Disinflation and Monetary Transmission in Cyprus1
Inflation in Cyprus dropped in 2023 due to the diminishing impact of supply-side shocks and moderating demand. But some domestic price pressures persist, mostly from non-fiscal aggregate demand. ECB tightening has significantly impacted interest rates, including for outstanding mortgages. Deposit rates saw delayed and smaller increases, likely driven by high banking sector liquidity and low competition. Continued commitment to containing aggregate demand is supporting the final stage of disinflation.
A. Introduction
1. Like many other countries, Cyprus experienced a post-pandemic inflation surge. Inflation has been sensitive to energy price pressures following Russia’s war in Ukraine and the strong rebound in tourism after the pandemic (Beyer 2023). Headline inflation peaked at over 10 percent in July 2022. While initially driven by external shocks, inflation pressures became more broad-based over time. Robust wage growth also drove up core inflation, which surpassed 6 percent in July 2022.
Different Inflation Measures in Cyprus Over Time
Citation: IMF Staff Country Reports 2024, 138; 10.5089/9798400276439.002.A001
Sources; burostat and IMF staff calculations.2. Inflation declined markedly in 2023. By the end of the year, headline inflation had dipped below 2 percent, while core inflation declined to 2.4 percent (Figure 1). Last year’s inflation dynamics compared favorably to those in the euro area, where in December headline and core HICP inflation still stood at 2.9 percent and 3.4 percent, respectively. The strong decline in headline inflation in Cyprus has been supported by energy price normalization.
3. Some domestic price pressures remain. Inflation of goods and services with low import content, the so called LIMI inflation index, is a good measure of domestic price pressures (Fröhling, O’Brien, and Schaefer 2022). It has been stickier than core HICP inflation and remained above 4 percent at the end of last year (Figure 1). LIMI inflation excluding restaurants and accommodation (for which second-round effects play an important role), provides an even narrower measure of domestic price pressures. It remained stable during the pandemic, when other prices fell, and throughout 2021, when other prices surged. However, it started picking up slowly in 2022 and has been close to 3 percent since mid-2022, also indicating remaining internal price pressures.
4. Against this backdrop, this Selected Issue Paper analyzes the drivers of disinflation and discusses implications for the inflation outlook. Section B analyzes the role of supply and demand, and Section C differentiates fiscal from non-fiscal demand. Section D considers the role of unit profits and labor costs. Section E discusses monetary transmission to lending rates, a key driver of weaker (non-fiscal) demand pressures. Section E uncovers weaker pass-through to deposit rates and touches upon reasons and implications. Section F discusses policy conclusions.
B. Supply and Demand Drivers of Inflation
5. Inflation dynamics can be decomposed into supply and demand drivers. Bayesian Vector-Error-Correction models (BVARs) with output growth and inflation—in which aggregate supply shocks move prices and quantities in opposite directions and aggregate demand shocks move them in the same direction—allow for such a decomposition (see Appendix I for more details). The cumulative effect of demand and supply shocks, and their lagged effects on inflation, can then be investigated with a historical decomposition of excess inflation (i.e., deviation of inflation from target).2 While such a simple BVAR model can provide useful insights, it has shortcomings. First, such models can be sensitive to model specification. In addition, it can be difficult to properly identify supply and demand shocks. This is especially the case for headline inflation and when supply and demand shocks hit at the same time, as arguably was the case during the pandemic and subsequent recovery. We hence restrict this analysis to core inflation.3
6. The analysis suggests that high core inflation in 2023 was driven both by demand and supply factors (Figure 2). The post-pandemic inflation surge is attributed to both supply and demand factors, with the latter dominating most of the time. Supply factors peaked in 2022Q3— contributing 2.4 percentage points to core inflation—and then declined, but still accounted for over a third of excess core inflation in 2023 (Figure 2, left panel). Demand factors, while also declining since the second half of 2022, have been sticker, even though their remaining quarter-over-quarter impact was tiny at the end of the year (Figure 2, right panel). The significant impact of demand aligns with a positive output gap estimate (IMF 2024b) and lingering effects of monetary and fiscal policy easing in response to the pandemic.
