WAVES AND OIL TANKERS —
DYNAMICS OF THE MARKET FOR OIL TANKERS TO
THE YEAR 2000
Jgrgen Randers
Norwegian School of Management
Hans Burumsvei 30, 1340 Bekkestua
Oslo, Norway
Introduction
The oil tanker market is interesting from a system dynamics point
of view. The market exhibits regularities which appear to be caused by an
underlying structure which has been stable for at least 30 years, and pro-
bably longer. This seemingly stable structure is primarily the result of
the systematic, but not particularly rational, behaviour of the main acter
in the oil tanker market: the community of shipowners. The collective
effect of their individualistic actions, I believe, is a rather violent
and rhythmic development in the market - on a timescale of years to decades.
The regularity is, of course, superimposed on a non-recurring pattern of
developments caused by events entirely outside the control of the oil
tanker community. In this paper I describe the stable structure and discuss
what it means for the likely development of the oil tanker market over the
next decade.
This paper is based on work done in the Resource Policy Group over
the last 6 years. I owe great thanks to Mess. Ulrich Golike, Christoph
Endress, Per Axel Prydz and Lasse Franck who have all toiled at various
times, to concretize what was initially a hunch to what is now a well
documented model of one possible explanation of the development of the
market for oil tankers.!
The market for oil tankers
As is well known, much cf the world's crude oil is produced far
away (in the Arabian Gulf and Venezuela) from where the oil is used (in
Europe, Japan and in the US). “T° get the oil to its destination, the oil
is pumped aboard oil tankers. These are large ships of ever increasing
size, currently capable of carrying up to 500 000 tons of oil in a single
load. Today there are around 1000 oil tankers in the world, varying
between approximately 20 000 and 500 000 tdw (tons dead weight). Together
they transport around 2000 million tons of oil every year over an average
distance of 6000 miles. This means that roughly one half of all the oil
used in the world has once spent up to a month on board a vessel travelling
across the high seas. Figure i illustrates the pattern of oceanborne oil
transport in 1978.
The oil companies own 40% of the world's oil tankers. “The rest are
held by independent shipowners - in Greece, Hong Kong, Norway, Liberia and
elsewhere. The independents let their ships to the oil companies, either
on a period basis ("time charter") or for a single trip ("Spot"). Typi-
cally, 40% of the tanker fleet is engaged in time charters of varying
duration, “only the remaining 20% (varying between 10 and 30% over the last
30 years) of the total fleet operates in the spot market, being available
for single trips. But in this small part of the market ordinary vari-
ations in the total supply and demand for oil tankers are amplified into
violent booms and busts. And here fortunes are quickly made and lost.
In the spot market one can fee] the pulse of the market, not only from
hour to hour, but from minute to minute. The spot market for oil tanker
transport is probably one of the best existing approximations to the
“perfect" market, with hundreds of brokers and thousands of telex lines
continuously transmitting information about available ships and cargoes.
-3-
The shipping crisis of the 1970's
The extent of oceanborne oil transport grew spectacularly before
1973. First because the industrialized countrics Consumed growing quanti-
ties ef oil in their cars, houses and factories, but also because they
exhausted traditional oil reservoirs near home and were forced to go
farther and farther away for new supplies. At the time of the OPEC
quadrupling of oil prices in 1973, the US shipped in 1/3 of its cil consump-
tion and Europe most of hers. Much of the oil comes from the Arabian
Gulf, 12 000 and 10 000 miles away, respectively. After 1973 the tanker
market has been less expansive. Both because growth in oil consumption
has been lower and because new oil fields have started producing nearer
home (Alaska, Mexico, The North Sea). The demand for oil transport over
the last 30 years is illustrated to the left in figure 2.
