The price puzzle is the recurring empirical finding, in structural vector autoregressions (SVARs) used to study monetary transmission, that a contractionary monetary policy shock — an unexpected rise in the policy interest rate — is followed by a rise rather than a fall in the price level. Tas (2011) proposed that the puzzle arises from asymmetric information between the central bank and the public: because the central bank has private information about future inflation and output, the public reads a policy tightening as a signal that inflation will be higher, so expected and then actual inflation rise. This problem asks for the unresolved quantitative and comparative completion of that research program: (i) a decomposition of the observed price puzzle into the component attributable to central-bank asymmetric information (public learning from policy actions) and components attributable to competing explanations, obtained by formal estimation rather than illustrative calibration; (ii) a direct test of whether central-bank transparency systematically shapes the inflation response to monetary policy; and (iii) evidence on central-bank private information outside the Federal Reserve, particularly in emerging economies, where it remains essentially unmeasured. Later literature has confirmed the information channel qualitatively for the United States and the United Kingdom, but no accepted quantitative decomposition exists, the explanation fails in Australia, and the transparency and emerging-economy prongs are unexecuted.

Background

Since the early 1990s, empirical monetary economics has measured the effects of monetary policy with structural vector autoregressions (SVARs) — linear systems in which a handful of macroeconomic time series (output, prices, a policy interest rate, monetary aggregates) evolve jointly, and a 'monetary policy shock' is isolated as the unexpected component of the policy rate. Identification typically uses a Cholesky (recursive) decomposition, which orders the variables so that the policy rate can respond within the period to output and prices while the latter respond to policy only with a lag; the traced-out effect of the shock on each variable over time is its impulse response. Sims (1992) and Eichenbaum (1992) documented the counterintuitive result that became known as the price puzzle: after a policy tightening, the estimated impulse response of the price level is positive for several periods, contradicting the standard prediction that restricting demand lowers inflation.

Proposed remedies include adding forward-looking variables to the VAR: Giordani (2004) proved that omitting a measure of the output gap (the deviation of actual from potential output) spuriously produces a positive price response in a wide class of models, and showed the prediction holds in US data. A parallel line argues the puzzle is not a misspecification artifact at all but a genuine expectations channel operating through central-bank information: the central bank sets rates in response to its own forecasts of future inflation and output, so an observed tightening partly reveals that the bank expects higher inflation. Romer and Romer (2004) exploited the Federal Reserve's internal (Greenbook) forecasts to purge the policy rate of systematic responses to this private information, and the price puzzle disappeared in US data; Cloyne and Hürtgen (2016) obtained the same result for the United Kingdom using the Bank of England's real-time forecasts.

Tas (2011) formalized this mechanism in an explicit asymmetric-information model. Inflation and output follow VAR(1)-style dynamics, the Federal Reserve receives private noisy signals about persistent shock components, and it sets the federal funds rate by a forward-looking Taylor rule (a rule making the policy rate respond to expected inflation and output). The public observes only the interest rate and uses a Kalman filter (the standard recursive algorithm for inferring hidden state variables from noisy observations) to learn the bank's private information from policy actions. Under asymmetric information, a rate hike makes the public revise expected inflation upward, so actual inflation rises — reproducing the puzzle — whereas under a symmetric-information benchmark, in which the public sees the same signals as the bank, the inflation response to a tightening is negative. Empirically, Tas estimated a six-variable monthly VAR for the US (1978:1–2006:11) including a direct survey measure of expected inflation (the median from the Survey of Consumer Attitudes) and found that expected inflation rises significantly for three months after a funds-rate shock, and that adding expectations to the VAR reduces but does not eliminate the positive price response. The model was calibrated for illustration rather than estimated, so the exact quantitative contribution of asymmetric information to the puzzle remains open; the paper notes that its results have practical implications for central-bank transparency — the degree to which a central bank discloses its objectives, data, forecasts, and decisions, typically summarized in indices such as that of Dincer and Eichengreen (2010) — and monetary policy, but whether transparency explains the time-varying SVAR coefficients linking inflation to the policy rate is untested; and central-bank private information outside the Federal Reserve, especially in emerging economies, remains essentially unmeasured.

