Middle East Airlines – Strategic Diversifiers or Serial Destroyers of Capital? (Part 1)

In 2019 one would be hard pressed to find a frequent international traveler who hasn’t either flown with or heard of one of the four large Middle Eastern airline carriers (or ‘super-connectors’ as The Economist magazine coined this category in a recent article). The small group of airlines includes: Emirates Airlines, Qatar Airways, Etihad Airways and Turkish Airlines; we collectively refer to them in this post as ‘ME Carriers’.  

In the last 15 years (2004-2018), these four carriers led a region that has added more seat capacity (available seat kilometers or ‘ASKs’) and flown more incremental passenger kilometers (revenue passenger kilometers or ‘RPKs’) for international routes than any other region globally.  The region's share of global ASKs rose from just 2.6% in 2004 to nearly 9% by the end of 2018, accounting for a staggering 14.7% of global capacity additions during this time, only marginally lower than the capacity growth printed by the top four Chinese carriers.  Not only was this growth in capacity additions world-beating (14.9% CAGR 2004-18), these airlines also collectively dominated the widebody segment of the aircraft market, with over 75% of widebody purchases during this period.  According to Boeing, as a result of this voracious growth, not only do the ME carriers currently boast one of the youngest aircraft fleets on the planet, the region also accounts for close to half (46%) of global future widebody deliveries to 2038.  So, if you haven’t flown on one of them in the last decade, chances are you will in fact be a passenger in the coming decade. These airlines, which were relatively obscure national flag carriers as early as the late 1990s, have grown into several of the largest and most recognized airline brands on the planet.

This incredible growth story is also a multi-faceted one.  Their unique evolution elicits several important thematic questions, ranging from the relative efficacy and efficiency of sovereign-led allocation of capital to the merits of the chosen ‘super-connector’ business models that these state-owned carriers have pursued.  Since the airlines’ chosen competitive landscape is also a truly global and connected one, this is also a story of rampant and burgeoning globalization at play – the winners and losers from the rapid expansion of global human connectedness, mobility and trade.

In this post, we will attempt to not only analyze the progression of this growth phenomenon over the past several decades, we will also try to address a number of critical questions that have been ‘front of mind’ for us since the industry first caught our analytical attention almost a decade ago (having been avid consumers of their travel offerings since the late 1990s): 1) What are the implications of the ME carrier-driven capacity growth on the supply and demand dynamics of the global airline sector? 2) Are the strategies underpinning this growth sufficiently differentiated and sustainable? and 3) Has this explosive investment in capacity created sufficient economic value added (‘EVA’) and/or generated meaningful positive externalities for each of the national stakeholders (shareholders, citizens and consumers alike) that in turn justifies the deployment of this enormous magnitude of resources?  The central question of sustainability in turn begs addressing another broader question - have these investments in fact been effective diversifying strategies for the host ME countries thus far or is the jury still out on this assessment?

The first two topics have been the subject of intense global debate recently, not only amongst the different national airline regulators, but also among the varying competing legacy and emerging airline carriers around the world.  As one of the new destabilizing forces in the global aviation sector, the ME carriers have been the target of widespread analysis, criticism, envy and even downright fear as they have leapt into the top ranks of global international carriers in a relatively short period of time.  The last set of questions however are the most critical ones for us as GCC and Turkey observers and investors and thus it is where we spent most of our analytical energy.

This topic also ties into our series of posts on the GCC Growth Model which we introduced earlier this year and fits nicely into our narrative arc where the merits of the GCC’s deployment of its windfall capital is tackled. The aviation sector (airlines, airports and related industries) has been an early and prime recipient of these sovereign capital flows since the 1980s-90s while also commanding an important strategic priority from several of the GCC countries as an instrument of- and catalyst for- diversification away from hydrocarbons.  The aviation sector, like oil & gas exporters, petrochemicals, fertilizers and sea ports, is a globally integrated and competitive sector which the GCC countries have strategically targeted to compete in.  As a result, there are a number of important implications that go well beyond the Gulf or even the Middle East regions.  We therefore thought that this sector deserved special attention in our narrative arc for the GCC.

Before we dive into the ME carrier evolution and analysis, it is useful to take a giant step back and look at the global airline sector from 100,000 feet for context.  