The period of stagnation throughout the 1970's did create problems
of overcapacity in the oil tanker market. There were too many ships
available competing for a limited number of cargoes. The immediate reason
is obvious: Tankers must be ordered between 1 and 3 years before they are
delivered from the shipyard. And the boom years just prior to the OPEC
embargo gave the shipowners all the cash and ail the incentive necessary
to place orders for new, and larger, ships. At the peak, in late 1973,
there was nearly as much tonnage on order as there was tonnage afloat.
Ang once an order has been placed, it is expensive to cancel. Rather than
actepting "unnecessary" cancellation charges, the shipowners chose to hope
for a brighter future and took delivery on a lasting stream of new ships,
even though many went directly to year-long stays anchored in some remote
Norwegian fjord or. elsewhere.
The ships that still ‘did transport oil received very low rates.
As a consequence they often moved at reduced speed ("slow-steaming") to
minimize. fuel costs. At the bottom'of the shipping crisis the actual
transport work performed was only one half of the capacity of the existing
fleet. The other half of the tonnage was either laid up or absorbed in
slow-steaming. In short, capacity utilization of the world's oil tankers
was down to 508. 3
The development is visualized in the top part of figure 3. Today's
(1981) situation is a continuation of the gloomy 1970's, although the
order backlog for new tonnage now is minimal.
Recurring crises
The conventional wisdom of the shipping community is. that the cri-
sis of the 1970's was a one-time affair caused by unpredictable develop-
ment outside the control of the community, namely the "cil crisis" created
by OPEC.
7 I do not think that conclusion holds. If one looks further back
in time, the developments following 1973 seem to be mirrored by the
events between 1956 and the middle 1960's. Both periods display the same
pattern: An initial couple of years of exceedingly high profits, followed
by phenomenal growth in the order backlog for oil tankers, followed by a
lasting period of low rates, significant lay-ups, and low capacity utili-
zation. The depression lasts for as long as the steady flow of new ships
create and maintain overcapacity. It only ends when growth in the demand
for oil transport and scrapping of o1@ ships finally restore balance in the
market. The development from 1956 onward is also visualized in figure 3.
The similarity to the 1970's is obvious.
If we go even further back, verbal @escription tell of a similar
sequence, of events in the 1930's, which gave Tjalling Koopmans an incen~
tive to write his classic Tanker Freight Rates and Tankship Building in
1939,
In summary, a decade long period of overcapacity seems to bea re-
curring phenomenon in the market for oil tankers.
More waves in the tanker market
The recurring periods of overcapacity show up, of course, in the
time development of the freight rate (that is, in the price paid for trans—
port:of.:0il, measured in $/ton-mile). Figure 4 portrays the actual
development of the freight rate over the last 30 years’.
During the two periods of overcapacity, lay-ups and slow-steaming,
(1958-1967 and 1974 till today), the freight rate remained consistently
low - as one would expect. The Worldscale index fluctuated around 40.
Whenever the index increased some, new ships were lured out of lay-up
while others increased their speed. And predictably, the Worldscale
numbers once more were pushed back to the minimum level that could keep
sufficient tonnage operating.
To the system dynamicist, however, figure 4 is more interesting
for what it says about the intervals between the periods of overcapacity.
It would be reasonable to expect intervals of undercapacity - intervals
‘with a lack of tonnage as a result of insufficient investment in new ships
during the preceeding period of overcapacity and low profits. But figure
4 does not show a high and stable rate during these intervals (1950-1957,
1967-1973). ‘Rather we see violent fluctuations between peaks near Worlid-
scale 400 (where a few month-long trips yield sufficient profit to pay
down an entire tanker) and troughs below Worldscale 100. So although the
average rate during these intervals is higher, it is very unstable.
