Events over the past year demonstrated not only that volatility in shipping markets never went away, but that it’s back… big time. Thus, the Baltic Exchange Dry Index nearly quadrupled in value in the brief period from the end of January to the end of June 2020, as the aftershocks of COVID-19 first ravaged and then spurred international trade in bulk commodities (Heiland and Ulltveit-Moe 2020). Likewise, the index nearly doubled in value in the even briefer period from mid-February to the end of March 2021, when the containership Ever Given blocked the Suez Canal and momentarily captured the world’s attention.
Alongside such considerations of dramatic intra-annual movements in freight rates, professional sentiment has long argued for the existence of alternating booms and busts in the maritime shipping industry (Metaxas 1971, Cufley 1972, Stopford 2009). What is more, a burgeoning academic literature in behavioural finance and industrial organisation has taken these claims to heart, finding that such boom/bust activity goes a long way in understanding the dynamics of ship building, ship earnings, and ship prices in the dry bulk sector (Kalouptsidi 2014, Greenwood and Hansen 2015).
However, is it possible to rationalise the often extreme inter-annual changes we observe in dry bulk freight rates by considering fundamentals in the sector?
A new series of dry bulk freight rates
Here, we focus on activity in the dry bulk sector — that is, commodity cargo like coal, grains, and ore, which is shipped in large, unpackaged parcels — for two principal reasons. For one, this sector represents roughly 50% of world trade by volume in the present day (UNCTAD 2015). Historically, this share would have been only higher, given that the composition of trade by value did not begin to favour manufactured goods until the late 1950s (Jacks and Tang 2018). Thus, developments in the dry bulk sector loom large in our understanding of the global economy and its evolution.
For another, dry bulk markets are decentralised spot markets whereby exporters, importers, and traders must engage in a search process in order to hire a ship for a specific itinerary. Thus, their hire rates reflect real-time conditions in the supply of and demand for their services. This is in contrast to some other means of maritime transport, like containerships or liners, which operate in between fixed ports on fixed schedules and which can be bound to long-term contracts among exporters, importers, and shippers (Cosar and Demir 2017).
Using a new and comprehensive dataset on global dry bulk freight rates from 1850 to 2020 (Jacks and Stuermer 2021), we develop a new real dry bulk index. To our knowledge, this is the longest consistently measured and continuous series on the costs of shipping goods in the literature. The final series is depicted in Figure 1 below.
Figure 1 Real dry bulk index, 1850-2020 (1850=100)
Notes: The solid black line represents the real dry bulk freight rate index, constructed as described in the full paper. The dotted black line is an estimate of the long-run trend derived from the Christiano-Fitzgerald band pass filter, which assumes a cyclical component of 70 years duration in the real dry bulk freight rate index.
Figure 1 allows us to document the following important facts:
- Real dry bulk freight rates followed a downward but undulating path over time: They fell by 55% from 1850 to 1910, rose by 62% from 1910 to 1950, and fell – once again – by 71% from 1950, with a cumulative decline of 79% between 1850 and 2020.
- Behind these slowly evolving trends, there were often abrupt movements, with real dry bulk freight rates in some instances nearly tripling on a year-to-year basis.
We relate this secular decline to a historical literature which documents significant productivity growth as radical changes in goods-handling and storage in ports, naval architecture, and propulsion took place in the 19th and 20thcenturies (Harley 1988, Mohammed and Williamson 2004, Tenold 2019). Over the past 20 years, changes in corporate governance (Vuillemey 2020), fuel policy externalities (Keen et al. 2014), and industrial policy (Kalouptsidi 2017) may have further contributed to this decline.
Abstracting away from this long-run trend and its potential productivity-related determinants, we then turn to understanding the drivers of booms and busts in the dry bulk shipping industry, which occur at a relatively higher frequency.
We build on a canonical structural vector auto-regressive (SVAR) model with sign restrictions. Faust (1998), Canova and De Nicolo (2002), and Uhlig (2005) pioneered this model, which has become a standard of the applied macroeconomics literature. The same methodology makes it possible to set-identify the various shocks that drive real dry bulk freight rates at any one moment and might have an offsetting impact. Based on assumptions related to supply-and-demand analysis, we specify four orthogonal shocks to real maritime freight rates, which we interpret as a fuel price shock, a shipping demand shock, a shipping supply shock, and a residual shock.
In particular, we assume that a positive aggregate demand shock represents an unexpected expansion in global economic activity as in periods of rapid industrialisation and urbanisation. This, in turn, leads to not only higher global GDP, but also higher world mercantile tonnage, higher real fuel prices, and higher real freight rates. One key mechanism at work here is that an increase in dry bulk freight rates due to an increase in shipping demand triggers not only investment in new shipping capacity but also technological change in the wider industry, which augments effective supply. In contrast, a negative shipping supply shock represents an unexpected inward shift of the shipping supply curve. We associate such shocks with declines in world mercantile tonnage and assume that they negatively affect global GDP and real fuel prices but positively affect real maritime freight rates. Likewise, we assume that positive fuel price shocks negatively affect global GDP and world mercantile tonnage but lead to an increase in real maritime freight rates.
Finally, the residual term captures all remaining uncorrelated shocks, including changes in expectations. For our purposes, it can also be interpreted – at least partially – as a utilisation shock (see Kilian et al. 2020).