During the past 47 years, the volume of air travel (measured by worldwide RPK volumes – see Chart 1 below) has expanded almost 17x (6.2% annual average growth rate), far outstripping world trade growth (5.0%) and world GDP growth (3.1%).  Not only is this growth rate one of the highest and least volatile amongst global sectors, it is also testament to the rapid pace of globalization during these five decades.  One could argue that airline connectivity is in fact also a big contributor to globalization itself, acting as a key enabler for global trade, human connectivity and capital mobility.  As longstanding investors in emerging markets, we cannot remember the last time we felt comfortable deploying capital into a country before visiting it at least 6-12 times.  Capital and know-how invariably follow human connectivity and mobility, which in turn should eventually drive productivity-led growth – or so the economic theory would suggest. 

Noteworthy in Chart 1 is the steeper slope of the growth curve of traffic in the aftermath of the global financial crisis of 2008-9, where RPK growth (6.8% from 2009-17) far outstripped both trade growth (4.7%) and GDP growth (3.0%).  This indicates a higher average income elasticity of demand in the past decade, a reversal of the trend in demand elasticity witnessed in the prior decade and most likely driven by accelerating emerging market demand. Whenever we see such a parabolic acceleration of growth, we tend to be a bit weary of its short-run sustainability.  More on this later in the post when we look at how capacity growth has responded to accelerating traffic demand.                                                                                                                                             

Chart 1 – Global Air Traffic, Capacity, Passengers vs. Trade & GDP – 1970-2017

On the point of demand elasticity, Charts 2 and 3 below paint a snapshot of where current travel penetration is relative to income levels, while Chart 4 plots the past 15-year growth in traffic relative to current levels of per capita income.  At the $5k-$10k income per capita range, markets enter a steeper part of the ‘S’ curve in travel demand, while demand growth tends to plateau only beyond the $20k level, suggestive of more mature demand dynamics relative to increases in disposable income.  Most of the developed markets within the OECD have in fact begun to approach saturation, with the larger passenger markets (in Chart 2 and Chart 4, the size of the bubble is indicative of the national passenger market size) of Europe, North America and Oceania in particular starting to level-off in terms of incremental demand growth.   Many of the larger emerging markets have entered the high growth corridor over the past 10 years (e.g. China, Russia, Brazil, Turkey, Mexico, Argentina, Malaysia) and look set to continue to contribute disproportionately to global demand growth in the upcoming years before they eventually plateau.

The group of large population countries behind them (income levels of $3k-$10k), including Thailand, Indonesia, Colombia, Peru, Philippines, and Vietnam, are currently experiencing the fastest acceleration of growth.  At the far left of the income spectrum (< $2,500 in per capita income), we have the large South Asia economies (India, Pakistan, Bangladesh) and the big Sub-Saharan African countries of Nigeria, Ethiopia and Kenya.  Despite some of the highest growth rates globally (most from very low base levels), some remain in a pre-acceleration phase.  All however hold tremendous potential to contribute to future growth as their disposable income re-rates and they have ample discretionary spending power to afford travel.  According to Boeing’s CEO, Dennis Muilenburg, in a recent interview on CNBC, over 80% of the world’s population has never taken a single flight, with over 100 million people in Asia flying for the first time in 2017 alone.  The well of underpenetrated opportunity for growth is truly enormous. 

Finally, it is important to highlight the countries in the upper right quadrant of Chart 2 as they are the real present-day outliers (Ireland, Qatar, UAE, Singapore, Hong Kong).  All of them are sparsely populated states (4 of the 5 are in fact city-states) that have two main things in common: 1) they have invested tremendous resources in cultivating a leading international aviation sector and 2) they have fed off the un-met demand of one or several large regional neighbor markets, effectively tying their growth path to them (e.g. Europe for Ireland, China for Hong Kong and Singapore, S. Asia for the UAE & Qatar).  