Our central hypothesis
The instability is not random, however, at least for one who wants
to see regularity. The rate appears to fluctuate with a four-year period
which immediately brings to mind inventory oscillations and the 3-6 year
business cycle. And this was the initial hunch behind our study of the
dynamics of the oil tanker market. We interpreted the freight rate develop-
ment as a sum of one 20 (actually 15-20) year wave and one 4 (actually
3-6) year wave, as sketched in figure 5. The top curve is ment to repre-
sent the 4 year business cycle which we assumed is felt in the tanker
market as a 4-year oscillation in the demand for oil. The middle curve
shows a 20-year investment cycle, where we assumed that the observed 10-
year periods of overcapacity are succeeded by intervals of undercapacity
in the supply of ships. The lower curve in figure 5 shows the interaction
of the two waves.
Notice that we assumed that the business cycle does not show up in
the freight rate during periods of overcapacity. When there is still free
capacity, ~ ships laid-up or going at less than full -steam - increased
demand simply leads to increased supply, possibly after a brief period of
higher rates. But in the intervals of scarce capacity, when there are no
lay-ups Or Silow-steaning:, even small demand increases will push rates sky-
high. This was our hypothesis for the asymmetric ‘ rate development seen in
figure 4 - our reference mode.
We did kuow, however, that the amplitude in the fluctuations in
GNP and hence in the demand for oil is at most * 5% over the 4-year cycle.
The freight rate, on the other hand, varies with several hundred per cent.
One possible explanation could be the existence of a 4-year oscillatory
tendency in the tanker market itself, tending to amplify the swings in
the demand.
Further, an investment wave period of 20 years seemed long given
that the order delay for new ships rarely exceeds 3 years, and is around
1 in the normal situation with ample shipyard capacity. One possible
explanation could be that perception and decision delays beccme especially
veag because the market development appears.random, violent and confusing
to the community of shipowners.
The basic structure
Much work, discussion with industry people, review of existing
literature, and numerous tests and revisions of our model, have led us to
believe in the basic structure below. In short, this is a structure
dominated by a 20-year investment loop with very significant decision
delays. There is, however, no Single strong 4-year loop {although a
number cf weak ones) and we have concluded that the violent fluctuations
in the freight rate in response to soft fluctuations in the demand for oil
is largely a consequence of very inelastic supply once the oil tanker
market is near or above full capacity utilization.
The basic structure of our model is shown in figure 6.. This
apparently trivial diagram warrants a number of comments.
1 First, the freight rate is assumed to depend on the demand for oil
transport compared to the capacity of the existing fleet of tankers. The
higher the demand, the higher the rate. The important point here is our
belief that the freight rate depends on demand relative to the capacity of
the whole existing fleet, and not relative to the capacity of the tonnage
that is currently sailing. In other words: we assume that freight rates
will not soar as long as there is capacity available in lay-ups or slow-
steaming, even if the ships that do sail are working at full capacity. The
experience of the 1970's establishes this view beyond dispute.
Still, a short term dynamic exists, whereby the freight rate goes
up temporarily when the sailing fleet is pressed beyond reasonable utili-
zation. But the higher rate will imneaiately draw additional ships out of
lay-up. Soon there will be enough tonnage sailing to perform the current
oil transports with reasonable vessel utilization and.- hence - fora
reasonable rate. This short term dynamic can be seen as a tendency for a
i-year oscillation in figure 4. It is included in the model.
Second, returning to figure 6, the demand for oi] transport ob-
viously depends on the consumption of oil, which in turn depends on the
level of. economic activity. An important structural element, however, is
the fact that there is no (or at most, a very weak) link between the freight
rate and the demand for oil transport (dotted in figure 6). The cost of
transporting oil (usually below 1 $/barrei) is so much lower than the value
of oil (around 35$/barrel), that even record high rates have little effect
on the market price, and hence consumption, of oil.
Third, we believe there is some effect from the inventory of oil,
decoupling the consumption of oil from the oil transports performed. We
see all oil in transit between oil field and end consumer as the relevant
inventory. The size of this inventory appears to be between 3 and 6
months - 1 month's supply are on board the oil tankers; refineries and
retailers store another 1-2 month's sales, while the consumer stores the same
in his house or factory. When there is much oil in transit, we would
expect this to reduce the orders for new cargoes of oil. The delay around
the loop is roughly one year. The loop therefore is interesting because.
it might amplify an exogenous 4-year oscillation in the oil consumption.