Supply and Demand Drivers of Inflation
Citation: IMF Staff Country Reports 2024, 138; 10.5089/9798400276439.002.A001
7. This finding of the importance of demand factors aligns with a bottom-up decomposition of core inflation developed during the COVID-19 pandemic. HICP core components can be classified into four distinct groups at the 2-digit level: 1) those sensitive to supply-chain disruptions; 2) those influenced by re-opening dynamics, indicating pent-up demand; 3) rent; and 4) miscellaneous items (Gonçalves and Koeste 2022). We aggregate the components in each category using the corresponding HICP weights. The classification is kept invariant during the sample period and does not directly account for second-round effects from energy prices. In line with previous results, the contribution from supply disruptions declined during 2023 and were very small at the end of the year (Figure 3). As before, demand factors have been around twice as large on average last year, with supply disruptions still accounting for around a third of 2023 excess core inflation.
C. Fiscal and Non-Fiscal Demand Pressures
8. The theoretical links of fiscal expansion and inflation are well established. Expansionary fiscal policy tends to be inflationary through the textbook Keynesian channels of private consumption and investment. This relationship is also inherent in a typical downward sloping Phillips curve. If expansionary fiscal policy raises inflation expectations, the impact on inflation is amplified. The relationship also holds in Representative Agent New Keynesian models, in which fiscal policy shocks move output and inflation in the same direction. In the Euro Area, model-based simulations suggest that a uniform cut of public expenditure by 1 percent of GDP across member states is equivalent to a monetary policy tightening of around 50 basis points, both reducing inflation by around 0.2 percentage points (Beyer et al. 2023).
9. As other countries, Cyprus employed fiscal measures to mitigate the impact of the pandemic and energy price shock. Cyprus has maintained fiscal discipline since the 2013 crisis, resulting in fiscal space that enabled Cyprus to mount an effective policy response via temporary and mostly targeted support measures (IMF 2023a). The fiscal balance turned strongly negative in 2020 and 2021 but returned to a surplus in 2022.
10. The contribution of fiscal policy shocks to inflation can be assessed empirically. To do so, we estimate a similar BVAR model as before. It now includes four variables: output growth, inflation, the short-term interest rate, and the fiscal balance. A contractionary (expansionary) fiscal shock is restricted to lead to an increase (decrease) in the primary balance and a negative (positive) impact on output growth and inflation (see Appendix II for more details). This specification can capture the response of inflation to a fiscal shock, and hence be used to decompose excess inflation into fiscal and non-fiscal drivers (again with a historical decomposition). Importantly, fiscal policy in the model also responds endogenously to supply and non-fiscal demand shocks (i.e., to the business cycle). The contribution of fiscal policy shocks to excess inflation is hence always in addition to how fiscal policy usually reacts to the other shocks (i.e., in addition to the estimated endogenous response).4 In addition, the fiscal balance only captures the costs of the fiscal measures.5 Concerns about model stability and proper shock identification become more severe as models gets more complex. We hence confirm the robustness of the main findings with respect to model specification (see Appendix D).
Bottom-up Decomposition of Core Inflation
Citation: IMF Staff Country Reports 2024, 138; 10.5089/9798400276439.002.A001
Sources: Eurostat and IMF staff calculations.11. By the end of 2023, demand pressures were likely dominated by non-fiscal factors. For 2022, our estimation attributes around 1.2 percentage points of core inflation to fiscal policy (Figure 4, left panel).6 However, this contribution declined in 2023 and dropped to 0.4 percentage points in 2023Q4, even declining slightly quarter-over-quarter (Figure 4, right panel). A corresponding partial yet substantial reduction in non-fiscal demand pressures has likely been supported by monetary tightening, discussed in Section F.
Fiscal and Non-Fiscal Drivers of Inflation
Citation: IMF Staff Country Reports 2024, 138; 10.5089/9798400276439.002.A001
D. The Role of Profits and Labor Costs
12. Persistent inflation can also arise from conflicts over relative prices between firms and workers. Due to nominal rigidities, particularly with wages being stickier than prices, cost-push shocks can lead to persistent inflation if firms and workers sequentially adjust prices and wages. For instance, following an import price shock, increased profit shares from initial price hikes may subsequently drive workers to demand higher nominal wage increases to restore their purchasing power (Blanchard 1986; Lorenzoni and Werning 2023).
