Re-measuring the natural interest rate for the Turkish economy: Does using the new GDP series matter?

Abstract

The Turkish Statistical Institute announced a new GDP series in Turkey at end-2016, which produced a remarkable upward revision in real GDP growth during the 2009-2015 period. The divergence between the old and the new series also pointed to a higher potential output, bringing the possibility that the deviation of output from its potential, i.e. the output gap has changed in the meantime. This necessitated the re-estimation of some indicators, the measurement of which is based on GDP, potential output and the output gap. Natural interest rate is one such important indicator that is used to assess the monetary policy stance, yet it is unobservable. Hence, this paper measures the natural interest rate for the Turkish economy using the new GDP series by applying extended Kalman filter to an unobserved components model in a state-space form. In order to determine how natural interest rate is affected by the GDP revision, natural interest rate is also estimated by using the old GDP series. The estimation results of the state-space model show that the revision in GDP series is reflected on the natural interest rate. Accordingly, a difference has been observed in the natural interest rate estimate as of 2009, and this difference is even more apparent in 2015. The results also imply that the recent monetary policy conduct is consistent with the economic fundamentals. In the meantime, other estimated series are plausible and capture the significant turning points of the economy, while the time-varying parameters are also reasonable. In sum, findings confirm the significance of using natural interest rate as a benchmark in the conduct of monetary policy. Yet, the imprecision of the natural interest rate estimate due to its unobservable nature implies that despite being an important gauge for the monetary policy stance, major policy decisions should still not be based solely on the natural interest rate.