The same group however when viewed through a lens of the past 15-year growth (Chart 4) start to exhibit a divergence.  Hong Kong and Singapore have converged to the normal decelerating global growth curve, most likely due to the robust endogenous growth in capacity that China has experienced of late.  Qatar and the UAE continue to leverage S. Asian spill-over growth, while Ireland (the home of the dominant European low-cost carrier or ‘LCC’ Ryanair) also continues to grow at a large premium due to persistent LCC share gains within a mature European market.  We believe the LCC business model is not only in fact a superior growth story, but has a powerful secular driver propelling it with real ‘legs’.  We will discuss the LCC phenomenon later on in the post when we look at global airlines side-by-side through a ROIC/EVA (return on invested capital and economic value-added) lens and analyze Ryanair’s financial metrics.

Chart 2 – Airline Passenger Penetration, Market Size & GDP Per Capita - 2017/18

Chart 3 - Airline Passenger Traffic & GDP Per Capita - 2017/18

Chart 4 – Airline Passenger Traffic Growth, Market Size & GDP Per Capita

For us, the big takeaway from these charts is the strength and secularity of these long term global demand trends and the formidable anchors behind them (air travel demand is tied to growing disposable income and wealth and rising urbanization rates) which in turn reinforces demand for travel through globalization-led growth, supported by higher trade and investment growth that is enabled by greater connectivity.  It is not only a virtuous self-reinforcing cycle, it is also one that generates huge positive externalities that support higher national economic growth and wealth.  On its face, it is tough to find a more global, impactful and compelling top-down secular investment theme than the aviation sector.

One of the biggest positive externalities that the sector produces is much deeper global trade integration, particularly for smaller countries with smaller endogenous populations and thus smaller captive domestic markets. The theory here is that greater global trade integration in turn fosters impactful economic development, making countries more competitive and productive, which in turn supports sustainable income growth and an eventual reduction in poverty through wealth convergence.  Global trade has also been a tremendously strong growth engine in the last seven decades, almost tripling as a share of global output since the early 1950s.  As more emerging countries join this fast-moving train of globalization, there should be no doubt that the overall pie of value and wealth creation from trade will only grow larger over time. 

The long-term growth dynamics paint a very clear picture of this relationship (see Charts 5 & 6 below).  Chart 5 shows 43-year (1974-2017) average growth in air freight tonnage (mil ton-km) versus the growth in airline passengers carried.  The size of each bubble indicates the relative size of the country’s airline capacity (proxied by passengers carried).  With greater passenger capacity comes the capability to engage in global trade via the air freight channel.  Even though air cargo only accounts for about 1% of the total volume of global trade, it accounts for an estimated 35% of the value (according to IATA), enabling small countries to engage in the highest value-added part of the global value chain.  This underlies the strategic weight placed on national airlines by emerging countries, especially those that are keen to leapfrog over the commoditized portion of the global value chain and integrate more swiftly into advanced industrial high value-added products.  Chart 6 illustrates the resultant robust linkage of higher air freight volume penetration (x-axis) to growth in the value of merchandise exports (y-axis). Several of the leading export power houses of the last few decades lie in this top-right cluster of countries (China, Vietnam, Singapore, S. Korea, Indonesia, Malaysia).  They are joined by several new entrants to the high growth air freight market (UAE, Qatar, Turkey and Ethiopia) as these newcomers engage in this highly strategic corridor of trade with real heft (e.g. the UAE & Qatar collectively shipped more airfreight tonnage than China in 2017).  

Chart 5 - Air Freight Growth & Passenger Growth: 1974-2017

Chart 6 - Air Freight Growth & Merchandise Export Growth: 1974-2017

As mentioned previously, for small countries in particular, the trade integration driver has proven to be a really powerful lever for sustainable income and wealth creation over time.  Chart 7 below captures this dynamic in a static fashion, with trade penetration (total trade or exports + imports as a fraction of GDP in 2018) plotted versus GDP per capita (adjusted for purchasing power parity or PPP, 2017); note that the size of the bubble indicates the relative size of the country’s population.  The large population countries, since they can drive ample endogenous growth from their large captive markets, tend to live in the bottom quadrant (below 50% trade to GDP) of the trade openness spectrum.  Those countries with smaller populations, and particularly those which are not endowed with substantial competitive resource endowments, are incentivized to drive per capita income growth through higher trade integration.  Above 50% trade penetration and especially once a country enters the ‘middle income trap’ range of income ($10k-$20k), the steepening curve underpinning the relationship between trade and wealth is pretty clear. Larger countries that have achieved this feat (e.g. S. Korea, Vietnam, Mexico, Thailand, Turkey, Philippines) are not surprisingly some of the highest productivity-led growth countries of the past two decades.  More on productivity-led growth in a bit.