13. Increasing prices can consequently reflect higher profits or higher labor compensation per unit of GDP. From the income side, GDP is the sum of labor compensation, gross operating surplus, and net taxes:
GDP = GVA + Net Taxes = Profits + Compensation of Employees + Net Taxes
Based on this identity, the GDP deflator (nominal GDP divided by real GDP) can hence be expressed as the sum of unit labor costs, unit profits, and unit net taxes:7
GDP/Real GDP = GDP Deflator = Unit Profits + Unit Labor Costs + Unit Taxes
14. So far, inflation has been accompanied by higher unit profits and lower labor costs. Since 2021, unit profits increased while unit labor costs declined (Figure 5, left panel). In 2022, nearly three-quarters of inflation was driven by unit profits. The remaining portion came from higher unit taxes, with a negligible contribution from unit labor costs. Following the spike in energy prices, firms have hence passed on more than just the immediate nominal import cost shock. The increasing profit share implies that firms have so far been relatively more shielded from inflation than wage earners. Across the euro area, the increase in nominal profits was largest in sectors benefiting from rising international commodity prices and those exposed to recent supply-demand mismatches (Hansen, Toscani, and Zhou 2023). Last year, unit profits and unit labor costs contributed equally.
Unit Profits, Labor Costs and GDP Deflator Decomposition
Citation: IMF Staff Country Reports 2024, 138; 10.5089/9798400276439.002.A001
15. Decomposing the consumption deflator introduces a role for import prices and reinforces previous findings. The consumption deflator, like HICP but unlike GDP, includes imports but excludes exports. The consumption deflator can hence provide an additional perspective on inflation, particularly regarding the post-pandemic inflation surge. By employing a set of simplifying assumptions, the deflator can be dissected into contributions from domestic profits, labor, foreign factors, and net taxes (Hansen, Toscani, and Zhou 2023). In 2022, foreign factors contributed significantly, though less than domestic profits (Figure 6). Consequently, domestic profits remain a significant driver of inflation, even after explicitly considering the import price shock. Consistent with the decomposition of the GDP deflator above, unit labor costs started contributing last year.
Consumption Deflator Decomposition
Citation: IMF Staff Country Reports 2024, 138; 10.5089/9798400276439.002.A001
Sources: EurostatOECD, IMF staff calculations.E. Monetary Tightening and Lending Rates
16. In response to the inflation surge, the ECB tightened its monetary policy stance. The ECB started a tightening cycle in July 2022 and raised its policy rate from 0 to 4.5 percent. This has been the steepest tightening since the inception of the euro by far. Not surprisingly, bank retail rates increased as well (Figure 7). The ‘pass-though’ of changes in policy rates to bank retail rates is a crucial intermediate step in the transmission of monetary policy to aggregate demand.
Policy and Bank Retail Interest Rates Over Time
Citation: IMF Staff Country Reports 2024, 138; 10.5089/9798400276439.002.A001
17. The pass-through to bank retail rates can be assessed with so-called beta coefficients. They are defined as the ratio of the cumulative change in bank retail rates to the cumulative change in the policy rate. To assess pass-through during a tightening cycle, it is useful to compute beta coefficients for different bank retail rates for both households and non-financial corporations (NFCs).8
18. The pass-through to new lending rates in Cyprus has been comparable to that in other European and Euro Area countries. Although quite similar, there has been a somewhat weaker pass-through to new loans for NFCs. The pass-through to consumption and mortgages, in contrast, has been somewhat stronger than in other European and euro area economies (Figure 7, left panel).
19. The pass-through to outstanding mortgages has been much stronger. For mortgages, which usually have a longer maturity than other loans, we also assess the pass-through to the outstanding stock. Due to the high share of mortgages in household debt, pass-through to outstanding mortgages plays an important role in the transmission of monetary policy to output and prices (IMF 2024a). This pass-through has been more than twice as strong as in Europe and the euro area (Figure 8, left panel).