We have, however, not yet been able to gather the necessary statistics to
investigate this possibility. Oil transport statistics do not exist in an
aggregated form on a quarterly (or monthly) basis, and the industrialized
countries did not start to gather aggregated inventory statistics until
very recently. So although included in the model, the relation between
inventory and the demand for oil transports has been tuned so as not, to
generate a dominant ‘-ycear oscillatory tendency.
Fourth, we believe that the impact of the freight rate on the
market for.oil tankers is primarily on the supply side. Given the cost
structure of the ships, the freight rate determines what is optimal be-
haviour for the shipowners concerning whether to lay up, to slow steam,
to perform extended maintenance and repairs, etc. The aggregated result
of the individual decisions of hundreds of shipowners. is a certain program
of utilization for the existing fleet. In short: the freight rate deter-
mines the capacity utilization of the fleet. The effect is quick: the
decision to slow-steam can be made within hours (if the contract allows),
and all decisions on vessel utilization can be redone when the current
contract expires & %
Fifth, as long as the fleet is operating below capacity, the
amount of oil transported will be equal to the demand for oil transports.
Over the last 20 years, there has only been a very few months where the
existing Fleet operated above 80% of capacity. In other words, there has
consistently been ample spare. capacity, and shipments have equalled orders.
One likely effect of this is that fluctuations in the oil consumption do
pot result so much in varying oil inventories as in varying capacity utili-
zation in the tanker market. The market may, so to speak, be the buffer.
And not only absorbing the inventory oscillations, but also the
4-year business cycle. We have done much work to try and illustrate this
effect statistically, but even our best results remain unitipreawive due to
lack of data and the shrouding effect of technical developments”.
Sixth, we assume that the total transport capacity, i.e. the total
number of ships, develops primarily in response to the shipowners' percep-
tion of the capacity utilization. High freight rates may tempt owners to
= 10 -
order new ships, but high rates is not enough, we believe, to make him
commit tens of millions of dollars. First he will make sure that all exis~
ting ships have (or will have) employment, i.e. that capacity utilizat
is (or will be) at acceptable levels.. Only then will he go ahead and place
his. order.
Still, we do not imply that the ordering is a cool and rational
Process. . There are large uncertainties associated with demand development,
future technology, scrapping of old ships, and potential cancellation of
orders. And there are psychological effects which isles it difficult to
remain unperturbed by the current mood of the trade. All these factors
:may add up to Large scale ordering even when cool analysis reveals over-
capacity in the near future. But basically we believe that ‘new ships are
ordered when continuation of current trends indicates excessive levels of
capacity utilization within a couple of years.
Finally, we stress the dominant role of the investment loop (freight
rate —» capacity. utilization —3 total transport capacity —> freight rate)
cin figure 6. The total number of ships can only be changed through orders
for new ships or scrapping of old ships. These are slow processes. Still,
I believe, they dominate developments - even if low freight rates may result
in quick cancellations of new orders, and high rates may lead to immediate
postponement of scrapping plans. Thus, in very rough terms, I see the oil
tanker market.as a 20-year investment loop, driven by an exogenous demand
for oil which includes a 4-year business cycle component, Short term
variations in the demand for oil transport are reflected in the capacity
utilization.of the fleet, and may be amplified some by endogenous inventory
dynamics, During periods of scarce transport capacity, the variations also
show up in the freight rate. =
Our model reflects these views.
ae 1g &.
The model
The current version of our DYNAMO model (documented in GRS-315,
see note 1) of the market for oil tankers comprises 300 equations. Diffe-
rent parts of the model describe the demand for oil transport; the capacity,
age structure and cost structure of the existing fleet; the shipowners’
decisions on how to utilize the fleet; the formation of expectations -and. the
Gecisions. on investment in new ships/scrapping of.old ships; and, finally,
the ability of the shipyards to handle new orders for ships.