Chart 7 – Trade Openness, Population Size & Per Capita Income ($PPP) – 2017/18

Beyond trade in goods, the other obvious positive externality that is derived from higher airline connectivity is tourism.  Chart 8 below plots tourism receipts per capita relative to airline passengers carried per capita in 2017, while Chart 9 shows the relationship between per capita tourism receipts and per capita income (in $PPP terms).  Naturally, as airline capacity grows, this in turn enables tourism receipt growth to contribute to income growth, both directly (usually with a high tourism spending multiplier as tourism is labor-intensive) and indirectly via facilitation of business travel and the associated capital and know-how flows that accompany non-tourist visitors.  

Once again, the countries that experience the biggest ‘bang for the buck’ from growing their inbound tourism ecosystems tend to be (on balance) the countries with the highest connectivity rates and thus greatest openness to both trade and human traffic flows.  The outliers in Chart 9 are few and far between and tend to encompass either very large countries with relatively underdeveloped inbound tourism sectors in the context of an overwhelmingly large economic activity (e.g. Russia, Brazil, China and India) or very small countries that are over-reliant on tourism receipts in their limited economic activity (e.g. Maldives, Iceland, Luxembourg).  

Chart 8 – Tourism Receipts & Airline Passengers - Per Capita, 2017

Chart 9 – Tourism Receipts & GDP Per Capita ($PPP) - 2017

Underpinning a country’s openness to trade and human flows is the appropriate and scalable fixed infrastructure that is necessary for airline and shipping connectivity.  This infrastructure is both hard assets (airports, seaports, warehouse parks, railroads, highways, etc.) but also soft assets in the form of efficient customs regulations and authorities, logistics clusters, etc.  The World Bank has developed a survey-based scoring for country logistics named the Logistics Performance Index (‘LPI’) which ranks both the quality of hard and soft assets for each country (1=low to 5=high).  Chart 10 below plots the LPI relative to airport departures per capita (size of the bubble is the annual airport departure capacity), while Chart 11 shows the relationship between LPI and container port throughput per capita. 

Chart 10 – Airport departures Per Capita & Logistics Performance Index – 2016/17

Chart 11 – Container Port Throughput Per Capita & Logistics Performance Index – 2016/17

Both charts illustrate what one might expect – beyond a certain minimum quality threshold (around 2.5 – or roughly where countries like Russia, Colombia and Morocco score) the higher quality the logistics ecosystem is, the greater the connectivity performance tends to be – both for airport departures and container trade.  What is most interesting is how small the club of countries that score in the top quadrant of LPI scores (above 4.0) while also generating above-trend departures and throughput penetration.  The only three countries that make the cut on both asset categories (seaports and airports) are: the UAE, Hong Kong and Singapore.  It is no coincidence therefore that these three countries are also the biggest positive outliers on trade intensity (total trade to GDP – see Chart 12).

The other big takeaway observation from Charts 10, 11 & 12 are the large population countries that remain stubborn laggards on growing domestic connectivity.  Both in terms of airports and seaports, the regions of South Asia (India, Pakistan, Bangladesh) and Sub-Saharan Africa (Nigeria and Kenya) continue to be big laggards, particularly on the infrastructure investment side.  These countries have chronically underinvested in seaport and airport capacity, and have therefore stymied port throughput volumes and airport departures.  As a result, they have likely driven pent-up volumes away to alternative regional transshipment or transit hubs.  This has as a result held these countries back in terms of driving trade integration and openness and the multiplier effect of tourism (Chart 12 – bottom left quadrant).

Chart 12 – Total Trade Intensity & Per Capita Port Throughput – 2017/18

Since we decided to introduce seaports to the discussion at hand, we thought we would take this opportunity to make a quick detour to diagnose the dynamics that are unfolding in the port transshipment space. Even though ports and airlines are not strictly comparable (vastly different business models with distinctly different competitive dynamics, pricing dynamics and asset/labor intensity) there are some interesting parallels between the two when it comes to trans-connectivity.  In many ways, the hub-and-spoke model of connecting air traffic is quite similar to the transshipment model of transporting containers and bulk goods from their origins to ultimate destinations.  The growth of the Middle Eastern super-connector model in the aviation space, in some ways, has mirrored the growth in the transshipment of containers, along East-West routes that are connected effectively by geographically well-located Middle Eastern ports of transfer.     