Policy Pass-Through to Lending Rates and Flexible Rates Mortgages
Citation: IMF Staff Country Reports 2024, 138; 10.5089/9798400276439.002.A001
20. The high share of flexible rate mortgages plays a key role. Cyprus has among the highest share of flexible rate mortgages in Europe, and—in contrast to most other countries—the share has not declined over the last decade (Figure 8, right panel).9 This explains the higher pass-through than elsewhere to both new and outstanding mortgages. New flexible-rate mortgages are more closely related to short-term interest rates than fixed-rate, and outstanding flexible-rate mortgages are adjusted more often than other types of mortgages.
21. Despite a high share of flexible rate mortgages, only half of the policy rate change has been passed on to outstanding mortgages. Possible reasons include constraints on mortgage rate adjustments (interest rate collars) and the reference to a bank-specific base rate, tied to the deposit rate, which exhibited a notably weaker pass-through (see next section).
22. The impact of higher mortgage rates on aggregate demand has been somewhat mitigated. Usually, higher mortgage rates lower disposable income (due to higher interest rate payments) and wealth (due to declining house prices). However, robust external demand for houses have so far—property sales increased in 2023, with foreigners accounting for close to half the demand—offset the impact of higher mortgage rates on house prices, thus weakening the wealth channel.
F. Monetary Tightening and Deposit Rates
23. Deposit rates have increased less than lending rates. Around two-thirds of the increase in policy rates has been passed on to new mortgage and consumption loan rates. In contrast, only a quarter has been passed on to household time deposits, and none to household overnight deposits. As a result, the share of time deposits in all deposits started increasing again (after falling strongly before). The deposit pass-through has been somewhat higher for corporations, likely reflecting more sophisticated financial planning and more bargaining power, but their loan rates increased by more as well.
24. Deposit pass-through has been lower and slower than in other countries, contrasting with the relatively similar pass-through to lending rates (Figure 9, left panel). Usually, bank retail rates react quickly to changes in the policy rate, with nearly all the transmission happening within the initial three months following a rate change (Beyer et al. 2024). In Cyprus, however, rates on time deposits continued increasing for much longer. Even so, the pass-through remained much lower, even by the end of 2023 (Figure 9, right panel).10
Policy Pass-Through to Deposit Rates
Citation: IMF Staff Country Reports 2024, 138; 10.5089/9798400276439.002.A001
25. Structural characteristics appear to have contributed to the limited and delayed deposit pass-through in Cyprus. Low competition and high banking sector liquidity can weaken pass-through to deposit rates (Beyer et al. 2024). Cross-country analysis reveals Cyprus as an outlier in both banking sector concentration (a measure of competition) and the loan-deposit-ratio (a measure of liquidity and lending opportunities), likely influencing the lower and slower pass-through to deposit rates than elsewhere (Figure 10). The impact of low deposit pass-through is likely to weaken transmission to output and prices, as it disincentivizes saving. However, this is mitigated by lower income-flows to those saving in deposits.
Policy Pass-Through and Structural Characteristics
Citation: IMF Staff Country Reports 2024, 138; 10.5089/9798400276439.002.A001
26. The heterogeneous lending and deposit pass-through increased interest rate margins, supporting bank profits. Net interest income increased strongly after the start of the monetary policy tightening cycle (Figure 11). The lower deposit pass-through added to higher revenue from reserves held at the ECB, pushing up bank profits further.
G. Policy Implications
27. The last mile of disinflation would benefit from containing aggregate demand. While supply disruptions are no longer materially impacting inflation, domestic demand continues to put pressure on prices. The remaining contribution from fiscal policy is small. Given a still positive output gap and robust growth projections (IMF 2024b), a neutral fiscal policy stance is appropriate.
28. Transmission through the real estate sector should not be hindered. Without monetary tightening and significant pass-through to mortgage rates (around half), last year’s increase in house prices likely would have been larger. However, robust external demand for housing and the Interest Rate Subsidy Schemes for real estate announced in May 2020 have weakened the wealth channel of monetary policy. Policy measures like a lower VAT for first time house buyers or interest rate subsidization schemes would further dampen transmission through the real estate market.