Most, of the formulations have been discussed extensively with
practitioners in the shipping industry. Consequently, the equations are more
Girectly descriptive of real world procedures and less "streamlined system
dynamics formulations". Industry wording have been used throughout, and there
is little disagreement about the centrality of the relations that are. included
in the model. Practitioner criticism typically focuses on a number of mechanisms
that have not been included. These mechanisms were excluded because we do not -
believe they are sufficiently stable over time to be part of a fundamental
explanation of the causes behind our 30-year long reference mode. In short,
the detailed model formulations and the main parameter values have passed
the test of being plausible to practitioners. One evidence is’the fact that
the Norwegian Ministry of Trade and Shipping does make model runs as an input to
decisions which require an opinion on the long term development (say, to 1990)
of the market for oil tankers.
The model was tested by trying to reproduce historical tanker market
behaviour from 1953-1980. The model was initiated with the values for 1953
and only subjected to the actual rates of growth of the exogenous variables
{oil consumption and average transport distance) and random disturbances in
oil consumption. The result is summarized in figure 3.
As can be seen from the figure the model is able to recreate the
observed pattern of 10 years of overcapacity succeeded by intervals of under-
e42<
capacity. Furthermore, lay-ups, tonnage. on order, and the growth of the
tanker fleet are also reproduced. The phase relationships observed in
reality, high lay-ups at the beginning of the extended depression and
increasing orders during successive booms are also reflected in. the model
simulation. The model also does recreate the wild swings in the freight
rate, although only driven by a very soft tendency to a 4-year cycle in
the historical. figures for global. oil consumption. In short, our relatively
simple hypotheses are capable of recreating the rather complex historical
developments over the last 30 years.
Finally, the model system is robust against reasonable changes,
both in parameters and exogenous functions, This will be seen from the
model experiments discussed below.
A system's view
But. even without model experiments, we can draw some "system's"
conclusions about the. market for oil tankers. These are conclusions that
rely on the basic system structure and therefore, I believe, will remain.
valid for a long time, In a way these conclusions are the main practical
results of our study, because they can be transmitted and defended in words.
1. The level of activity in the tanker market is primarily determined
by the demand for seaborne transport of oil - which, of course, is
determined outside the shipping sector. In short, the volume of the
tanker warket is determined by the rest of the world.
2. The profitability of the sector, on the other hand, is determined
by the degree of match between supply of and demand for oil transport.
And the supply (the number of ships and the way they are used) is,
of course, fully determined by the shipowners themselves. In short,
the sector determines its own profitability.
3. Flippantly, but with a large element of truth: in the future the
level of activity in the tanker market will be determined by the
world's demand for oil transport, while the profitability will
be determined by the shipowners' supply of oil tankers.
‘When freight rates are low, it signals less than perfect ¢co-
operation among shipowners. This is particularly true when
‘ considering that transport costs represent a minute fraction
of the cost of oil.
4. Major freight rate increases will only occur during intervals of
undercapacity (i.e. when most vessels are fully employed). This
is true regardless of whether the boom is caused by a business
cycle upturn or by random events like the closing of the Suéz Canal
in 1967. I£ such boom impulses occur during a period of signifi-
cant hay-ups and slow-steaming, they will hardly result in anything
but fewer lay-ups and higher speed.
Application of the system's view
These general, structure determined, conclusions can be applied to
the current situation in the oil tanker market.
As mentioned, the oil tanker market is presently in its 7th year of
a period of overcapacity. Still, the equivalent of 40-50 miilion tons dead-~
weight is absorbed in lay-ups and slow-steaming (out of a total fleet of 350
million tons dw). The next interval of scarce capacity will not begin until
this overcapacity has been completely eliminated. This.is unlikely to occur
before 1982-3, and no soaring freight rates can be expected before then.
