“Any year that passes in which you don’t destroy one of your best loved ideas is a wasted year....
...real knowledge is knowing that you don't know anything." [Charlie Munger]
...real knowledge is knowing that you don't know anything." [Charlie Munger]
Work in progress: working hard... (or hardly working?)
- Learning about Monetary Policy using VARs? Some Pitfalls and Solutions (with Jan Bruha), [slides]
- Forecasting and Policy Analysis with Trend-Cycle Bayesian VARs [draft] (new) + [handout]
- Estimation of Large-Scale Nonlinear Models using Bayesian Simulated Method of Moments with System Priors, [slides]
(A) Unpolished Work*
- Forecasting and Policy Analysis with Trend-Cycle Bayesian VARs [draft] (May 2017). This paper introduces Trend-Cycle Bayesian VARs for forecasting and policy analysis. Variables are modeled as a trend and cyclical components, jointly, in a state-space formulation. The benefit of trend-cycle VAR is flexible handling of trends and cycles, with clearly determined steady state. This enhances forecasting performance, lower misspecification, and puts forecasters in control of the medium-run dynamics. [ECB Seminar Slides]
- Parameter and State Estimation with Singular DSGE Models [pdf] (Feb 2016), This short note proposes a flexible method for estimating structural shocks with models that are stochastically singular. The estimation is using a standard under-determined least-squares solution. The resulting 'SVD filter' can work well with both noisy and missing observations and allows for further L2 or L1 penalties on the parameter and shock estimation.
- Estimating Structural Shocks with DSGE Models [pdf]. [CEF-2014 slides] In this paper I propose an alternative way of estimating structural shocks in DSGE models, where the requirement on structural shocks to be uncorrelated is imposed. It is only when the shocks of the model are uncorrelated when model properties --IRFs, FEVDs, spectrum-- are meaningful. Most DSGE models' estimates of structural shocks are correlated and signal misspecification of the model. The uncorrelatedness requirement induces a meaningful distinction between structural shocks and residuals when DSGE models are brought to data. The paper also illustrates that most of typical DSGE models of these days would fail the test by a large margin. SVD Filter and PROCRUSTES Filter for DSGE models are introduced.
(B) Somewhat more polished...
2022
2020
2019
2018
2017
2016
2015
2014
2013
2012
2010
2008
- Fiscal Rules in the Disaster Prone Countries: Implications for the ECCU [pdf] This paper reviews the ECCU’s experience with rule-based fiscal responsibility frameworks and considers how fiscal frameworks and fiscal policy could be enhanced to explicitly account for the presence of sizeable natural disasters.
2020
- COVID-19 Hits the Poor Harder, but Scaled-Up Testing Can Help [IMF Blog with Allan. Dizioli and John Bluedorn] https://blogs.imf.org/2020/12/03/covid-19-hits-the-poor-harder-but-scaled-up-testing-can-help/
- Countering Future Recessions in Advanced Economies [IMF's World Economic Outlook Analytical Chapter, April] - [pdf]
Our analysis, prepared largely before the onset of the covid-19 crisis in 2020, focused on making the economies more resilient to recessions by implementing rules-based cyclical fiscal policy. We illustrate that rules-based cyclical fiscal transfers, triggered by a change in unemployment can largely offset the constraints of monetary policy at the effective lower bound. Not only direct fiscal transfers help to stabilize disposable income of households but the sole existence of the rule makes households and firms react less adversely to an economic shocks [expectation, insurance channel]. With the onset of covid-19 crisis, it became even clearer how important it is to have a rule that automatically triggers fiscal stabilization, not hampered by partisanship and delays associated with discretionary fiscal policy.
- Model-Based Globally-Consistent Risk Assessment [pdf] IMF Working Paper WP/20/64 June 2020 (joint with Ben Hunt).
This paper demonstrates a new approach to create a globally-consistent model-based growth at risk and changes in the growth distributions with policy changes. Using a multi-country structural DSGE model, we illustrate a new way to carry out stochastic simulations using a non-linear model (with the effective lower bound, etc.) and the shocks drawn from empirical distribution function of estimated structural shocks. The resulting distributions can be very flexible -- skewed and with fat tails. The paper also illustrates a simple method to estimate structural shocks using non-linear large-scale models under very general assumptions.
Slides: Globally-Consistent Risk Assessment [pdf] (path linearization, jittered bootstrap, variable constraints)
- Agricultural Market Integration in India [pdf] IMF Working Paper WP/20/115. (joint with Patrick Blagrave). We assess the degree of cross-market price discrepancy (a proxy for market integration), its evolution over time, and proximate determinants, using monthly price data for 21 agricultural goods and 60 markets in India.