29. The low competition in the banking sector and its interaction with financial literacy warrant further analysis. For instance, the current situation could be primarily an endogenous or exogenous outcome. On one hand, the small size of the Cypriot economy in comparison to the Minimum Efficient Scale of banks may be a significant factor. On the other hand, anecdotal evidence about substantial rigidities in switching between banks point to the latter. In any case, improving financial literacy, enhancing consumer protection measures, and altering perceptions of risks associated with smaller banks and new market entrants (e.g., fintech companies from outside of Cyprus) could offer substantial benefits to consumers and foster competition. With possible further banking consolidation in the future, the roles of financial literacy and consumer protection would become even more crucial.
30. Wage dynamics will influence the inflation outlook. While risks of a wage-price spiral have declined substantially, the extent to which remaining demand pressures will impact future inflation will partly depend on wage dynamics. Wage pass-through of inflation tends to be higher in countries with higher union density (Baba and Lee 2022). In line with this, wage pass-through has been strong in Cyprus in the past (Beyer 2023) and has been driving inflation last year. Wage pressures could intensify further this year due to labor market tightness, indicated by a record-high high vacancy-to-unemployment ratio. An increase of labor supply—including from immigration or from increased participation—could help meeting employment needs. Nevertheless, in the current situation, it would be prudent for the public sector to lead the way by controlling public sector wage growth. Above all, the cost-of-living allowance should not be raised further.
Banking Sector Profitability Over Time
Citation: IMF Staff Country Reports 2024, 138; 10.5089/9798400276439.002.A001
Sources: EBA and IMF staff calculations.31. Elevated corporate profits could offer some room for absorbing potential wage increases. In the euro area, illustrative simulations show that a profit share compression to historic norms will likely be necessary to achieve timely disinflation under plausible wage growth assumptions (Hansen, Toscani, and Zhou 2023). Similarly, disinflation in Cyprus will likely need to go hand in hand with an increase of wages relative to profits and a normalization of the profit share.
Appendix I. Bayesian Vector Autoregression Model
1. A BVAR to decompose inflation into supply and demand drivers.
For the estimation of the BVAR model, we rely on a toolbox developed by the Fiscal Affairs Department of the IMF (Davoodi, Nguyen, and Poplawski-Ribeiro 2023). The BVAR model can be described as follows:
where Xt is a vector of n endogenous variables, q is the lag length, which we set to three; B0 represents deterministic terms; Bl is a matrix of parameters associated with the lagged variables; A0 is a matrix of parameters, capturing the contemporaneous relationships between the endogenous variables; and εt is a n x 1 vector of orthogonal structural shocks with a Gaussian distribution of mean zero and identity covariance matrix.
2. The reduced-form representation can be expressed as:
where
Cyprus: Sign Restrictions for Supply and Demand Shocks
Cyprus: Sign Restrictions for Supply and Demand Shocks
| Shock type | Output growth | Inflation |
|---|---|---|
| Demand | - (+) | - (+) |
| Supply | - (+) | + (-) |
Cyprus: Sign Restrictions for Supply and Demand Shocks
| Shock type | Output growth | Inflation |
|---|---|---|
| Demand | - (+) | - (+) |
| Supply | - (+) | + (-) |
3. We estimate this model with quarterly output growth and both HICP and core inflation from 1996Q1 to 2023Q4. All variables enter as differences of log levels (of output and price indices, respectively) and we estimate the mode both for y-o-y and q-o-q changes. The results presented above are for the former.
Appendix II. Bayesian Vector Autoregression Model with Fiscal Variables
1. The second model follows the same structure as the first model but includes four variables: output growth (difference of log GDP), inflation (difference of log price indices), the short-term interest rate, and the primary fiscal balance (seasonally adjusted as percent of GDP). It adds monetary and fiscal policy to the simple AD/AS framework of the first model (as in Bianchi and Melosi 2017). It hence can capture the response of output growth and inflation to fiscal shocks and can be used to decompose inflation into fiscal and non-fiscal drivers.