It is worthwhile to stress that I believe this conclusion is largely
independent of the growth (or, more likely, decline) of seaborne oil transports
over the next couple of years. The time of “balance” in the market is deter-
mined by the supply side, and will, roughly speaking, remain equally far in the
future regardless of the rate of growth of demand. This is nothing but a re-
statement of conclusion 2 above: profitability. is determined by supply, and
supply will grow faster if demand grows faster. Opposite (and more likely),
ships will be scrapped faster and new orders cancelied more frequent.; if the
demand for oil transports softens even further.
-
Oil transport in the future
We have seen that the model can reproduce the past. And we have
made, stat nts ahout the future of the oil tanker market based on our
understanding of the structure of the ‘system. Let us now use the DYNAMO
model to trace possible futures.
-In. order to run the model forwards, we need to make assumptions
about the future development of the exogenous variables - primarily the
determinants of the demand for oil transport.
To this end we undertook a detailed study of the global petroleum
market (reported in GRS-214, see note 1). We split the world in 10 regions
and studied oi1 production and oil consumption in each region. Same regions
are net importers and other net exporters, and we assumed that the deficit
regions ship in necessary oil from the nearest exporting region. Needless
to say, the transport patterns can become complicated, and we used a linear
program to establish the pattern involving minimal transport costs.
Comparison with historical data proved that the sciution of the LP was within
5-10% of the transport pattern actually used.
The LP solution can easily be converted to the inputs needed in the
oil tanker model, namely global oil consumption, amount shipped and average
transport distance. As production and consumption of oil‘change over time
in the different regions, the LP solution will change and give time series
for the same inputs.
‘Figure 7 shows the result of using this procedure to calculate likely
developments to the year 2000. The figure is based on available statistics
up through 1978 and model calculations from then on. We see that the total
oil transport performed actually declines from 1976 to 1980. Then there is
slow growth toa peak in the 1990's, followed by a new decline. The decline
in the late 1970's is caused by the decrease in the transport distance
-~15-
which follows from the opening up of oil fields closer to the consumer
nations. The decline in the 1990's is caused by beginning exhaustion
of the world's petroleum reserves.
In the base case (solid line in figure2) we assumed economic Growth
rates around 3%/year. (Corresponding to growth in oil consumption around
2%/year.) If we assume higher rates, the peak transport work occurs earlier,
because the oil reserves are exploited faster. Lower economic growth
stretches the era of oil transports. Possible extremes in the use of canals
and pipelines to reduce shipping distances introduce about the same uncer-
tainty in the projection as the variation in economic growth rates.
It is interesting to notice that the maximum transport work (around
15,000 billion ton-miles/year in the middle 1990's} is not larger than it
could have been handled by today's fleet of oil tankers. The present over-
capacity is significant, to say the least.
Model generated futures
Figure 7 shows the result of running the oil tanker model to 1990,
from a new initialization in 1972, using the base case assumptions regarding
the demand for oil transports (with a soft business cycle and a small amount
of noise superimposed) .
The result is as expected. The current period of overcapacity is over
by 1983, and soon thereafter we see tendencies for rate increases whenever
boom impulses hit the market, now having a high capacity utilization. But
the period of acceptable freight rates does not last long before it triggers
a new wave of optimism, cash and orders for more tonnage. As in 1974 and 1958
a wave of new ships stops the bonanza (in 1988).
= 16 =
It is worth pointing out that the bonanza occurs in a period of
declining tonnage. The total fleet decreases steadily from 1979-1988,
but this Goes not, of course, prevent high freight rates. In the tanker
market profitability and volume are largely decoupled.
As a final curiosa it might be mentioned that figure 7 was made
in February 1980, 14 years ago. In the intervening period the real world
events have deviated little from the model run.
Alternative futures
To test the sensitivity of the world, we made an experiment assuming
an. unrealistically high growth rate (7%/year from 1981-1990) in oil con-
sumption. The result is. shown in figure 8.
