In many shipping markets it is widely accepted, and even well documented, that current earnings and values for a given vessel have been highly correlated in time. It is therefor surprising that market cycles continue to be driven by a seemingly myopic behavior.
During the winter of 2000/2001 tanker earnings reached an historic high. These rates, along with the IMO phase-out plan for elderly single-hull tankers, induced a huge level of orders. VLCC (Very Large Crude Carrier) prices ended as high as 84 million usd. A year and a half later, order prices dropped more than 10 million and the vessels delivered entered into a market where earnings were as low as a mere 5,000 usd per day, less than a fifth of their break-even rate.
In the thesis myopic expectations is defined as the case where agents are too focused on current levels of earnings when assessing vessel values. A static model for second hand value determination is deduced through combining the net present value criterion for accepting investment projects with a division of the present value into a short and a long term horizon of earnings. Our hypothesis is that if the market is dominated by myopic agents, the short term value component will represent a disproportionate share of value. The entire data-set was supplied by ViaMar AS, and contains two earnings indicators and four vessel prices on each of seven vessels, on a quarterly basis, from the first quarter of 1990 up to and including the fourth quarter of 2000.
The testing of the hypothesis was undertaken in two steps. First regression estimation of the model, using today’s indicators of short and long term earnings as variables, was carried out based on classic least squares assumptions and also under Cochrane-Orcutt transformation. Then the corresponding estimates of the short and long term value component shares of mean secondhand value, calculated in the same manner as assumed in deducing the model, were compared with theoretical shares. The short term was a priori assumed to equal the construction time of a vessel ordered today. As indicators of short term earnings during this period we used once the timecharter equivalent earnings (TCE) and once the one year timecharter (1TC). As an indicator of expected long term earnings the newbuilding price was used. Both indicators and values were taken from the same point in time.
During the first step the regressions under classic least squares assumptions showed evidence of autocorrelation. Thus regressions were performed under Cochrane-Orcutt transformation as well. In both instances the model generally fit best with the larger and younger vessels of the dry bulk and tanker markets.
In the second step short and long term value components were calculated based on the estimated slope and intercept coefficients, and average values of the indicators. Based on the results from the regressions, the earnings indicator used is the TCE and 1TC for dry bulkers and tankers respectively. Focusing on the larger and younger vessels and comparing the calculations of estimated value components to the theoretical shares of short and long term earnings in vessel pricing, the tanker secondhand markets seem to be myopic indeed – on average placing five times more weight on short term earnings than on long term earnings. Though also highly correlated to current earnings, the secondhand prices of dry bulk carriers do not show the same behavior, and the agents seem to place an even weight on earnings in the short and long term.
Looking at the estimated intercept terms one finds large and positive estimates for tankers, but negative estimates for capesize and panamax dry bulk vessels. It would seem that the tanker regressions either explain less of valuation, or show a lesser dependency on current earnings indicators. For large dry bulk carriers it would seem that valuation is too responsive to current earnings, while on average placing a correct (theoretically) proportion of value on the short and long term horizons respectively.
It is difficult to conclude whether or not the markets are myopic through a verification of the hypothesis, though estimation results for evaluation of large and young vessels seem strong. The results additionally seem to indicate pro-cyclical valuation in dry bulk markets and a myopic behavior in the tanker markets. The one thing that seems clear is that there is more research which can be done on vintage pricing of vessels. Some possibilities are mentioned towards the end of this thesis, along with regression results on secondhand values for an aframax tanker – assuming first adaptive expectations, and second a focus on the largest vessel in a shipping market.