According to a recently published report from Drewry Maritime Research, the global transshipment incidence ratio (or the proportion of global container port throughput that is a transshipment activity; the balance being origin/destination or ‘gateway’ import-export traffic) started to decelerate markedly in 2012 (see Chart 13 below).  The incidence ratio rose consistently since 1990, but peaked at 27% in 2008 and again in 2012 – two big inflection points for global trade growth momentum. Driving this recent deceleration are three main drivers: 1) consolidation in the shipping industry which created large alliances that in turn prefer more direct port pair connections, 2) slowing overall global container port growth (from 14% in 2010 to just above 6% in 2017) and most importantly 3) some signs of saturation in gateway port capacity growth, specifically in Chinese ports and ports feeding off of Chinese container trade flows. 

Chart 13 – Global Transshipment Incidence – 1990-2017 (Drewry)

During the period 2000-2017, Chinese throughput growth compounded at a rate exceeding 10% per annum, while accounting for just under one-third of total global capacity additions.  To put this in context, China added 172.7 million TEUs (twenty-foot equivalent unit) in container capacity during this 17-year period, equivalent to 77% of global port throughput capacity in 2000. In the more recent past 5 years (2012- 2017), this growth rate decelerated to 5.1%, while China’s share of global port capacity additions rose to a whopping 40% of the world total (from 31% during 2000-2012).  In the meantime, growth rates for the main transshipment ports that have fed off of China’s excess demand for the better part of the past 40 years (mainly Hong Kong and Singapore) have begun to decelerate meaningfully, and in the case of Hong Kong, are actually shrinking (see Chart 14 below, left panel).  Singapore which grew its TEUs at 4.1% during 2000-17 (7.3% from 2000-2008), witnessed a meager 1.2% growth rate from 2012-2017.  Hong Kong’s TEUs declined by -1.8% during 2008-2017 and by -2.1% from 2012-2017.  This is despite still very healthy growth rates amongst most of the remaining gateway ports in the E. Asian region, which grew at an average annual rate of 3.7% (2012-2017), a small premium to global TEU growth globally of 3.5%.

Chart 14 – Port Throughput Growth: East Asia & South Asia/Persian Gulf (2000-2017)

Shifting the focus to Southern Asia and the Persian Gulf (Chart 14, right panel), the overall deceleration picture looks quite similar, though somewhat lower in magnitude.  Indian TEU growth in the past five years has slowed to 5.7%, from the 17-year trend growth rate of 10.4% (2000-2017), while most of the other S. Asia ports have exhibited accelerating volume growth (much smaller base vs. India; Sri Lanka is less than half of Indian capacity, Pakistan & Bangladesh are 20%-22%), with S. Asia as a region growing port traffic at 6.6% between 2012-2017.  

Ports in the GCC, several of which rely heavily on transshipment traffic to/from major S. Asian gateway ports, have also decelerated significantly and exhibit the same leverage (to big neighbors) dynamics as their E. Asian counterparts.  Traffic growth across the GCC ports declined to 3.1% in 2012-2017 vs. a 17-year average of 9.7%.  Ports like Oman that rely heavily on transshipment traffic have suffered the biggest decelerations in activity with growth rates falling from 8.7% (17-year) to 2% (12-17). Qatar and Kuwait have bucked the GCC trend as they have clawed back traffic from regional transshipment ports by aggressively building out domestic capacity; the former in response to the embargo imposed on it in 2017.   Finally, Saudi Arabia and Abu Dhabi have both embarked on very large and ambitious domestic port expansion plans that have only begun to reach completion stages in the past 2 years, further dampening the demand for overflow transshipment traffic from the neighborhood.