The market experiences a stronger boom than in the base case, but not
at an earlier point in time. Because supply expands when demand does. The
behaviour of the ship owners makes the system robust.
Like in all preceeding crises of overcapacity, the tanker industry :
is discussing a program of accelerated scrapping of old-ships. Figure 9
shows the effect of removing from service each year twice the normal tonnage.
The market recovers earlier, but the higher rates attract large in-
. vestments in new ships, and by 1990 the market is once more on its way
into a.period of low rates and lay-ups. And no one-has yet explained
how one makes 1000 viciously competitive and individualistic shipowners
agree on a program of scrapping.
Postscript
As can be seen from this paper, I view much of the development of
the tanker market es determined by systematic behaviour in the shipping
industry. Many will disagree and explain the events of the past 30 years
-17-
as the result of a sequence of uncorrelated occurences.
The main insights to be gained from my perspective are descriked
above. They add up to.a belief that in our imperfect world - where people
(including shipowners) do not easily collude cr cooperate - the actions
necessary to perturb the rhythmic development of the tanker market are so
gargantuan that they.will not.be realized during the {relatively few)
decades when oil will still be shipped.
NOTES
The main. reports from the study are:
1, Ulrich Golfke, Tanker Futures: A model of tanker market dynamics,
GRS-141, April 1978, 90 pages.
2. Ulrich Golike, Four-Year and Twenty-Year Cycles in the Tanker Market,
GRS-147, May 1978, 10. pages.
3. Christoph Endress, The Demand for Tanker Transport up to Year 2000,
GRS-214, June 1979, 124 pages.
4. Per Axel Prydz, Lasse J. Franck, Ulrich Golike, Bglger i tankfarten
~_Om tankmarkedets virkemate og fremtid, GRS-284, April 1980, 88 pages.
5. Lasse Franck, Per Axel Prydz, Waves in the Tanker Market (Technical
Report), GRS-315, October 1980, 110 pages.
All reports are available from The Resource Policy Group, Sagveien 21,
Oslo 4, Norway.
A ship of 100 000 tdw is capable of carrying approx. 100 000 tons of oil.
Actually, figure 1 exhibits a dimensionless index number called World-
scale. The Worldscale index is defined as the ratio between the freight
rate (in $/ton-mile) and a “standard” freight rate (in $/ton-mile) which
is adjusted continuously with inflation and technical developments in such
a manner that Worldscale index 100 equals a rate which gives a reasonable
return on investment in a tanker of 50 - 100 000 tdw. Bigger ships have
lower cost per ton-mile and can operate with a profit at much lower World-
scale numbers.
We define vessel utilization (ton-miles/tdw- year) through
vessel utilization
designed vessel utilization
capacity utilization = (l-fraction laid up)
Designed vessel] utilization (ton-miles/tdw-year) is the amount of transport
work that can be done per ton deadweight each year assuming normal speed,
harbor times, maintenance etc. Figure 6 seems to imply a one-to-one
correspondance between the freight rate and capacity utilization. This
is incorrect, both in the model and reality, and simply a consequence of
simplification. The real structure is shown below and is capable of re-
- 19 -
creating the main characteristic of the recurring 10 year depression,
namely gradual increases in the capacity utilization as ships come out
of lay-up while the freight rate remains essentially constant.
VESSEL
UrmLIZaTION Ny
CAPACITY
UTILIZATION
a OF FLEET
FREIGHT FRACTION
RATE LAID: UP
DESIRED
VESSEL
UTILIZATION TOTAL
'TRANSPORT
CAPACITY
DEMAND
aRANSPORT o
‘TRANSPORT
CAPACITY
5. See figure 9 and Appendix B in GRS-284 for a heroic effort to construct
a 20 year time series for the capacity utilization.
nautical miles
8000
4000-1
2000.
billion ton-miles/year
Figure 1.-Pattern of ocean-borne oil transports in 1978.
a
16000: T T TT high: E
Average transport distance moderate
(nautical miles). A ,
awl
12000. %. i oe
uo ae
1
\
i
iN H
F “Oil transports performed |
(ton-miles/year) i
1 .