Subsequent work transformed the measurement of the information channel: high-frequency identification uses asset-price movements in narrow windows around policy announcements to measure surprises. Campbell, Evans, Fisher, and Justiniano (2012) distinguished Odyssean forward guidance (commitments to future actions) from Delphic guidance (announcements of the central bank's economic outlook) and found both move the economy. Nakamura and Steinsson (2018) showed Federal Reserve announcements simultaneously convey information: after a tightening, private forecasts of output growth rise, an 'information effect'. Jarociński and Karadi (2020) separated policy shocks from central-bank information shocks using the co-movement of interest rates and stock prices around announcements and found that purified monetary shocks do not generate the price puzzle, while information shocks raise prices. Miranda-Agrippino and Ricco (2021) built an information-robust high-frequency instrument and found monetary tightening 'unequivocally contractionary'; Melosi (2017) embedded the signaling mechanism in an estimated dynamic stochastic general equilibrium (DSGE) model. At the same time, the explanation's generality is contested: Tulip and Bishop (2017) showed the forecast-purging approach of Romer and Romer fails in Australia, and Beckers (2020) found the Australian price puzzle survives controlling for the Reserve Bank's own inflation forecasts, attributing it instead to credit spreads. Competing explanations remain active: Castelnuovo and Surico (2010) linked the puzzle to passive monetary policy (a weak interest-rate response to inflation) and omitted expected inflation; the cost channel (firms' borrowing costs rising with the policy rate) is pursued by Cucciniello, Deleidi, and Levrero (2022); and Ida (2024) generates the puzzle in a New Keynesian model with household heterogeneity. How much of the puzzle the asymmetric-information channel actually accounts for, across countries and regimes, therefore remains an open quantitative question.

Problem Statement

Determine, quantitatively and with formal estimation rather than illustrative calibration, how much of the price puzzle — the positive response of the price level or inflation to a contractionary monetary policy shock identified in structural vector autoregressions — is attributable to central-bank asymmetric information, in which the public infers the central bank's private information about future inflation and output from its policy actions and revises expectations accordingly; and determine whether central-bank transparency systematically shapes this inflation response.

This objective — establishing the quantitative role of the asymmetric-information channel in monetary transmission — comprises three interrelated open questions.

  1. A decomposition of the estimated price response to policy tightenings into (a) the signaling/expectations component induced by the public's learning from policy actions and (b) the residual component attributable to competing candidate explanations treated as a single group (omitted forward-looking variables such as the output gap or expected inflation, passive monetary policy, the cost channel, and identification artifacts), for the United States and for at least one additional economy for which the requisite real-time forecast or communication records are available, with the uncertainty of the attributed shares made explicit. The asymmetric-information share is defined operationally as the fraction of the cumulative impulse response of the price level over a reported horizon 0 to H that is removed when public learning from policy actions is shut down — that is, under the symmetric-information benchmark in which the public observes the central bank's signals — with the competing-explanations residual the remainder of that same cumulative response, and a single normalization and horizon are used for every economy reported; the asymmetric-information share and the competing-explanations residual are estimated rather than calibrated over a reported horizon, with their uncertainty, for the United States and for at least one economy beyond it.

  2. A direct test of the transparency hypothesis: whether the time-varying or cross-country strength of the price-puzzle response is systematically related to central-bank transparency, estimated on a panel of countries, with attention to the endogeneity of transparency to the inflation environment and to the uncertainty of the estimated relation.

  3. Evidence on central-bank private information outside the Federal Reserve — for at least one economy outside the Federal Reserve for which real-time forecast or communication records are available, including at least one emerging economy for which central-bank forecasts or survey inflation expectations are available, where the presence and severity of the price puzzle is documented but the asymmetric-information mechanism has not been tested. The non-US economy used here may be the same as the additional economy of question 1, provided question 3 poses a question that question 1 does not: a direct test of whether the central bank's forecasts contain predictive information about future inflation or output beyond publicly available data — that is, whether the estimated anticipatory component of policy reflects central-bank private information. The anticipatory component of policy is measured for that economy, and this private-information test is carried out for it.