The big regional incumbent in the S. Asia/ME sphere (the two regions straddling the Indian Ocean) is Dubai’s Jebel Ali port, the world’s 9th ranked port by traffic and a real hegemon in the extended region (see Chart 15 below for a map of surrounding ports).  The strength of Jebel Ali’s strategic location and first mover advantage cannot be overstated (see Chart 16 for a graphic side-by-side comparison of logistics strength vs. neighboring MENA peers).  Most of Dubai’s regional neighbors are large gateway destinations with either insufficient port infrastructure in place or extremely burdensome customs and clearing environments; for the most part – it is both. Even though large regional gateway markets such as Saudi Arabia, Abu Dhabi and Morocco have spent heavily on port expansions in the past 5 years, markets such as the S. Asian markets, Iran, Iraq and Egypt will likely take decades to fully satisfy their hinterland needs.  So, the immediate transshipment growth opportunity for a strategically positioned port like Jebel Ali remains quite attractive in the short run.  Counterbalancing this opportunity to feed off regional pent up demand for logistics is the fact that the Arab World and South Asia are the two least integrated regions in the world due to protectionist trade policies towards neighbors that severely impedes intraregional trade and investment.  

Finally, Jebel Ali’s parent (Dubai Ports World) is coy about disclosing its own real transshipment incidence ratio (there is a grey area of what is really hinterland traffic and what is not, especially in light of the large share of re-exports to/from duty free zones in Dubai).  Most analysts we have spoken to estimate that Jebel Ali’s incidence ratio may be as high as 45%-55%.  This in turn may evolve into a vulnerability either as the major gateway markets grow capacity to eventually saturate their endogenous needs or as GCC peers like Abu Dhabi and Saudi Arabia quickly catch up to Jebel Ali in capacity build-out and begin cannibalizing their transshipment traffic.  Both of these trends are at play presently.

A recent report from BCG, titled “Will Middle Eastern Ports Continue to Succeed”, highlights several of these vulnerabilities.  The report warns that “From 2011 through 2016, the annual growth rate for container throughput was 4%, exceeding the global average.  But continued success is threatened by rising overcapacity, exposure to transshipment business and lagging port productivity.” The authors highlight the fact that “although the Middle East accounts for less than 3% of global GDP, its ports handle approximately 20% of global seaborne trade.”  On the important topic of overcapacity, the authors argue that “Unless growth accelerates significantly, it does not appear that additional capacity will be required for the foreseeable future. Ultimately, the overcapacity will lead to a lower return on capital than is normally expected from port infrastructure investments.”  On the equally important topic of transshipment exposure and vulnerability, the authors state “Transshipment accounts for more than half (53%) of the throughput of Middle Eastern Ports.  At ports in the UAE and Oman, transshipment represents the lion’s share of utilization.  A large volume of transshipped cargo is destined for ports in other countries within the region and outside the region….This year [2018] has seen several high-profile examples of destination ports taking such steps to counter the dominance of transshipment hubs…..If smaller destination ports succeed in attracting large volumes of direct calls, the current model of serving the entire region with a few transshipment hubs will be threatened.”  

Time will only tell how the region evolves from this challenge.  We believe many of these strains will likely manifest themselves initially through significantly lower utilization levels and weaker container pricing power stability - something we are watching carefully. Keep these three important dynamics in mind (utilization, pricing power and return on capital) as they will be important factors to recall when we embark on analyzing the airline sector from a micro-standpoint later in the post.

We chose to highlight the container seaport case here as a possible leading analog to the aviation space since we are starting to see similar dynamics begin to play out in the latter.  Not only is competing regional capacity growing at a massive rate (among the four regional super-connector ME airlines and their associated national airports), but the legacy airlines and airports on both the origin and destination ends have also begun to grow capacity in an effort to exploit their own endogenous demand for travel.  This in turn may challenge the hub-and-spoke & transshipment business models by disintermediating many of these transition hubs, particularly those with small domestic hinterland markets. Overcapacity in the region may ensue as a result, dampening pricing power for all and debasing returns on investment. More on this when we look more closely at the GCC aviation's competitive dynamics later in the post.

