

Notes: t=0 is the year of the shock. The graphs plot for the relevant state variable the inflation expectations response to a 10 percent surprise in the debt-to-GDP ratio. Included as controls are the change in the primary balance to GDP ratio as well as the change in a measure of the output gap. The red (blue) dot corresponds to the estimates evaluated at high (low) debt and inflation levels, with the whiskers representing 90 percent confidence intervals.

The main analyses of the paper use government debt to GDP levels as a proxy for debt sustainability concerns. While this is a reasonable proxy, debt capacity varies across counties, and the mapping from debt levels to default risk is not one-to-one. As a robustness exercise, we use the Financial Risk Rating (FRR) of the International Country Risk Guide (ICRG) instead of government debt levels to proxy for debt sustainability concerns. Specifically, we replace ????????௜,௧ିଵ in equation (2) with the inverse of the risk rating for country ?? at time ?? − 1 and higher levels indicate higher risk (??????௜,௧ିଵ). The FRR is an index that attempts to measure a country’s “ability to finance its official, commercial, and trade debt obligations” (ICRG Method 2022). Its components include foreign debt as a percentage of GDP, foreign debt service as a share of exports, the current account as a share of exports, reserves as a share of imports, and currency depreciation.

As was the case with government debt, we find that the sensitivity of inflation expectations to government debt surprises rises with the level of the ICRG risk measure (Table 9). Graph 10 displays the response of 5-year ahead inflation expectations to government debt surprises when the conditioning variable is the ICRG risk measure. To highlight the impact of the conditioning variable, we again show the coefficients of the local projections for the highest risk decile (90th percentile) and the lowest risk decile (10th percentile). After the first year, the impact of debt shocks on long-term inflation expectations is significantly higher for the highest risk decile.

Response of Inflation Expectations to Government Debt Surprises, Emerging Market Economies – ICRG Financial Risk Rating

(Basis points, annual rate)

Graph 10

Notes: t=0 is the year of the shock. The graph plots for the 10th and 90th percentiles of the ICRG Financial Risk Rating, the inflation expectations response to a 10 percent surprise in the debt-to-GDP ratio. The red (blue) dot corresponds to estimates evaluated at a high (low) financial risk rating, with the whiskers representing 90 percent confidence intervals.

As noted earlier, government debt forecast errors may not represent true surprises, because they may be forecastable. We follow the method used in IMF (2017) and Magud and Pienknagura (2022) in regressing the debt forecast errors on forecast errors of inflation and real GDP growth and using the residuals from this regression as our debt shock. We find that the results are qualitatively similar, and that the peak estimated sensitivities retain statistical significance and are of slightly larger magnitude than in the baseline specification. This might be due to downward biased estimates in the baseline results from the disinflationary effects of adverse GDP growth shocks. That is, a negative GDP growth surprise that would imply a larger debt-to-GDP likely lowers inflation and inflation expectations. We also follow the approach of Auerbach and Gorodnichenko (2013) and project the debt forecast errors on lagged macroeconomic variables, using the residuals from this regression as the debt shock. The results and conclusions discussed above are quantitatively and qualitatively robust to this approach as well.

Regression results, by ICRG financial risk rating, Emerging Market Economies

	(1) h = 4	(2) h = 10	(3) h = 16	(4) h = 22	(5) h = 28	(6) h = 34
Debt Surprise	0.0493 (0.0363)	-0.0043 (0.0528)	-0.1042 (0.0634)	-0.2697*** (0.0895)	-0.2357*** (0.0821)	-0.0217 (0.0440)
Financial Risk Rating	13.1319 (19.8761)	22.8277 (25.9743)	23.5666 (25.4310)	18.5699 (42.7289)	-4.7757 (44.8312)	-39.6436 (27.5824)
Debt Surprise # ICRG FRR	-1.7191	0.8848	5.2006**	11.8482***	9.9753***	1.5808
No. of Obs.	(1.3463) 342	(1.9772) 280	(2.4406) 279	(3.6415) 252	(3.2268) 251	(1.6284) 232
Adjusted R sq.	0.63	0.40	0.20	0.28	0.15	0.03

Conclusion

We have explored a new way of assessing how concerns about government debt levels can affect inflation expectations, based on unanticipated changes in government debt. Our findings that debt surprises raise long-term inflation expectations in Emerging Market Economies, particularly when initial debt levels are already high, when inflation is already elevated, and when monetary policy frameworks are weaker, have important policy implications.

