Turkish Straits and an Important Oil Price Benchmark: Urals.

1. INTRODUCTION

   Oil tanker traffic through the busy Turkish Straits (the Istanbul Strait, the Sea of Marmara, and the Canakkale Strait) is notable. Approximately 4% of the world's crude oil trade and 3% of the world maritime trade passes through the Turkish Straits. There are four major crude oil grades passing through the Turkish Straits: Russia's Urals, Kazakh's Caspian Pipeline Consortium (CPC) blend, Russia's Siberian Light and Azerbaijan's Azeri Light. Wlazlowski, Hagstromer and Giulietti (2011) note that "Mediterranean Russian Urals is, in spite of little public and media attention, the third global price setter." after West Texas Intermediate (WTI) and Dated Brent. They also point out that "Mediterranean Russian Urals is a very clear price setter in the [medium] segment." (1) The busyness of the straits causes delays in oil traffic. The average waiting time of the oil tankers (sum of both ways) was 19 days each for the Istanbul and Canakkale straits in 2018, and this waiting time may increase up to 54 days combined. Since the price elasticity of crude oil demand is low, the straits carry a crucial strategic and economic importance in terms of oil trade for the oil-rich Russian Federation and oil-importing countries in the Mediterranean region for any possible supply-side disruptions for such an important benchmark. In this paper, we study how tanker traffic congestion in the Turkish Straits can affect oil prices.

   The empirical evidence gathered here suggests that congestion in the Turkish Straits increases Urals CTF prices for Mediterranean deliveries but similar supporting evidence could not be found for the important benchmark oil (Brent) or Iranian Light, which has similar characteristics and can be considered a close substitute for Urals in the Mediterranean refinery market. (2) This is the first paper that models the relationship between congestion in the Turkish straits and oil prices. To the best of our knowledge, this is also the first paper in the literature focusing on how congestion in a major chokepoint influences oil prices. We claim that the effect of congestion in the Istanbul and Canakkale straits on Urals prices differs above a certain level of congestion (13-15 days of waiting time) and increases oil prices up to 5.05% and 3.09% for the Istanbul and Canakkale straits, respectively.

   Oil has very inelastic demand and supply curves. Thus, a small change (or a shock) in quantity demanded or supplied will make a big change in oil prices. For this reason, following oil market trade flows is very important. While the academic debate on whether oil demand shocks drive oil prices more than oil supply shocks or vice versa continues. Caldara, Cavallo and Lacoviello (2018) argue that supply and demand are equally crucial for explaining fluctuations in both oil price and oil production. Furthermore, their estimate of oil price elasticity of supply is 0.08 on average, and the oil price elasticity of demand is also 0.08 (in absolute value) based on country-level, instrumental-variable panel regressions analyses. In addition to that, they estimate the oil price supply and demand elasticities as 0.08 and 0.14 in absolute value, respectively, based on a baseline structural Vector Autoregressive (VAR) model. Caldara, Cavallo and Lacoviello (2018) state that their oil price elasticity of supply estimates are within the range of other estimated elasticities in the studies available in the literature and in line with the estimates by Baumeister and Hamilton (2017) but nearly three times as high as those of Kilian and Murphy (2012) (3). Thus, a small disturbance in the straits concerning oil supply can have a sizable effect on oil prices. The existence of this supply side disturbance threat may have various other types of costs. For example, it can have after effects such a higher risk premiums for oil prices, higher bunker and transportation costs, higher inventories that need to be carried for any future disruptions, lower trade volume for oil as well as for other non-oil products, higher congestion for non-oil product shipments and higher risk of maritime accidents, which threaten the most populated city in Europe--Istanbul.

   Another set of evidence suggests that oil trade interruptions in the Turkish Straits affect oil prices. For instance, in March 1994, the oil tanker Nassia, carrying crude oil loaded in the Novorossiysk Oil Terminal in Russia, collided with the M/V Ship Broker at the North entrance to the Istanbul Strait. Twenty-nine people from Nassia's crew died, and 15,000 tons of crude oil were discharged into the sea, costing their insurers nearly $13 million as well as oil contamination in the sea waste. This accident caused a loss of almost $1 billion (Daly, 2008). Because of this accident, the Turkish Straits were closed to vessel traffic for 7 days from 13 March to 20 March in 1994, and during this period, Brent crude oil prices increased by 3.7%, and Urals Med 80KT prices increased by 6.95%.