2019
- Assessing House Prices in Canada: Borrowing Capacity and Investment Approach [pdf] IMF WP/19/248, November 2019. This paper is an extended analysis of Canadian house prices for 11 metropolitan areas using both the borrowing capacity approach but also the investment approach and discussion of macro-prudential policies.
- Assessing House Prices in Canada [pdf] IMF's Selected Issues Paper assessing house prices in Canada and its 11 Census Metropolitan Areas (CMAs) using the borrowing-capacity approach for the IMF's Article-IV surveillance mission. The paper illustrates that most of the metropolitan areas in Canada in early 2019 were fairly-priced, with the exception of Hamilton, Toronto, and Vancouver where the market has been severely over-valued. Decomposition of change in attainable house prices into interest rate and income contribution is available.
Related: IMF Blog "To Tackle Housing Affordability in Canada, Build More Houses", media coverage in Financial Post [pdf].
- Assessing House Prices with Prudential and Valuation Measures [pdf] and SLIDES [pdf] (July 2018) joint with M. Plasil. IMF Working Paper IMF WP/19/59. In this paper, we use the concept of borrowing capacity and net-present value models to assess and appraise the value of housing assets. The advantage of the approach over econometric procedures is clear motivation by economic and finance theory and lack of revision variability. The empirical application focuses on house prices in Prague, the Czech Republic. This research has been also mentioned in IMF Research Perspectives, Fall 2018.
- The Present Value of U.S. Corporate Profits: A Forecasters' Survey Perspective [pdf] (January 2019) [slides], IMF WP 19/12. In this paper, I use the five-year ahead forecast of corporate earnings growth from the Blue Chip survey to compute the implied net present value of the U.S. corporate earnings (dividends). I analyze the resulting dynamics of asset valuation and compare with the S&P-500 market prices. The results are encouraging for using this valuation approach to back-out implied asset valuations using forecasts of policy institutions.
2018
- Italy: Toward a Growth-Friendly Fiscal Reform [pdf] joint with S. Hebous, A. Kangur, M. Raissi, IMF Working Paper WP/18/59
- Italy: Quantifying the Benefits of a Comprehensive Reform Package [pdf] joint with A. Kangur, M. Raissi, IMF WP/18/60
2017
- Banks' Adjustment to to Basel III Reform: A Bank-Level Perspective for Emerging Europe [pdf][slides] (Jan 2017, joint with V. Tomsik and J. Vlcek), IMF Working Paper WP/17/24. The paper seeks to identify strategies of commercial banks in response to higher capital requirements of Basel III reform and its phase-in. It focuses on a sample of nine EU emerging market countries and picks up 5 largest banks in each country assessing their response. Worries echoed at the early stage of Basel III compilation, namely that commercial banks would shrink their balance sheet by reducing their lending to meet stricter capital requirements, did materialize only in banks struggling with profitability.
- Structural and Fiscal Measures to Increase Potential Output in Austria [pdf], IMF Selected Issues Papers, Feb. 2017. This short paper proposes and evaluates a comprehensive structural reform and fiscal reform package for Austria. Model analysis uses IMF GIMF model for Austria.
2016
- Output and Inflation Co-Movement: An Update on Business-Cycle Stylized Facts [pdf] (Dec 2016, joint with J. Bruha and S. Solmaz), IMF Working Paper WP/16/241. In the paper, we review business stylized facts about real and nominal co-movement. We argue there is a strong and rather stable co-movement of real variables among themselves and also co-movement of real output with output inflation dynamics around their long-term expectations. The paper also suggests that the high comovement of macroeconomic variables has important consequences for structural and reduced-form economic modeling.
- System Priors for Econometric Time Series [pdf][R-code] -- (2016/2017 with M. Plasil) IMF Working Paper WP/16/231 [CNB WP 2017 updated version] introducing system priors for Bayesian econometrics. System priors are priors about selected features of the economic or econometric model, be it IRFs, variance decompositions, frequency-response functions... This paper shows how to elicit interesting priors when estimating an time series model, in this case an AR(2). Extensions to BVARs are relatively simple and are forthcoming. One of the priors on parameters is that the system be stable and that at least 60% of variance is coming from the business cycle frequency. Such prior makes the prior IRF more economically-meaningful. The methodology is also very suitable and easy to use for CALIBRATION exercises. Old, original version cobbled up for teaching purposes: [pdf]
- On the Sources of Business Cycles: Implications for DSGE Models [pdf] (joint with J. Bruha and S. Solmaz), CNB Working Paper No 3/2016. {CEF-2014 slides} What are the drivers of business cycle fluctuations? And how many are there? By documenting strong and predictable co-movement of real variables during the business cycle in a sample of advanced economies, we argue that most business cycle fluctuations are driven by one major factor at business cycle frequency. This has some interesting implications for structural economic models in terms of their specification and estimation.