2. The fiscal shock is identified by restrictions on the sign of variables’ contemporaneous responses, with a contractionary (an expansionary) fiscal shock leads to an increase (a decrease) in the primary balance and has a negative (positive) impact on output growth and inflation. A negative (positive) non-fiscal aggregate demand shock decreases (raises) output growth and inflation, leading to a decline (an increase) in the primary balance-to-GDP (via automatic stabilizers):
3. Following Gambetti and Musso (2017), only n – 1 shocks are identified in system of n endogenous variables. This leaves one reduced-form residual shock unidentified. It captures the effects of omitted variables and other shocks, including monetary policy shocks.
Cyprus: Sign Restrictions for Fiscal and Non-Fiscal Shocks
Cyprus: Sign Restrictions for Fiscal and Non-Fiscal Shocks
| Shock type | Output growth | Inflation | Primary balance | Interest rate |
|---|---|---|---|---|
| Fiscal | - (+) | - (+) | + (-) | |
| Demand | - (+) | - (+) | - (+) | - (+) |
| Supply | - (+) | + (-) |
Cyprus: Sign Restrictions for Fiscal and Non-Fiscal Shocks
| Shock type | Output growth | Inflation | Primary balance | Interest rate |
|---|---|---|---|---|
| Fiscal | - (+) | - (+) | + (-) | |
| Demand | - (+) | - (+) | - (+) | - (+) |
| Supply | - (+) | + (-) |
4. We again estimate this model with quarterly data from 1996Q1 to 2023Q4, both for HICP and core inflation, and both for y-o-y and q-o-q changes. For the short-term interest rate, we replace the policy rates with the shadow rate (Krippner 2013) when the zero-lower bound has been binding.
5. Since Cyprus has had an independent monetary policy only until adopting the Euro in January 2008, it is not clear whether the response of the interest rate should be restricted as well. We hence estimated the model twice: with a sign restriction on the response (reflecting monetary policy autonomy) with a negative (positive) non-fiscal aggregate demand shock leading to a decline (an increase) in the interest rate, and once without that restriction (reflecting no monetary policy autonomy). The differences for the decomposition of inflation in fiscal and non-fiscal factors are minimal (Figure 1).
Robustness of Fiscal Contribution to Inflation
Citation: IMF Staff Country Reports 2024, 138; 10.5089/9798400276439.002.A001
6. We conducted an additional robustness check. We included expenditure and revenue separately (instead of the fiscal balance), which shows that the fiscal contribution to inflation can be attributed mostly to expenditure, with revenue playing some role as well (Figure 1).
References
Baba, C., & Lee, J. (2022). Second-Round Effects of Oil Price Shocks—Implications for Europe’s Inflation Outlook. IMF Working Papers, WP/22/173.
Beyer, R. (2023). Causes and Implications of Elevated Inflation in Cyprus (No. 2023/042). IMF Selected Issues Paper, 2023/042.
Beyer, R., Chen, R., Li, C., Misch, F., Ozturk, E. O., & Ratnovski, L. (2024). Monetary Policy Pass-Through to Interest Rates: Stylized Facts from 30 European Countries. IMF Working Papers, WP/24/9.
Beyer, R., Duttagupta, R., Fotiou, A., Honjo, K., Horton, M. A., Jakab, Z., ... & Zeng, L. (2023). Shared Problem, Shared Solution: Benefits from Fiscal-Monetary Interactions in the Euro Area. IMF Working Papers, WP/23/149
Bianchi, F., & Melosi, L. (2017). Escaping the Great Recession. American Economic Review, 107 (4), 1030–1058.
Blanchard, O. J. (1986). The Wage Price Spiral. The Quarterly Journal of Economics, 101 (3), 543–565.
Canova, F., & De Nicolo, G. (2002). Monetary Disturbances Matter for Business Fluctuations in the G-7. Journal of Monetary Economics, 49 (6), 1131–1159.
Colonna, F. Torrini, R. and E. Viviano. (2023). The profit share and firm mark-up: how to interpret them? Banca d’Italia Occasional Papers, 5/2023.
Davoodi, H., Nguyen, A. D. M. and Poplawski-Ribeiro, M. (2023). Fiscal and Non-fiscal Drivers of Inflation: A Decomposition Analysis. Internal IMF-FAD Toolkit and Note, International Monetary Fund, Fiscal Affairs Department, Washington DC.