In Emerging Market Economies with high government debt levels, bringing them to a sustainable path is likely to be important for containing inflation. This is particularly relevant at the current juncture, after a sharp rise indebtedness around the COVID-19 shock, and with inflationary pressures around the world having proven strong and persistent. In the medium term, adopting modern, forward-looking monetary policy frameworks such as inflation targeting can reduce inflationary concerns associated with government debt, creating more space for both monetaryand fiscal policy.

Further research could quantify the degree to which debt concerns affect the inflationary process itself through the expectations channel. Moreover, whereas in this study we have focused on professional forecasters’ expectations, it may be worth investigating further the degree to which households’ and firms’ expectations are also influenced by government debt levels.

Abiad, A., Furceri, D. and Topalova, P., 2016. “The macroeconomic effects of public investment: Evidence from advanced economies.” Journal of Macroeconomics, 50, pp.224-240.

Aiyagari, S. R., and Gertler, M., 1985. “The Backing of Government Bonds and Monetarism.” Journal of Monetary Economics, Vol.16 (1), pp. 19-44.

Arellano, C., Bai, Y., and Mihalache, G. P., 2020. “Monetary Policy and Sovereign Risk in Emerging Economies” (NK-Default), NBER Working Paper 26671, (Cambridge, National Bureau of Economic Research).

Auerbach, A. J. and Gorodnichenko, Y., 2012. “Measuring the Output Responses to Fiscal Policy.” American Economic Journal: Economic Policy, Vol. 4 (2), pp. 1-27.

(2013), “Output Spillovers from Fiscal Policy.” American Economic Review, Vol. 103(3), pp.141-146.

Banerjee, R., V. Boctor, A. Mehrotra, and F. Zampolli, 2023a. “Fiscal Deficits and Inflation Risks: The Role of Fiscal and Monetary Regimes.” BIS Working Paper No. 1028.

Banerjee, R., V. Boctor, A. Mehrotra, and F. Zampolli, 2023b. “Fiscal Sources of Inflation Risk in EMDEs: the Role of the External Channel.” BIS Working Paper No. 1110.

Beck, N. and J. N. Katz, 2011. “Modeling Dynamics in Time-Series–Cross-Section Political Economy Data.” Annual Review of Political Science 14(1), pp. 331-352.

Benigno, P., 2020. “A Central Bank Theory of Price Level Determination.” American Economic Journal: Macroeconomics, Vol. 12(3), pp. 258-83.

Bianchi, F., Faccini, R. and Melosi, L. 2022. “A Fiscal Theory of Trend Inflation.” NBER Working Paper 30727, (Cambridge, National Bureau of Economic Research).

Bank for International Settlements (BIS), 2019. “Monetary Policy Frameworks in EMEs: Inflation Targeting, the Exchange Rate and Financial Stability,” BIS Annual Economic Report, 31–53, (Basel, Bank for International Settlements).

Blanchard, O. and Perotti, R., 2002. “An Empirical Characterization of the Dynamic Effects of Changes in Government Spending and Taxes on Output.” The Quarterly Journal of Economics, Vol. 117(4), pp.1329-1368.

Bohn, H., 1988, “Why Do We Have Nominal Government Debt?” Journal of Monetary Economics, Vol. 21 (1), pp. 127-140.

Borio, C., M. Farag, and F. Zampolli, 2023. “Tackling the Fiscal Policy-Financial Stability Nexus,” BIS Working Paper No. 1090.

Buiter, W., and Grafe, C., 2001. “No Pain, No Gain? The Simple Analytics of Efficient Disinflation in Open Economies,” CEPR Discussion Paper 3038, (Washington, Center for Economic and Policy Research).

Canzoneri, M., Cumby, R. E., and Diba, B.T., 2001. “Is the Price Level Determined by the Needs of Fiscal Solvency?” American Economic Review, Vol. 91 (5), pp. 1221-1238.

Calvo, G., 1988. “Servicing the Public Debt: The Role of Expectations.” American Economic Review Vol. 78 (4), pp. 647-661.

Catão, L., and Terrones, M., 2005. “Fiscal Deficits and Inflation.” Journal of Monetary Economics, 2005, Vol. 52 (3), pp. 529-554.

Celasun, O., Gelos, G., and Prati, A., 2004. “Obstacles to Disinflation: What is the Role of Fiscal Expectations?” Economic Policy, Volume 19, Issue 40, October, pp. 442–481.

