| Tsolakis et al. [25] | Error correction model; comparison with AR; test for co-integration (Johansen, (1991) [3] ) 5% level is >0. Neither autocorrelation or heteroskedasticity; OLS estimator is BLUE | Used ships prices; annual data; stationarity secured by 1st differences in all relationships by ADF | New ships prices and time charters affect used ship prices; cost of capital valid only for bulk carriers; mixed results among sizes; order-book causes used ships’ prices to fall in cases |
| Tvedt [54] |
| Dry bulk market | He challenged the 1st differences stationarity of most shipping time series; if data transformed by a conversion from $ to yen, series became stationary… |
| Adland et al. [55] | Valuation model; Co-integration; pricing the forward ship value agreements; GARCH | 2nd hand ship market | Unbiasedness of implied forward prices in the 2nd hand ship market; market efficiency exists |
| Adland & Koekebakker [56] | A mathematical model to value forward time charter & spot rates | Option valuation; empirical volatility of forward freight rates; valuing ship prices. Brokers’ quotations | A model of forward freight rates dynamics; time charter rates |
| Adland & Cullinane [57] | VAR General non-parametric Markov diffusion model | Spot freight rates in tanker markets; this is mean reverting… in the extremes; volatility of freight rate changes increase with its level | They challenged that the 1st differencing in a VAR model is necessary, if series are truly stationary |