2015
- Reforming Fiscal Governance in the European Union [pdf] (joint with J. Bluedorn, L. Eyraud, T. Kinda, P. Koeva Brooks, G. Schwartz, and A. Weber), IMF Staff Discussion Note SDN/15/09. The purpose of the paper is to present options for simplifying the EU fiscal governance framework while enhancing its overall effectiveness. The current framework involves an intricate set of fiscal constraints. For example, both the preventive and corrective arms of the Stability and Growth Pact (SGP) constrain fiscal policies of EU member states through various targets, upper limits, and benchmarks. Empirical computations and model-based simulations are used to support the message of the paper.
- Banks in the Global Integrated Monetary and Fiscal Model (GIMF), [pdf] (joint with M. Kumhof, D. Laxton, and D. Muir), IMF WP/15/150. The paper describes implementation of banking sector into the IMF GIMF model and articulates the model's properties. The banks follow the loanable-funds principle, still. Incorporation of `fountain-pen' loan creation is left for further work. This model is still a `loanable-funds' model.
- The Flexible System of Global Models -- FSGM [pdf] (joint with Economic Modeling Division, IMF RES), IMF Working Paper WP/15/64. The Flexible System of Global Models (FSGM) is a group of models developed by the Economic Modeling Division of the IMF Research Department for policy analysis of the global economy. A typical module of FSGM is a multi-region, forward-looking semi-structural global model consisting of 24 regions. The FSGM is used for policy analysis and scenarios for the IMF World Economic Outlook (IMF), Spillovers Reports, Regional Economic Outlooks, and for country-level analysis within the Fund.
2014
- A Model-Based Analysis of Spillovers: The Case of Poland and the Euro Area, [pdf] (joint with Giang Ho and R. Garcia-Saltos) IMF Working Paper WP/14/186. We use simple model framework to investigate financial and real spillovers from Euro zone to Poland. Our model was used for Article IV mission to Poland. We model spillovers and global shock in a two country, semi-structural setup. See some previous results in IMF's Poland: Selected Issues Paperl [pdf] and in Article IV review (Box 3) [pdf]. The paper has been a long, arduous journey and there are still many gaps to be filled in terms of the model structure and properties. The paper is, however, very transparent where these gaps are and illustrates the process of building a small semi-structural policy model...
2013
- The Role of Domestic and External Shocks in Poland: Results from an Agnostic Estimation Procedure, [pdf] IMF Working Paper WP/13/220, (joint with R. Garcia-Saltos and Giang Ho). We motivate and estimate a trend-cycle vector auto-regressive model (TC-VAR) with block-restrictions to analyze how much do foreign shocks affect the Polish economy. The approach is novel in its emphasis on careful specification of the reduced form model using the restrictions implied by the economic theory and existing policy regimes.
- System Priors -- Formulating priors about DSGE models properties, (joint with Jaromir Benes) [pdf], IMF Working Paper WP/13/257 Slides CEF-2013 [pdf], slides IMF-ICD Nov 2017 [pdf]. We demonstrate how to estimate DSGE models with priors defined in terms of economic concepts like the sacrifice ratios, frequency response functions or other concepts. Our approach is very intuitive, economic and works surprisingly well on our models. The estimation method has also been partially implemented into IRIS Toolbox for Matlab by Jaromir Benes. NEW: System Priors Implementation for DYNARE (2017)
- Inflation and output comovement in the euro area: Love at second sight? [pdf], (joint with Jan Bruha and Serhat Solmaz), IMF Working Paper WP/13/192. We calculate trimmed mean inflation measure for the euro area, which matches cyclical frequencies of output, to point out a close link. Our results suggest a strong co-movement of output and inflation in the euro area, consistent with a simple Phillips curve intuition.
- Understanding DSGE filters in forecasting and policy analysis, DYNARE WP No. 16/2012 [pdf], updated in IMF WP/13/98[pdf]. The paper illustrates novel ways of exploring DSGE models and their implied filtering representation. Estimates of unobserved quantities are represented as a function of underlying observed data. Simple way of imposing expert judgment on Kalman filtering, using dummy observations is presented. Presented at 8th DYNARE conference. (Useless to academia, useful for policy and forecasting...)
- What Is in Your Output Gap? Unified Framework & Decomposition into Observables, 2013, IMF WP/13/105, [pdf] This paper presents a unifying framework for output gap estimation -- theory of linear filters. It also explains how to decompose the output gap measures, even using DSGE models, as a function of observed data inputs, e.g. inflation, output or unemployment.
