international econometric journal
in Russian language
modeling time series dynamics one has to decide on the class and type of models
to use which depends much on the object to be modeled. This essay briefly over
We discuss the Markov switching model, one of most popular nonlinear time series models. This model involves switching between multiple structures that characterize different time series behaviors in different regimes, the switching mechanism being controlled by an unobservable variable that follows a Markov chain process. We introduce a simple Markov switching model of conditional mean and describe its generalizations, study estimation methods, discuss how to conduct hypothesis testing, and elaborate on two empirical examples.
Repetitive interactions among players are common to many actual markets. If this is complemented by guaranteed sales markets and the presence of dominant players, then it creates natural conditions for coordination. Economic theory suggests that the decline in competition up to the cartel level can provide higher profitability. However, coordination in such markets may alternate with competition and price wars. The observed rice prices of major Asian exporters demonstrate similar dynamics, which implies coordination of their actions. However, it is an open question how close this behavior is to that of a cartel. Using monthly data on export sales of rice from 1997, the study examines the degree of imperfect competition and the possibility of collusion. Estimation is carried out using the generalized method of moments.
paper examines the demand structure for diapers in one of Russian cities with
the use of a discrete choice model with random coefficients in consumer
utilities. As a first step, logit and nested logit specifications are
estimated. The market-level information for sales volumes, prices and other
product characteristics is used. An unbalanced bimonthly panel for the period
from 2008 to 2011 is available. At the second stage, the full model adapted
studies whether politicians manipulate monetary instruments to win elections in
the new democracies. The question makes sense because the Central Bank in the new
democracy conditions is usually weak. A sample of 8 new democracies is analyzed
via individual country vector autoregressions and via simple autoregressions
for each variable of interest. I test various opportunistic political cycle
models, both with adaptive and rational expectations. My results reject the
political business cycle model with adaptive expectations due to the lack in
the data of any impact of the electoral monetary expansion on output. However,
there is evidence of opportunistic behavior in combination with rational
There is an opinion that the Russian Central Bank's actual policy in 2000Ц2008 was real exchange rate targeting. At the same time, the Central Bank regularly declared inflation targets, but regularly missed them. We estimate a simple structural threshold VAR model of the Russian economy to test for these two regimes of the CB's policy. Our testing procedure is based on applying the bootstrap to the estimated TVAR model. We find significant nonlinearity (two policy regimes) caused by endogenous switching between regimes based on past month inflation. The Central Bank by changing its targets was not able to commit to inflation reduction in that period of time, hence facing the issue of credible monetary policy. After the global financial crisis the CB declared its commitment to inflation targeting. But, based on our findings, the policy will be successful only if the Central Bank actually commits itself to reaching inflationary goals.