Celasun and others, 2021 for GDP forecasts

Chen, J., Gornycka, L., and Zdarek, V., 2022. “Biases in Survey Inflation Expectations: Evidence from the Euro Area,” IMF Working Paper No. 2022/205, (Washington, International Monetary Fund).

Christiano, L., and Fitzgerald, T., 2000. “Understanding the Fiscal Theory of the Price Level,” Economic Review, Issue Q II, pp. 2-38

Cochrane, J., 2001. “Long‐Term Debt and Optimal Policy in the Fiscal Theory of the Price Level,” Econometrica, Vol. 69 (1), pp. 69-116.

Cochrane, J., 2005. “Money as Stock.” Journal of Monetary Economics, Vol. 52(3), pp. 501-528.

Cochrane, J., 2023. The Fiscal Theory of the Price Level, (Princeton, NJ, Princeton University Press).

Coibion, O. and Gorodnichenko, Y., 2012. “What Can Survey Forecasts Tell Us About Information Rigidities?” Journal of Political Economy, Vol. 120(1), pp.116-159.

Coibion, O., Gorodnichenko, Y. and Weber, M., 2021. “Fiscal Policy and Households’ Inflation Expectations: Evidence From a Randomized Control Trial” NBER Working Paper No. w28485 (Cambridge, National Bureau of Economic Research).

de Mendonça, H. F., and Machado, M. R., 2013. “Public Debt Management and Credibility: Evidence from an Emerging Economy.” Economic Modelling, Vol. 30, pp. 10-21.

Dinçer, N., Eichengreen, B. and Geraats, P., 2019. “Transparency of monetary policy in the postcrisis world.” The Oxford handbook of the economics of central banking, 2019, p.287.

Dornbusch, R., 1982. “Stabilization Policies in Developing Countries: What Have We Learned?” World Development, Vol. 10 (9), pp. 701-708.

Driscoll, J. C., and Kraay, A. C., 1998. “Consistent Covariance Matrix Estimation with Spatially Dependent Panel Data,” Review of Economics and Statistics, Vol. 80(4), pp.549-560.

Frankel, M., Mauch, M. and Rülke, J.C., 2017. “Forecaster Rationality and Expectation Formation in Foreign Exchange Markets: Do Emerging Markets Differ from Industrialized Economies?” Working Paper 17/04, WHU - Otto Beisheim School of Management.

Flores, J. E., Furceri, D., Kothari, S. and Ostry, J.D., 2023. “Worse Than You Think: Public Debt Forecast Errors in Advanced and Developing Economies,” Journal of Forecasting, Vol. 42 (3), pp. 685-714.

Furceri, D. and Li, B.G., 2017. “The macroeconomic (and distributional) effects of public investment in developing economies.” IMF Working Paper No. 2017/217, (Washington, International Monetary Fund).

Gordon, D.B., and Leeper, E., 2002. “The Price Level, the Quantity Theory of Money, and the Fiscal Theory of the Price Level,” NBER Working Paper 9084 (Cambridge, National Bureau of Economic Research).

Grigoli, F. and D. Sandri, 2023. “Public Debt and Household Inflation Expectations.” BIS Working Paper No. 1082

International Monetary Fund (IMF), 2017. “World Economic Outlook”, International Monetary Fund.

, 2021. “Review of the Debt Sustainability Framework for Market Access Countries,” International Monetary Fund.

_, 2022. “Fiscal Monitor”, International Monetary Fund.

Jordà, Ò., 2005. “Estimation and Inference of Impulse Responses by Local Projections.” American Economic Review, Vol. 95(1), pp.161-182.

Kwon, G., McFarlane, L., and Robinson W., 2009. “Public Debt, Money Supply, and Inflation: A Cross-Country Study.” IMF Working Paper No 2006/121 (Washington, International Monetary Fund).

Landau, J. P., 2020. “Money and Debt: Paying for the Crisis.” VoxEU, June 23 (Vox EU, CEPR).

Leeper, E.M., 1991. “Equilibria Under ‘Active’ and ‘Passive’ Monetary and Fiscal Policies.” Journal of Monetary Economics, Vol. 27(1), pp.129-147.

Loyo, E. 1999. “Tight Money Paradox on the Loose: A Fiscalist Hyperinflation.” Unpublished manuscript, John F. Kennedy School of Government, Harvard University

Magud, M.N.E. and Pienknagura, S., 2022. “The Return of Expansionary Austerity: Firms’ Investment Response to Fiscal Adjustments in Emerging Markets,” IMF Working Paper 2022/070 (Washington, International Monetary Fund).

