Abstract
This paper aims to study the connectedness and potential synchronization between financial and business cycles in the Saudi economy during the period 1970-2022. To surpass univariate filters such as HP and CF, which allow the extraction of cycles only when frequencies are predetermined, we opt for three novel techniques: the continuous wavelet coherence, the wavelet quantile correlation, and the maximum overlap discrete wavelet transform with turning point analysis. These methods assume that the characteristics of cycles are not fixed over time and allow the dominant frequencies to change from one period to another. Moreover, the adopted methodology enables the determination of the existence of common cycles between variables and the evaluation of their intensity over time. To understand real business cycles, movements in non-oil per capita real GDP are used, while financial cycles are captured through the credit-to-GDP ratio. Our results reveal that, in Saudi Arabia during 1970-2022, financial cycles are longer and more extensive than real business cycles, and the two cycles are synchronized approximately 55% of the time. Wavelet coherence and phase difference indicate that the synchronization between financial cycles and business cycles fluctuates significantly across frequencies and over time. Wavelet quantile correlations demonstrate that the relationship between business and financial cycles varies across different quantile frequencies and time horizons. These findings are of primary importance for authorities to design appropriate macroprudential policies.