طراحی مکانیزم رفتاری برای مدیریت تورم انتظاری در شرایط عقلانیت محدود: شواهدی از اقتصاد ایران (۱۳۹۵–۱۴۰۳)

The persistent inflationary environment in Iran—characterized by recurrent currency crises, institutional rigidities, and behavioral frictions—has increasingly undermined the conventional effectiveness of monetary policy. Inflation expectations, as forward-looking determinants of actual inflation, have therefore become central to policy debates. Nevertheless, prevailing analytical frameworks remain grounded in rational expectations assumptions that neglect cognitive frictions and heterogeneous belief systems. Against this backdrop, the present study seeks to design an adaptive, behaviorally informed monetary policy mechanism suited to Iran’s unique macroeconomic context. It examines the dynamic interaction between inflation expectations and core monetary and nominal variables, while explicitly incorporating time-varying structures and cognitive biases in the formation of expectations.

Methodologically, the research follows a three-stage empirical strategy. In the first stage, exploratory analysis is conducted using monthly data from 2012 to 2024 on inflation, liquidity (M2), monetary base (M0), and both official and free-market exchange rates, along with proxies for expectations such as consumer confidence indices and behavioral indicators derived from Google Trends queries. Variables are harmonized to a common monthly frequency and log-transformed to ensure both stationarity and interpretability in elasticity form. In the second stage, two baseline models—Autoregressive Distributed Lag (ARDL) and Random Forest regression—are employed to detect initial associations and nonlinearities. However, both approaches prove inadequate for capturing temporal shifts in coefficients and structural volatility, rendering them unsuitable for policy-oriented forecasting in a crisis-prone environment such as Iran. Consequently, they are not retained in the final modeling stage. In the third stage, the study adopts a Time-Varying Parameter Vector Autoregressive model with stochastic volatility (TVP-VAR-SV), estimated through Bayesian inference using Markov Chain Monte Carlo (MCMC) techniques. This framework allows for evolving impulse responses, dynamic variance–covariance matrices, and coefficient trajectories over time, making it particularly appropriate for economies characterized by recurrent shocks and adaptive behavioral patterns.

The empirical results reveal pronounced time-variation in the relationship between inflation expectations and macroeconomic fundamentals. Among these, the free-market exchange rate exerts the most dominant influence, particularly during speculative episodes and periods of policy incoherence such as the 2018 currency shock and the post-2020 sanctions tightening. The monetary base emerges as the second most influential factor, whereas broad liquidity (M2) exerts only a delayed and structurally weaker effect. The estimated impulse responses indicate nonlinear and asymmetric reactions: in high-volatility regimes, expectations respond more sharply to exchange rate fluctuations and perceived policy inaction. Variance decomposition further confirms that unanticipated shocks to nominal variables account for more than 60 percent of the variation in expectations during peak inflationary episodes.

A critical contribution of this study lies in its behavioral interpretation of expectation formation. Using auxiliary evidence from digital search trends, media sentiment, and central bank communication patterns, the findings provide strong support for bounded rationality, salience-driven attention, and adaptive learning. These behavioral frictions help explain the divergence between official inflation targets and realized expectations, particularly in contexts marked by fiscal dominance or a credibility deficit.

The conclusions underscore the necessity of moving beyond conventional inflation-targeting frameworks toward hybrid approaches that integrate behavioral dimensions and account for time-varying dynamics. To this end, the study proposes a policy mechanism based on three complementary pillars: exchange rate anchoring, whereby the free-market exchange rate is stabilized as the most salient and psychologically influential nominal anchor; cognitive-aligned communication, involving forward guidance tools that employ intuitive framing, salient language, and visual reinforcement to enhance credibility and reduce interpretive noise; and behavioral monitoring, which institutionalizes real-time tracking of expectations through digital trace data, sentiment analysis, and search-based indices.

Taken together, these findings call for a dual-track monetary policy framework—one oriented toward macroeconomic fundamentals such as interest rates and the monetary base, and another explicitly dedicated to managing public perceptions. Practical recommendations include enhancing the transparency of foreign exchange operations, simplifying inflation reports, and embedding behavioral insights teams within the central bank structure. Ultimately, the study argues for a paradigm shift in which managing expectations is regarded as equally critical as controlling fundamentals. In structurally uncertain economies such as Iran, where uncertainty and information asymmetry are entrenched features, the integration of behavioral economics into monetary policymaking is not optional but essential.