Nickell, S., 1981. “Biases in dynamic models with fixed effects.” Econometrica, Vol. 49 (1), pp.1417-1426.

Nunes, R., Ozdagli, A.K., and Tang, J., 2022. “Interest Rate Surprises: A Tale of Two Shocks.” Working Paper 22/2 Federal Reserve Bank of Boston, (Boston, Federal Reserve Bank of Boston).

Panizza, U., and Taddei, F., 2020. “Local Currency Denominated Sovereign Loans: A Portfolio Approach to Tackle Moral Hazard and Provide Insurance,” IHEID Working Papers 09-2020, Economics Section (Geneva, Graduate Institute of International and Development Studies)

Rodriguez, C.A.,1982; ‘The Argentine stabilization plan of December 20th’, World Development, Vol. 10, Issue 9, pp. 801-811.

Sargent, T.J., and Wallace, N., 1981. “Some Unpleasant Monetarist Arithmetic.” Federal Reserve Bank of Minneapolis Quarterly Review 5, No. 3: 1–17, (Minneapolis, Federal Reserve Bank of Minneapolis)

Schmitt-Grohe, S., and Uribe, M., 2000. “Price Level Determinacy and Monetary Policy Under a Balanced-Budget Requirement,” Journal of Monetary Economics, Vol. 45 (1), pp. 211-246

Singh, A., and others 2005. “Stabilization and Reform in Latin America: A Macroeconomic Perspective on the Experience since the Early 1990s”, IMF Occasional Paper 238 (Washington, International Monetary Fund).

Sims, C., 1994. “A Simple Model for Study of the Determination of the Price Level and the Interaction of Monetary and Fiscal Policy.” Economic Theory, Vol. 4(3), pp. 381- 999.

Sunder-Plassmann, L., 2020. “Inflation, Default and Sovereign Debt: The Role of Denomination and Ownership.” Journal of International Economics, Vol. 127, 103393.

Unsal, D. F., Papageorgiou, C., and Garbers, H., 2022, “Monetary Policy Frameworks: An Index and New Evidence,” IMF Working Paper WP/22/22, (Washington, International Monetary Fund).

Van Bon, N. 2015. “The Relationship Between Public Debt and Inflation in Developing Countries: Empirical Evidence Based on Difference Panel GMM.” Asian Journal of Empirical Research, Vol. 5, No. 9, pp. 128–42.

Woodford, M., 1994. “Monetary Policy and Price Level Determinacy in a Cash-in-Advance Economy,” Economic Theory, Springer; Society for the Advancement of Economic Theory (SAET), Vol. 4(3), pp. 345-380.

_, 2001; "Fiscal Requirements for Price Stability," Journal of Money, Credit and Banking, Blackwell Publishing, Vol. 33(3), pp. 669-728.

Country list
Advanced Economies		Emerging & Developing Economies
Australia	Korea	Albania India	Developing Countries Bangladesh
Austria	Latvia	Argentina	Indonesia Honduras
Belgium	Lithuania Netherlands	Armenia	Kazakhstan Moldova
Canada	New Zealand	Belarus Malaysia	Nicaragua
Cyprus	Norway	Bolivia Mexico	Nigeria
Czech Republic	Poland	Bosnia and	North Macedonia Uzbekistan
Denmark	Portugal	Herzegovina Pakistan	Vietnam
Estonia	Singapore	Brazil Panama
Finland	Slovak Republic	Bulgaria Paraguay
France	Slovenia	Chile Peru
Germany	Spain	China	Philippines
Greece	Sweden	Colombia	Romania
Hong Kong SAR	Switzerland	Costa Rica Russia
Ireland	Taiwan Province of China	Croatia	Saudi Arabia
Israel	United Kingdom	Dominican Republic Serbia
Italy	United States	Ecuador	South Africa
Japan		Egypt Sri Lanka
		El Salvador Thailand
	Georgia	Turkiye
		Guatemala Turkmenistan
		Hungary Ukraine

5-Year Inflation Expectations, by Inflation Level, Advanced Economies

	(1) h=4	(2) h = 10	(3) h= 16	(4) h = 22	(5) h = 28	(6) h = 34
Debt Surprise	0.0050 (0.0043)	0.0066* (0.0038)	0.0065 (0.0040)	0.0022 (0.0049)	-0.0085 (0.0061)	-0.0079* (0.0045)
CPI, y-o-y	0.0217**	0.0463*** (0.0101)	0.0640*** (0.0123)	0.0662*** (0.0178)	0.0323* (0.0173)	0.0458*** (0.0145)
Debt Surprise # Inflation	(0.0110) -0.0010	-0.0041***	-0.0023	0.0000	0.0049*	0.0033*
No. of Obs.	(0.0017) 381	(0.0010) 349	(0.0015) 349	(0.0023) 314	(0.0027) 314	(0.0018) 293
Adjusted R sq.	0.33	0.13	0.02	-0.02	-0.07	-0.08

