Are recent tanker freight rates stationary?

Abstract

The evidence concerning the stationary nature of freight rates is quite ambiguous: some studies conclude that freight rates are stationary, and others conclude the opposite. However, none uses spot and futures freight rates jointly to determine whether freight rates are stationary. Studying the stationary nature jointly in spot and futures freight rates is important because, theoretically, futures freight rates are the risk-neutral expectation of spot freight rates, and if this close relationship is not maintained in practice, arbitrage opportunities could arise. In this article, we investigate the main empirical properties of several freight rate time series, and when we calibrate stochastic models using joint spot and futures freight rates, we find that there are serious doubts about the stationary nature of freight rates. Moreover, a non-stationary factor-model for freight rates is proposed and estimated, and we show that it achieves better results than the stationary model. This means that if freight rates are non-stationary and a stationary model is applied to characterize their dynamics, model fit will be poor and, consequently, valuations based on a stationary model will not be sufficiently accurate. Similar arguments apply in the case of hedging: whenever the construction of a hedging portfolio is attempted, based on non-stationary assumptions, applied to stationary freight rate series, the hedging strategy will not be effective because the residual (basis) risk will be very large.