References

  1. Tas, Bedri Kamil Onur (2011). 'An explanation for the price puzzle: Asymmetric information and expectation dynamics.' Journal of Macroeconomics 33(2):259–275. DOI: 10.1016/j.jmacro.2010.10.005. https://doi.org/10.1016/j.jmacro.2010.10.005
  2. Sims, Christopher A. (1992). 'Interpreting the macroeconomic time series facts.' European Economic Review 36(5):975–1000. DOI: 10.1016/0014-2921(92)90041-T. https://doi.org/10.1016/0014-2921(92)90041-T
  3. Eichenbaum, Martin (1992). 'Interpreting the macroeconomic time series facts: The effects of monetary policy.' European Economic Review 36(5):1001–1011. DOI: 10.1016/0014-2921(92)90042-U. https://doi.org/10.1016/0014-2921(92)90042-U
  4. Giordani, Paolo (2004). 'An alternative explanation of the price puzzle.' Journal of Monetary Economics 51(6):1271–1296. DOI: 10.1016/j.jmoneco.2003.09.006. https://doi.org/10.1016/j.jmoneco.2003.09.006
  5. Romer, Christina D., and David H. Romer (2004). 'A New Measure of Monetary Shocks: Derivation and Implications.' American Economic Review 94(4):1055–1084. DOI: 10.1257/0002828042002651. https://doi.org/10.1257/0002828042002651
  6. Castelnuovo, Efrem, and Paolo Surico (2010). 'Monetary Policy, Inflation Expectations and The Price Puzzle.' The Economic Journal 120(549):1262–1283. DOI: 10.1111/j.1468-0297.2010.02368.x. https://doi.org/10.1111/j.1468-0297.2010.02368.x
  7. Campbell, Jeffrey R., Charles L. Evans, Jonas D. M. Fisher, and Alejandro Justiniano (2012). 'Macroeconomic Effects of Federal Reserve Forward Guidance.' Brookings Papers on Economic Activity 2012(1):1–80. DOI: 10.1353/eca.2012.0004. https://doi.org/10.1353/eca.2012.0004
  8. Melosi, Leonardo (2017). 'Signalling Effects of Monetary Policy.' The Review of Economic Studies 84(2):853–884. DOI: 10.1093/restud/rdw050. https://doi.org/10.1093/restud/rdw050
  9. Nakamura, Emi, and Jón Steinsson (2018). 'High-Frequency Identification of Monetary Non-Neutrality: The Information Effect.' The Quarterly Journal of Economics 133(3):1283–1330. DOI: 10.1093/qje/qjy004. https://doi.org/10.1093/qje/qjy004
  10. Cloyne, James, and Patrick Hürtgen (2016). 'The Macroeconomic Effects of Monetary Policy: A New Measure for the United Kingdom.' American Economic Journal: Macroeconomics 8(4):75–102. DOI: 10.1257/mac.20150093. https://doi.org/10.1257/mac.20150093
  11. Jarociński, Marek, and Peter Karadi (2020). 'Deconstructing Monetary Policy Surprises—The Role of Information Shocks.' American Economic Journal: Macroeconomics 12(2):1–43. DOI: 10.1257/mac.20180090. https://doi.org/10.1257/mac.20180090
  12. Miranda-Agrippino, Silvia, and Giovanni Ricco (2021). 'The Transmission of Monetary Policy Shocks.' American Economic Journal: Macroeconomics 13(3):74–107. DOI: 10.1257/mac.20180124. https://doi.org/10.1257/mac.20180124
  13. Tulip, Peter, and James Bishop (2017). 'Anticipatory Monetary Policy and the Price Puzzle.' SSRN working paper. DOI: 10.2139/ssrn.2971775. https://doi.org/10.2139/ssrn.2971775
  14. Beckers, Benjamin (2020). 'Credit Spreads, Monetary Policy and the Price Puzzle.' Reserve Bank of Australia Research Discussion Paper 2020-01. DOI: 10.47688/rdp2020-01. https://doi.org/10.47688/rdp2020-01
  15. Reusens, Peter, and Christophe Croux (2017). 'Detecting time variation in the price puzzle: a less informative prior choice for time varying parameter VAR models.' Studies in Nonlinear Dynamics and Econometrics 21(4). DOI: 10.1515/snde-2015-0018. https://doi.org/10.1515/snde-2015-0018
  16. Estrella, Arturo (2015). 'The Price Puzzle and VAR Identification.' Macroeconomic Dynamics 19(8):1880–1887. DOI: 10.1017/S1365100514000200. https://doi.org/10.1017/S1365100514000200
  17. Muhanji, Stella, Christopher Malikane, and Kalu Ojah (2013). 'Price and liquidity puzzles of a monetary shock: Evidence from indebted African economies.' Economic Modelling 33:620–630. DOI: 10.1016/j.econmod.2013.04.048. https://doi.org/10.1016/j.econmod.2013.04.048
  18. Cucciniello, Maria Chiara, Matteo Deleidi, and Enrico Sergio Levrero (2022). 'The cost channel of monetary policy: The case of the United States in the period 1959–2018.' Structural Change and Economic Dynamics 61:409–433. DOI: 10.1016/j.strueco.2022.02.019. https://doi.org/10.1016/j.strueco.2022.02.019
  19. Ida, Daisuke (2024). 'Household heterogeneity and the price puzzle in a new Keynesian model.' Journal of Macroeconomics 79:103587. DOI: 10.1016/j.jmacro.2024.103587. https://doi.org/10.1016/j.jmacro.2024.103587
  20. Tümtürk, Oğuz (2025). 'Monetary policy, the price puzzle and inflation expectations: evidence from Türkiye.' Trakya Üniversitesi İktisadi ve İdari Bilimler Fakültesi Dergisi. DOI: 10.47934/tefeas.14.01.02. https://doi.org/10.47934/tefeas.14.01.02
  21. Dincer, Nergiz, and Barry Eichengreen (2010). 'Central Bank Transparency: Causes, Consequences and Updates.' Theoretical Inquiries in Law 11(1):75–123. DOI: 10.2202/1565-3404.1237. https://doi.org/10.2202/1565-3404.1237