4000 t
° ~
1950 1960 1970 1980 1990 2000
Figure 2. Global oil transports: historical development (1950 - 78)
E and model calculation (1978-—.2000) for moderate,.high and
low economic growth rates. .
sy otal fleet’ (mill. tons dw)
“ coed Scale: 0-400
{Oil shipments (mill, tons/year)
Scale: 0-2400
| ~Fraction of fleet laid up (%)
Seale: 0-20
Capacity utilization of fleet (4)
Scale: 60-120
Freight rate index (Worldscale)
Scale: 0-S00
ips on order (mill. tons dw)
Scale: 0-200
Total fleet (mill. tone dw)
|" scate: 0-400
|_-Oi1 shipments (mill. tons/year)
Scale: 0-2400
Fraction of fleet laid up (%)
Scale: 0-20
Capacity utilization of fleet (%)
Scales 60-120
‘| Freight rate index (worldseale)
Va Seale: 0-500
Ships on order (mill. tons dw),
Seale: 0-200
1973 1978
Piaure 33. Historical (top) and simulated model (bottom) behaviour.
-1%-
- 22 -
Freight rate index
400 (Worldscale)
Toe cons oa [er a er
1950 1955 1950 1965 1970 1075 i980
Figure 4. The Worldscale index for oil tanker freight rates, 1950-80.
Worldscale=(current rate in $/ton-mile)/("normal" rate in $/ton-mile).
Sp eee fe eee es a ——~- ---- Trend
b) 20-year investment wave in the supply of oil tankers
+t +
50 Iteo mo 190
c) Resulting pattern in the freight rate.-
Figure 5. The effect on the development of the freight rate
of. the interaction between a 4-year wave in demand anda 20-
LEVEL OF ECONOMIC
ACTIVITY
OIL PRICE
Ms d CAPACITY
Si UTILIZATION
orn nN , OF FLEET
CONSUMPTION FREIGHT
RATE
- -— eo
> —
OIL IN DEMAND FOR TOTAL
TRANSIT ‘OIL TRANSPORT TRANSPORT
OIL TRANSPORT CURRENT 4
PERFORMED ee TRANSPORT
CAPACITY
Figure 6, Basic structure of our model of the market for oil tankers.
-€%-
Oil shipments (mill. tons/year)
Scale: 0-2400
me,
INrotal fleet (mill. tons dw)
Scale: 0-400
| Aapacity utilization of fleet (%)
Scale: 60-120
Pant Q P 7
\ ee nnny Freight rate index (Worldscale)
Scale: 0-500
penne
{
Fraction of fleet laid up (%)
Scaie: 0-20
f—Ships on order (mill. tons dw)
Scale: 0-200
1973
1978 1983
Loil shipments (mill. tons/year)
Scale: 0-2400
Total fleet (mill. tons dw)
Scale: 0-400
Capacity utilization of fleet (%)
Scale: 60-120
ca
__f
a
‘. reight rate index (Worldscale)
paneer
Scale: 0-500
Fraction of fleet laid up (%)
Scale: 0-20
hips on order (mill. tons dw)
Scale: 0-200
Figure 8. Run assuming 7 %/year growth in oil consumption. .
|-Oil shipments (mill. tons/year)
? Scale: 0-2400
Total fleet (mill. tons dw)
Scale: 0-400
Capacity utilization of fleet (%)
Scale: 60-120
ln“, Pa
ee Freight rate index (Worldscale)
nanan
Scale: 0-500
raction of fleet laid up (8)
Scale: 0-20
ips on order (mill. tons dw)
Scale: 0-200
978 "$983 1888
Figure 9. Run assuming dovble scrapping rates.