5-Year Inflation Expectations, by Financial Risk Rating, Advanced Economies

	(1)	(2)	(3)	(4)	(5)	(6)
	h = 4	h = 10	h = 16	h = 22	h = 28	h = 34
Debt Surprise	-0.0139	-0.0209	0.0020	-0.0227	-0.0335**	-0.0120
	(0.0113)	(0.0213)	(0.0201)	(0.0280)	(0.0163)	(0.0248)
Financial Risk Rating	8.8700*	-2.4138	6.9317	-9.0687	-2.9544	-14.0266*
	(4.6527)	(8.5152)	(6.3806)	(8.9443)	(7.4589)	(8.0936)
Debt Surprise # ICRG FRR	0.7321*	0.8730	0.0575	1.0676	1.4386**	0.4676
	(0.4136)	(0.8052)	(0.8014)	(1.1156)	(0.5765)	(0.9031)
No. of Obs.	371	340	340	306	306	286
Adjusted R sq.	0.26	0.07	-0.09	-0.11	-0.13	-0.13

Regression Results, 3-Year Inflation Expectations, Full Sample

	(1) h = 4	(2) h = 10	(3) h = 16	(4) h = 22	(5) h = 28	(6) h = 34
Debt Surprise	-0.0024	0.0143*	0.0246**	0.0436***	0.0328***	0.0217***
No. of Obs.	(0.0057) 780	(0.0081) 673	(0.0114) 673	(0.0130) 608	(0.0102) 608	(0.0081) 564
Adjusted R sq.	0.61	0.45	0.23	0.21	0.20	0.15

Notes: All regressions include 2-month and 8-month lags of 3-year ahead inflation expectations, a 12-month lag of the debt shock, and country and time fixed effects. Driscoll-Kraay standard errors in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01

Regression Results, 3-Year Inflation Expectations, Emerging Market Economies

	(1) h = 4	(2) h = 10	(3) h = 16	(4) h = 22	(5) h = 28	(6) h = 34
Debt Surprise	-0.0050	0.0280**	0.0479***	0.0800***	0.0637***	0.0530***
	(0.0111)	(0.0128)	(0.0173) 281	(0.0189) 253	(0.0137)	(0.0170)
No. of Obs. Adjusted R sq.	354 0.61	281 0.47	0.25	0.26	253 0.23	233 0.17

Regression Results, 3-Year Inflation Expectations, Advanced Economies

	(1) h = 4	(2) h = 10	(3) h = 16	(4) h = 22	(5) h = 28	(6)
Debt Surprise	-0.0008	-0.0003	0.0006	0.0022	-0.0016	h = 34 0.0005
	(0.0017)	(0.0030)	(0.0036)	(0.0034)	(0.0032)	(0.0037)
No. of Obs. Adjusted R sq.	426 0.39	392 0.13	392 -0.01	355 -0.04	355 -0.08	331 -0.13

(Basis points, annual rate)
Regression Results, 3-Year Inflation Expectations, By Initial Debt Level, Emerging Market Economies
Table A.7

	(1)	(2)	(3)	(4)	(5)	(6)
	h = 4	h = 10	h = 16	h = 22	h = 28	h = 34
Debt Surprise	0.0411	0.0228	0.0231	-0.0341	-0.0404	0.0575*
	(0.0323)	(0.0308)	(0.0422)	(0.0518)	(0.0459)	(0.0330)
Debt to GDP	-0.0133**	-0.0112	-0.0239***	-0.0283**	-0.0299**	-0.0217**
	(0.00588)	(0.00794)	(0.00823)	(0.0111)	(0.0128)	(0.0110)
Debt Surprise # Debt	-0.000955	0.000122	0.000506	0.00252*	0.00220*	-0.000198
	(0.000788)	(0.000765)	(0.00107)	(0.00131)	(0.00116)	(0.000674)
No. of Obs.	335	277	277	249	249
Adjusted R sq.	0.62	0.47	0.26	0.31	0.27	229 0.17

Response of 3-Year Inflation Expectations to Public Debt Shocks, Baseline

Graph A.1

Notes: t=0 is the quarter of the shock. The graphs plot for the relevant horizon the 3-year ahead inflation expectations response to a 10 percent surprise in the debt-to-GDP ratio. The blue dots denote the inflation expectations response for Emerging Market Economies in our sample while the red dots denote the corresponding response for advanced economies. The whiskers represent 90 percent confidence intervals. The chart on the left shows the response for the full sample.