European Central Bank (ECB). 2021. Monetary-fiscal Policy Interactions in the Euro Area. ECB Occasional Papers, 273.
Fröhling, A., O’Brien, D., and S. Schaefer. (2022). A New Indicator of Domestic Inflation for the Euro Area. ECB Economic Bulletin, 4/2022.
Gambetti, L., & Musso, A. (2017). Loan Supply Shocks and the Business Cycle. Journal of Applied Econometrics, 32 (4), 764–782.
Gonçalves E. and G. Koeste. (2022). The Role of Demand and Supply in Underlying Inflation– Decomposing HICPX Inflation into Components. ECB Economic Bulletin, 7.
Hansen, N., Toscani, F., and Zhou, J. (2023). Euro Area Inflation after the Pandemic and Energy Shock: Import Prices, Profits and Wages. IMF Working Papers, WP/23/131.
International Monetary Fund. (2023a). Cyprus: Staff Report for the 2023 Article IV Consultation. Washington, DC.
International Monetary Fund. (2023b). Inflation and Disinflation: What Role for Fiscal Policy?, Chapter 2 in Fiscal Monitor, April. Washington, DC.
International Monetary Fund. (2024a). World Economic Outlook, April 2024. Washington, DC.
International Monetary Fund. (2024b). Cyprus: Staff Report for the 2024 Article IV Consultation. Washington, DC.
Krippner, L. (2013). Measuring the Stance of Monetary Policy in Zero Lower Bound Environments. Economics Letters, 118 (1), 135–138.
Lorenzoni, G. and Werning, I. (2023). Inflation is Conflict. NBER Working Paper, w31099.
Uhlig, H. (2005). What are the Effects of Monetary Policy on Output? Results from an Agnostic Identification Procedure. Journal of Monetary Economics, 52 (2), 381–419.
Prepared by Robert Beyer (EUR). The paper benefitted from helpful comments and suggestions by Prof. Marios Zachariades (University of Cyprus), Iacovos Sterghides (Central Bank of Cyprus), Mark Horton, Alex Pienkovski, and Moheb Malek (all IMF), as well as from participants at a seminar held at the Central Bank of Cyprus. Part of the analysis is based on a toolkit made available by the Fiscal Affairs Department of the IMF. Any remaining errors are my own.
For the estimations we use an internal toolbox developed by the Fiscal Affairs Department of the IMF (Davoodi, Nguyen, and Poplawski-Ribeiro 2023).
Other shortcomings include the exclusive reliance on statistical relationships (and not economic theory), potential oversimplification, and biases if demand and supply shocks are not fully exogenous.
One could consider an alternative identification scheme and assume that within-quarter fiscal policy does not respond to the business cycle or supply shocks within that same quarter. This would change the interpretation of the fiscal factor and increase its contribution.
To the extent that the fiscal measures had a positive impact on growth and—in turn—tax revenue, the overall contribution of the fiscal factor is likely to have been somewhat larger.
Note that the model shows a negative contribution of fiscal shocks at the beginning of the pandemic in 2020. This is due to a smaller fiscal response to the large non-fiscal shocks in 2020Q2 than the model predicts based on historical relationships. The endogenous response contributed positively to inflation.
National accounts report profits as gross operating surplus and mixed income, which differ from profits shown in company accounts in several ways. For a discussion of the different concepts and their relationships, see Hansen, Toscani, and Zhou (2023).
Usually, the transmission is nearly complete after three months, so for international comparison we compute beta coefficients from the first increase of the policy rate to three months after the last increase (see Appendix C for more information and robustness checks). For the international comparison, the data ends in August 2023.
The stable share could either signal strong market power of banks, an inability of banks to absorb long-term funding risks, or low financial literacy. Most likely it has been a combination of these factors. Interestingly, the share has declined recently, with the share of fixed-rate mortgages for house purchases increasing from 7 percent in June 2022 to 32 percent in January 2024.
For lending rates, the pass-through in the baseline and until December 2023 are very similar, indicating delayed transmission.