- Forecasting and Monetary Policy Analysis in Low-Income Countries: Food and non-Food Inflation in Kenya, 2013, IMF Working Paper, WP/13/61. [pdf]. (joint with Andy Berg, Rogelio Morales, Rafael Portillo and Jan Vlcek) We develop a semi-structural new-Keynesian open-economy model, with separate food and non-food inflation dynamics, for forecasting and monetary policy analysis in low-income countries and apply it to Kenya.
- Money Targeting in a Modern Forecasting and Policy Analysis System: an Application to Kenya, 2013, IMF WP/13/239[pdf] (joint with Andy Berg, Enrico Berkes, Armando R. Morales, Rafael A. Portillo, and Jan Vlcek). This paper uses a simple model to show that even a very sophisticated attempt for flexible money-targeting is a bad idea. The paper also discusses why money-targeting regime is essentially impossible to implement due to the nature of the commercial banks' balance sheets.
2012
- As you sow, so shall you reap: Public investment surges, growth, and debt sustainability in Togo, [pdf] IMF Working Paper, 2012, WP/12/127 (joint with Antonio C. David, Raphael Espinoza, Marshall Mills and Luis-Felipe Zanna) The paper outlines a model for analysis of debt sustainability in low-income countries. Public investment, learning by doing and public finance constraints are explicitly model. An empirical pilot exercise is carried using the Togolese economy to uncover sources of growth and check IMF DSA consistency. [version francais!]
- Structural correlation decomposition for business cycle analysis, [link] Economics Letters, vol. 115, issue 3, pp. 390-391, 2012 Simple, yet practical method of decomposing correlations into contributions of group of stochastic shocks. Originally coded to handle a stochastic version of the IMF GIMF model ;) Simple, yet fun...
2010
- A note on identification patterns in DSGE models, European Central Bank WP No. 1245/August 2010 [pdf] Discussion of parameter identification in dynamic general equilibrium models. Practical methods how to detect non- and weak identifiability in models and how to rank parameters by their strength of identification, using repeated singular value decomposition of Fisher Info. matrix. Insights from the paper have been used in Identification toolbox for DYNARE, viz. (link)[pdf]
2008
- Implementing the New Structural Model of the Czech National Bank, Czech National Bank WP 2009/2 [pdf] (joint with T. Hledik, O. Kamenik and J. Vlcek) detailed write-up on a development and practical use of the core DSGE model of the Czech National Bank. The model has been used for forecasting and published interest rate path since 2008, after a year of shadow forecasting. The path of interest rates published is the outcome of the judgmental forecast produced using the model, see our first forecast [link]...
- Adding Indonesia to the Global Projection Model, International Monetary Fund WP/09/253, 2009, [link] (joint with Ch. Freedman, Roberto Garcia-Saltos, Denny Hermawan, Douglas Laxton and Harris Munandar) Work produced when providing technical assistance to Bank Indonesia, Jakarta, with Chuck Freedman and Doug Laxton. GPM is a multi-country projection tool of the IMF Research dept.
- The role of trends and detrending in DSGE models: Emerging countries need trendy models, CNB mimeo, 2008, [pdf] A short, not-so-well written note on consequences of stochastic trends in DSGE models. A permanent shock propagates to business cycle frequency dynamics of the data and cycles in emerging economies are due to permanent shocks. Frequency domain techniques are used how much dynamics due to permanent shocks is left in the data after HP filtering or related method.
Hard to finish, too geeky, or sleeping...
- Estimating singular DSGE models using likelihood-based dimensionality reduction, draft, [pdf] and [slides, CEF, 2012] The paper uses dynamic principal components in combination with likelihood function to estimate DSGE models with more data than shocks, i.e. models featuring stochastic singularity. The principal component approach endogenously pins down structure of measurement errors.
- Cheers to the good health of the US short-run Phillips Curve: 1960 -- 2012, draft [pdf], [slides, Notre Dame, Midwest Macro] The paper document strong output and inflation co-movement in the US, using spectral measures. Statistical and economic arguments are used to define a measure of inflation target in the US. A simple New Keynesian DSGE model built fits inflation dynamics using demand shocks.[Here] you can find some simple graphs demonstrating that despite, or due to, active monetary policy there is a tight link between output and inflation in the United states.
- Exact nonlinear filter for DSGE models, [deprecated, newer efficient method forthcoming...], 2012 [tutorial pdf + code] A simple way to estimate shocks in nonlinear DSGE models, a nonlinear Kalman filter based on the least squares principle. This is a TOY code... Currently working on an different, efficient algorithm exploiting the structure of the stack-time algorithm devised by Juillard-Laffarque-Bouccekine...
- DSGE2VAR: Mapping Structural Models to VAR Models [pdf] -- a super short note of mapping structural shocks of DSGE models into a structural VAR model (SVAR)
*Disclaimer
The views expressed herein are those of the author and should not be attributed to the International Monetary Fund, its Executive Board, or its management.
