Suchergebnisse

IN A RECENT ARTICLE. GERTLER (1982) PRESENT A MODEL THAT EXPLAINS CERTAIN STYLIZED FACTS CONCERNING THE RELATIVE BEHAVIOUR OF THE MONEY SUPPLY AND NOMINAL WAGES. ONE ADDITONAL FEATURES OF THE MODEL IS THE BEHAVIOUR OF REAL OUTPUT, WHICH IS FOUND TO HAVE A SERIALLY CORRELATED RESPONSE TO A MONEY SUPPLY SHCOK. THIS SERIAL CORRELATION APPEARS TO ARISE FRTOM AN INFORMATION STRUCTURE THAT PRECLUDED AGENTS FROM DIRECTLY DISENTANGLING PERM ANENT VERSUS TRANSITORY MOVEMENTS IN RELEVANT STATE VARIABLES. THIS INFORMATION STRUCTURE COMPLICATES THE FORECASTING PROBLEM THAT MUST BE SOLVED TO ANALYZE THE MODEL AND AFFECTS THE BEHAVIOUR OF WAGES AND THE VAIRANCE OF OUTPUT. IT WILL NOT. HOWEVER, LEAD TO SERIAL CORRLEATION IN REAL OUTPUT. AN ALGEBRAIC ERROR IS RESPONSIBLE FOR THE SERIAL CORRELATION FOUND BY GERTLER. THIS NOTE CORRECT THE ERROR AND PRESENTS AN ALTERNATIVE SOLUTION TO THE FORECASTING PROBLEM. GERTLER CONSTRUCTS FORECASTS OF THE RELEVANT STATE VARIABLE THROUGH THE USE OF A KALMAN FILTER. WHILE THIS TECHNIQUE IS APPROPRIATE FOR THE PROBLEM, A SIMPLER SOLUTION IS AVAILABLE. THIS SOLUTION IS BASED ON AN EQUIVALENT REPRESENTATION OF THE STOCHASTIC PROCESS GENERATING THE DISTURBANCES OF THE MODEL. THIS EQUIVALENT REPRESENTATION CAN THEN BE USED, IN A STRAIIGHTFORWARD WAY, TO DESCRIBE WAGE DYNAMICS AND THE VARIANCE OF REAL OUTPUT.

IN THIS PAPER THE AUTHORS WILL SHOW THAT, EXCEPT IN A BORDERLINE CASE, MUNDELL'S MODEL HAS TWO STEADY STATES (INSTEAD OF ONE, AS WAS TAKEN FOR GRANTED) AND THAT THE NUMERICAL EXAMPLES FOCUSED ON THE ONE THAT EXHIBNITS THE LOWEST LEVEL OF REAL MONETARY BALANCES AND GROWTH. IF INSTEAD THE ANALYSIS WERE TO BE BASED ON THE SECOND STEADY STATE, A DRAMATICALLY DIFFERENT RESULT IS FOUND: THERE EXISTS NO UPPER BOUND TO THE STEADY-STATE SEIGNIORAGE-FINANCED RATE OF GROWTH. AND THE LATTER INCREASES WITHOUT BOUND AS THE RATE OF EXPANSION OF THE MONEY SUPPLY. APPROACHES ZERO.