Ship Valuation Using Cross-Sectional Sales Data: A Multivariate Non-Parametric Approach

Abstract

Despite the illiquid and heterogeneous nature of the second-hand market for bulk ships and the resulting difficulty of creating reliable time series of ship prices for generic ships, the literature on ship price dynamics relies heavily on time series models. In this paper we present, for the first time, an analysis of ship valuation using cross-sectional data based on actual sale and purchase transactions in the second-hand market for bulk ships. This allows us to investigate price formation in the second-market free of broker bias and measurement error. Moreover, we do not impose a strict linear model specification but allow for the presence of non-linearity in a flexible non-parametric vessel valuation function. Based on data from more than 1,850 individual sales of Handysize bulk carriers from January 1993 to October 2003, we find that the second-hand value of a vessel can be well described as a partially non-linear function of three main factors: DWT, age, and the state of the freight market. Maritime Economics & Logistics (2007) 9, 105–118. doi:10.1057/palgrave.mel.9100174

Does energy efficiency affect ship values in the second-hand market?

“…Buyers are obviously expected to pay more for vessels with higher transportation capacities because they obtain higher daily earnings all else equal (Köhn and Thanopoulou, 2011;Adland et al, 2017b), but we note that the effect is nonlinear with a decreasing rate of increase per DWT as the vessel size gets larger. A similar effect is found in the non-parametric vessel value model of Adland and Koekebakker (2007). This could be related to the non-linearity sometimes observed with regards to DWT utilization (Adland et al, 2016b), where vessels whose size is far away from the fleet average have lower utilization due to the standardization of cargo sizes.…”

“…However, the positive estimate for the second-order age term suggests that the average depreciation curve is slightly convex 5 . Adland and Koekebakker (2007) find a similar relationship and argue that this is due to the lower bound represented by the scrap value of a vessel. The number of times a vessel has been purchased previously on the second-market, which may be considered as an indicator of demand for that specific vessel, has a positive effect on price.…”

“…First, we account for market conditions using the logarithm of the one-year TC rate as proxy. This is a common approach in the literature on vessel valuation due to the strong positive correlation between the TC rate and sale price (Adland and Koekebakker, 2007). Second, we introduce information on the most important attributes of Handysize bulkers: vessel age at sale, deadweight tonnes, capacities of cranes/derricks, design speed and builder country.…”

“…For instance, we expect the asset value depreciation to be lower with a longer economic life of the ship when markets are booming. Adland and Koekebakker (2007) solve this concern by estimating a multivariate nonparametric function of age, DWT and the TC rate. In our parametric framework, we add an interaction term corresponding to the product of the TC rate and age at sale.…”

The development of a performance index for KG funds and a comparison with other shipping-related indices

“…Accordingly, the value of a one-ship company is primarily determined by the market value of the fund vessel. The market values of ocean-going vessels are usually assessed using standard valuation multipliers (Adland and Koekebakker, 2007;Esty and Sheen, 2010), and therefore the value of the fund vessel is determined based on a comparison with the transaction prices of similar vessels. Ship-broking companies maintain databases with prices from comparable transactions (or broker valuations).…”