FAVAR Model Research Articles

Two-step estimation of a factor model in the presence of observable factors

The dynamic factor model of Stock and Watson (2005) and the FAVAR model of Boivin (2009) include both observable and unobservable factors, and they estimate the model by using an iterative procedure. This paper presents a two-step procedure.

Drivers of Exchange Rate Dynamics in Selected CIS Countries: Evidence from a FAVAR Analysis

We investigate the likely sources of exchange rate dynamics in selected CIS countries (Russia, Kazakhstan, Ukraine, Kyrgyzstan, Azerbaijan, and Moldova) over the past decade (1999-2008). The analysis is based on country VAR models augmented by a regional common factor structure (FAVAR model). The models include nominal exchange rates, the common factor of exchange rates in the CIS countries, and global drivers such as gold, oil and share prices. Global, regional and idiosyncratic shocks are identified in a standard Cholesky fashion. Based on the decomposition of the variance of forecast errors, their relevance for exchange rates is explored. As a quite robust finding, CIS exchange rates have become more vulnerable to global shocks towards the end of the sample.

MONETARY POLICY AND ASSET PRICES IN A SMALL OPEN ECONOMY: A FACTOR-AUGMENTED VAR ANALYSIS FOR SINGAPORE

The ongoing global financial turmoil has revived the question of whether central bankers ought to tighten monetary policy pre-emptively in order to head off asset price misalignments before a sudden crash triggers financial instability. This study explores the issue of the appropriate monetary policy response to asset price swings in the small open economy of Singapore. Empirical analysis of monetary policy based on standard vector autoregression (VAR) models, unfortunately, is often hindered by the use of sparse information sets. To better reflect the extensive information monitored by Singapore's central bank, including global economic indicators, we augment a monetary vector autoregression (VAR) model with common factors extracted from a large panel dataset spanning 122 economic time series and the period 1980q1–2008q2. The resulting FAVAR model is used to assess the impact of monetary policy shocks on residential property and stock prices. Impulse response functions and variance decompositions suggest that monetary policy can potentially be used to lean against asset price booms in Singapore.

Hypercholesteremia: Its Clinical Significance and Management

We review the literature and examine the effects of shocks on bitcoin returns. We assess the effects of factors such as stock market returns, exchange rates, gold and oil returns, FED's and ECB's rates and internet trends on bitcoin returns. Alternative VAR and FAVAR models are employed and generalized as well as local impulse response functions are produced. Our results reveal (i) a significant interaction between bitcoin and traditional stock markets, (ii) a weaker interaction with FX markets and the macroeconomy and (iii) an anemic importance of popularity measures. Lastly, we reveal the increased impact of Asian markets on bitcoin compared to other geographically-defined markets, which however appears to have waned in the last two years after the Chinese regulatory interventions and the sudden contraction of CNY's share in bitcoin trading volume.