This thesis reflects on the recommendation by financial regulators to shift from estimation-based to transaction-based benchmarks following the 2012 London Interbank Offered Rate (LIBOR) manipulation scandal and structural changes in underlying markets after the 2007-08 global financial crisis. This paradigm shift is based on the neo-classical view that actual rates reflect perfectly competitive money market conditions and are more robust in monetary policy transmission. The first core chapter uses the LIBOR and available novel data from the Bank of Zambia to empirically test whether actual rates are susceptible to manipulation and/or collusion using a mathematical screening methodology, Benford’s Law. The evidence shows that both estimation-based and real rates are susceptible to manipulation and/or collusion and calls for regulators to remain vigilant. The second core chapter empirically investigates the interconnectedness and transmission of monetary policy of foreign exchange swap implied rates, LIBOR, and the overnight index swaps for 1-month, 3-month, and 6-month maturity categories for three major currencies, namely, Euro (EUR), Pound Sterling (GBP), and the Japanese Yen (JPY), using the time-varying vector autoregression model. The results show that for all currencies, interconnectedness is time-varying and is high in periods of market stress and volatility. Further, the transmission of shocks varies per currency and instrument, implying that no “one size fits it all” model exists for effective monetary policy transmission. Using liquidity metrics and network analysis, the third core chapter uses novel granular data to test the following hypotheses: limited liquidity and absence of transactions in interbank money markets constrain representative pricing of benchmarks; liquidity provision (access) is segmented; and there are price differentials in interbank market segments. The liquidity metrics show that the money market comprises a large and liquid ultra-short-term market and a small illiquid term market. Therefore, representative rates can only be achieved in the former. Further, there is evidence of market segmentation and price differentials, with a preference for trading within the bank categories.
A Paradigm Shift from Estimation-based to Transaction-based Money Market Benchmarks: An Empirical Assessment of Collusion, Robustness, and Representativeness
Muchimba, L. (Author). 15 May 2023
Student thesis: Doctoral Thesis