(Basis points, annual rate)

Response of 3-Year Inflation Expectations to Public Debt Shocks, by Initial Debt Level Classification

Graph A.2

Notes: t=0 is the year of the shock. The graphs plot for the relevant country groups the 3-year ahead inflation expectations response to a 10 percent surprise in the debt-to-GDP ratio. The blue dots denote the inflation expectations response for economies with a debt level at the 10th percentile of the relevant country group sample, while the red dots denote the corresponding response for economies with a debt level at the 90th percentile of the relevant country group sample. The whiskers represent 90 percent confidence intervals.

Response of 3-Year Inflation Expectations to Public Debt Shocks, by Initial Inflation Level Classification

(Basis points, annual rate)

Graph A.3

Notes: t=0 is the year of the shock. The graphs plot for the relevant country groups the 3-year ahead inflation expectations response to a 10 percent surprise in the debt-to-GDP ratio. The blue dots denote the inflation expectations response for economies with an initial inflation at the 10th percentile of the relevant country group sample, while the red dots denote the corresponding response for economies with an initial inflation level at the 90th percentile of the relevant country group sample. The whiskers represent 90 percent confidence intervals.

Response of 3-Year Inflation Expectations to Public Debt Shocks, Emerging Market Economies – by Initial FX Share of Government Debt

(Basis points, annual rate)

Graph A.4
Notes: t=0 is the year of the shock. The graphs plot for the relevant country group the 3-year ahead inflation expectations response to a 10 percent surprise in the debt-to-GDP ratio. The blue dots denote the inflation expectations response for economies with an initial FX share of government debt beneath the 10th percentile, while the red dots denote the corresponding response for economies with an initial FX share of government debt above the 90th percentile. The whiskers represent 90 percent confidence intervals.

Previous volumes in this series

1140 November 2023	Relationship discounts in corporate bond trading	Simon Jurkatis, Andreas Schrimpf, Karamfil Todorov, Nicholas Vause
1139 October 2023	A journal ranking based on central bank citations	Raphael Auer, Giulio Cornelli and Christian Zimmermann
1138 October 2023	Dealer capacity and US Treasury market functionality	Darrell Duffie
1137 October 2023	International portfolio frictions	Wenxin Du
1136 October 2023	Expectations and the neutrality of interest rates	John Cochrane
1135 October 2023	Artificial intelligence, services globalisation and income inequality	Giulio Cornelli, Jon Frost and Saurabh Mishra
1134 October 2023	Bank competition, cost of credit and economic activity: evidence from Brazil	Gustavo Joaquim, Bernardus van Doornik and José Renato Haas Ornelas
1133 October 2023	Remote work and high-proximity employment in Mexico	Lorenzo Aldeco Leo and Alejandrina Salcedo
1132 October 2023	Pandemic-induced increases in container freight rates: assessing their domestic effects in a globalised world	José Pulido
1131 October 2023	System-wide dividend restrictions: evidence and theory	Miguel Ampudia，Manuel Munoz, Frank Smets and Alejandro Van der Gothe
1130 October 2023	What can 20 billion financial transactions tell Pongpitch Amatyakul， Panchanok us about the impacts of COVID-19 fiscal Jumrustanasan,	Pornchanok Tapkham
1129 October 2023	transfers? Big techs in finance	Sebastian Doerr，JonFrost, Leonardo Gambacorta， Vatsala
1128	The cumulant risk premium	Shreeti Albert S. (Pete) Kyle and Karamfil Todorov
October 2023 1127 September 2023	The impact of green investors on stock prices Gong Cheng，Eric Jondeau， Benoit	Mojon and Dimitri Vayanos

All volumes are available on our website www.bis.org.

群内每日免费分享5份+最新资料

300T网盘资源+40万份行业报告

为您的创业、职场、商业、投资、亲子、网赚、艺术、健身、心理、个人成长 …… 全面赋能！

关注公众号获取更多资料

添加微信，备
注“入群”
立刻免费领取
200套知识地图+最新研报

致终身学习者社群

行业报告/思维导图/电子书/资讯情报