Based on the sign-restricted VAR model, we compute structural impulse response functions and historical decompositions for real dry bulk freight rates. The historical decompositions depicted in Figure 2 below are of particular interest: They show the cumulative contribution at each point in time of each of the four structural shocks in driving booms and busts in the market for dry bulk shipping services. Thus, they serve to quantify the independent contribution of the four shocks to the deviation of our real dry bulk freight rate index from its base projection after accounting for long-run trends in the same.
Figure 2 also allows us to visually discern the historical drivers of booms and busts in the dry bulk shipping industry. The vertical scales are identical across the four sub-panels, so that the figures clearly illustrate the relative importance of a given shock. Another way of intuitively thinking about these historical decompositions is that each of the sub-panels represents a counterfactual simulation of what real dry bulk freight rates would have been if driven only by this particular shock.
Figure 2 Historical decompositions of real freight rates
Notes: The chart shows the historical decompositions from the 68% joint highest posterior density sets obtained from the posterior distribution of the structural models, as in Inoue and Kilian (2013, 2019). The cumulative effects implied by the most likely structural model (modal model) are depicted in black. The results shown are based on 5,000 draws from the reduced-form posterior distribution with 20,000 draws of the rotation matrix each.
Table 1 more precisely quantifies these impressions by numerically summarising the contribution of each shock by period. Our results indicate that shipping demand shocks strongly dominate all others as drivers of booms and busts in real dry bulk freight rates. For the period from 1880 to 2020, shipping demand shocks explain 49% of the variation in real dry bulk freight rates while shipping supply shocks explain 22%. Thus, these two fundamental shocks, which are related to basic supply and demand conditions, explain a significant majority (71%) of the medium- and long-run variation in real dry bulk freight rates. Fuel price shocks and residual shocks respectively explain 11% and 18% of the same.
Table 1 Shares of shocks in explaining booms and busts in freight rates by period
Notes: Table 3 reports the share of variation in the real dry bulk index explained by the four structural shocks for the period from 1880 to 2020 and three sub-periods.
It is also possible to replicate this decomposition for shorter spans of time. Table 1 shows that the contribution of shipping demand shocks to variation in real dry bulk freight rates increased substantially in the interwar years and remained elevated in the post-WWII era. Likewise, the contribution of shipping supply shocks decreased substantially in the interwar years and remained suppressed in the post-WWII era. Finally, the contribution of both fuel price shocks and residual shocks remained roughly constant through the three sub-periods.
Canova, F and G De Nicolo (2002), “Monetary Disturbances Matter for Business Fluctuations in the G7”, Journal of Monetary Economics 49(6): 1131-1159.
Cosar, K and B Demir (2017), “Containers and globalisation: Estimating the cost structure of maritime shipping”, VoxEU.org, 13 June.
Cufley, C (1972), Ocean Freights and Chartering, London: Staples Press.
Faust, J (1998), “The Robustness of Identified VAR Conclusions about Money”, Carnegie-Rochester Conference Series on Public Policy 49: 207-244.
Greenwood, R and S G Hanson (2015), “Waves in Ship Prices and Investment”, Quarterly Journal of Economics 130(1): 55–109.
Harley, C K (1988), “Ocean Freight Rates and Productivity, 1740-1913”, Journal of Economic History 48(4): 851-876.
Heiland, I and K-H Ulltveit-Moe (2020), “An unintended crisis: COVID-19 restrictions hit sea transportation”, VoxEU.org, 17 May.
Inoue, A and L Kilian (2013), “Inference on impulse response functions in structural VAR models”, Journal of Econometrics 177(1): 1-13.
Inoue, A and L Kilian (2019), “Corrigendum to Inference on impulse response functions in structural VAR models”, Journal of Econometrics 209(1): 139-43.
Jacks, D S and M Stuermer (2021), “Dry Bulk Shipping and the Evolution of Maritime Transport Costs, 1850-2020”, CEPR Discussion Paper 15956.
Jacks, D S and J P Tang (2018), “Trade and Immigration, 1870-2010”, NBER Working Paper 25010.
Kalouptsidi, M (2014), “Time to Build and Fluctuations in Bulk Shipping”, American Economic Review 104(2): 564-608.
Kalouptsidi, M (2017), “Detection and impact of industrial subsidies: The case of Chinese shipbuilding”, VoxEU.org, 9 September.
Keen, M, I Parry and J Strand (2014), “The (non-) taxation of international aviation and maritime fuels”, VoxEU.org, 9 September.
Kilian, L, N Nomikos and X Zhou (2020), “A Quantitative Model of the Oil Tanker Market in the Arabian Gulf”, CEPR Discussion Paper 14798.
Metaxas, B (1971), The Economics of Tramp Shipping, London: Athlone Press.
Mohammed, S I S and J G Williamson (2004), “Freight Rates and Productivity Gains in British Tramp Shipping 1869-1950”, Explorations in Economic History 41(2): 172-203.
Stopford, M (2009), Maritime Economics, New York: Routledge Press.
Tenold, S (2019), Norwegian Shipping in the 20th Century, London: Palgrave Macmillan.
Uhlig, H (2005), “What are the Effects of Monetary Policy on Output?”, Journal of Monetary Economics 52(2): 381-419.
UNCTAD (2015), Review of Maritime Transport, Geneva: United Nations Conference on Trade and Development.
Vuillemey, G (2020), “Corporate irresponsibility in the shipping industry: A dark side of global trade”, VoxEU.org, 27 September.