   In order to incorporate the effect of congestion, we employ a Logistic Smooth Transition Model (LSTR) model such that higher waiting time in the Turkish Straits will force oil prices to follow another data generating process that is more likely to have higher implied prices. We consider two cases: the existence of congestion to such an extreme state that it closes the straits completely, or ships pass through the straits without any delay. Thus, we assume that crude oil prices follow different data generating (paths) processes under these two different situations. We also assume that the transition between the two states is a continuous non-linear process to allow regime-switching behaviour adequate for modelling the oil prices under congestion in the Turkish Straits. It is expected that above a certain threshold level of congestion (waiting time) in the straits, oil prices will increase and may reach a higher plateau in comparison with the no congestion scenario. The primary reason behind this expectation is that as the waiting time of the oil tankers at the entrances of the Istanbul and Canakkale straits (or congestion) increases, it causes oil supply delays, which lead to higher oil prices at the delivery port due to higher transportation cost as well as volatility. Therefore, the response of oil prices to increasing congestion in the Turkish Straits is often asymmetric and can be modelled as a non-linear process with a smooth transition.

   This paper is organized as follows. Section 2 elaborates on the characteristics and importance of the Turkish Straits. Sections 3 and 4 introduce the economic model and the data used in the analysis. In section 5, the empirical evidence is presented and the last section is for the conclusion. Appendix A introduces the econometric specification that we employ.

2. CHARACTERISTICS AND IMPORTANCE OF THE TURKISH STRAITS

   Considering the Turkish Straits' strategic location for the global oil trade, they have become one of the busiest waterways in the world (Mavrakis and Kontinakis, 2008). The Turkish Straits are extremely narrow in terms of physical characteristics in comparison with many other important straits that are used for international maritime trade, such as the Suez and Hormuz Straits. It is also one of the hardest waterways to navigate and famous for its sharp turns. Moreover, surface currents from the Black Sea to the Sea of Marmara have strong counter currents under the surface that create swirls and eddies.

   Connecting the Black Sea with the Aegean Sea, the Turkish Straits are comprised of the Istanbul Strait, the Sea of Marmara, and the Canakkale Strait. The Istanbul Strait is in the north-western part of Turkey and forms a boundary between Asia and Europe. It connects the Black Sea to the north with the Aegean Sea to the south through the Sea of Marmara. The Istanbul Strait's length is approximately 31km, and its average width is 1.5 km. The traffic volume in the Istanbul Strait is five times heavier than the traffic in the Panama Canal (Birpinar, Talu, Su and Gulbey, 2006). The Canakkale Strait is located to the south of the Istanbul Strait connecting the Sea of Marmara and the Aegean Sea. The Canakkale Strait is 70km long, and its average width is 1.3-2km (Aybay and Oral, 1998). It is common knowledge that the Canakkale Strait is one of the busiest, most dangerous and most difficult water passages for mariners to navigate because of its very narrow and winding shape and famous Cape Nara, which requires oil tankers to turn up to 90 degrees. The areas between the capes of Nara and Kilia of Gelibolu and the Cape of Cardak have strong currents which vessels should approach with caution by reducing speed or stopping to give way to oncoming vessels.

   Turkey has full sovereignty over the Turkish Straits. Since they are Turkey's internal waterways, Turkey has complete authority to regulate the maritime traffic there (Aybay and Oral, 1998). The Turkish Straits are also subject to the 1936 Montreux Convention, which defines their international legal status. Although the Montreux Convention confirms the principle of the freedom of transit and navigation by sea in the Turkish Straits ("Convention Regarding the Regime of the Straits", 1937), it does not enforce any restrictions regarding the merchant vessel traffic in the Turkish Straits. Turkey maintains its right to set Maritime Traffic Regulations and control all passage through the Turkish Straits.

   There are several factors that cause congestion in the Turkish Straits. The main factor is the restrictive transit regulations enforced by the Turkish authorities to ensure maritime safety in the Turkish Straits. Those restrictive transit rules relate specifically to vessels carrying dangerous/hazardous cargo, such as crude oil and petroleum products or derivatives. Vessels carrying dangerous cargo with an overall length of 200 meters or more (restricted vessels) are only allowed to transit during daylight...