Chart 15 – Gulf Ports in MENA and South Asia

Chart 16 – MENA Logistics Infographic

No one can deny the impressive scale and enormous resources that have been deployed in infrastructure in the region in the past 4 decades, much of it focused on the logistics and transportation sectors. According to the World Economic Forum’s Arab World Competitiveness Report 2018,  “Over the past decade, GCC countries have led the Arab world in investing heavily in infrastructure and technology in order to diversify their economies and create the conditions for more innovation-driven and high-value-added businesses.  In 2017, in spite of the decrease in oil revenues, the total value of projects either in the planning stage or in the delivery stage across the GCC amounted to $2.7 trillion.” To put that number in context – it is 1.7x the cumulative 2018 GDP for the GCC bloc ($1.63 trillion).  It is also a multiple of global infrastructure spending in 2015 of $2.3 trillion (according to Oxford Economics), a figure that includes all seven categories of infrastructure: electricity, road, telecom, rail, water, seaports and airports.  Oxford’s estimate is, as a result, heavily skewed to electricity which accounts for one-third of the global total.

So, with all this massive investment in GCC infrastructure (central to this have been the seaport and aviation investments) – it begs the question - has this in fact yielded strong dividends in the form of diversification of exports?  According to World Bank data, so far it does not appear to be.  

As a region, the Arab World boasted the highest ratios of total exports to GDP in 2017 (44%), with countries like the UAE (above 100%) and Bahrain (75%) leading the pack (Chart 17 below).  However, the region also has the highest ratio of fuel exports as a share of total exports (58.9% in 2017) while also suffering from the lowest ratio of manufactured exports to total exports (16.4%) – far lower than even Sub-Saharan Africa (26.8%).  As a region, this uniquely skewed export structure has remained quite stagnant for the better part of the last 15 years, with some improvement in manufacturing intensity of exports only achieved during the last decade (10% in 2007 to 16.4% in 2017).  On a 15-year trend basis, the big positive outlier has been the UAE, almost tripling its manufacturing intensity in exports from 3.6% in 2001 to almost 10% in 2017, while reducing its fuel intensity in exports from 92% in 2001 to 32.5% in 2017 – an impressive achievement for what remains a predominantly hydrocarbon reliant economy. 

Chart 17 – Export Structure, GCC, MENA vs. Global Regions – 2001, 2007, 2017

Despite this somewhat mixed progress on export diversification, real economic diversification improvements continue to be lacking, and in some cases, have deteriorated over the last decade.  Chart 18 below charts the trend in economic complexity, an index developed by the folks at MIT Media Laboratory’s Observatory of Economic Complexity and Harvard University’s Kennedy School.  The Economic Complexity Index (ECI) is a measure of the relative knowledge intensity of an economy by considering the knowledge intensity of the products it exports. ECI is calculated as an export-share weighted average of sophistication levels of the country’s export basket. The sophistication of each good is measured as the weighted average of real GDP per capita (a proxy for the level of sophistication) of all countries that export that good.  The ECI index attempts to capture not only the quantity of exported products, but their quality as well, scaling export diversification by the number of countries that export those products. Low values on the ECI represent a small number of exports and/or exports of common products (e.g., agriculture, natural resources). High values, by contrast, represent a large number of exports, including products exported by very few countries (e.g., high technology, complex goods).

For most Arab and GCC countries, the long run trend in ECI in the past few decades can only be described as one of stagnation. Even though many of these economies boasted levels of economic complexity in line with several of their E. Asia peers in the 1970s (e.g. Jordan had a higher ECI score than Singapore in 1977, the UAE had higher export complexity than Malaysia in that year), the trend since then and particularly up to the mid-1990s was one of sharp deterioration. Since the mid 1990’s however, a few countries have made meaningful strides in improving their ECI scores (UAE & Morocco especially).  Much of this improvement is likely due to the heavy investments in enabling infrastructure, but we would argue that it is also due largely to aggressive structural economic reforms being implemented during this time-frame, where the UAE and Morocco stand out.  The laggards on reforms and investments (e.g. Jordan, Lebanon, Egypt) have witnessed material deterioration in their ECI scores (in cases like Jordan they have turned negative) while countries like Saudi Arabia and Tunisia have essentially stagnated for almost four decades.  Qatar is a prominent negative outlier here, most likely due to the fact that the explosion in LNG exports in the mid-1990s completely overwhelmed any meaningful growth in non-hydrocarbon export diversification.

Chart 18 – Economic Complexity Index (ECI) Trend – 1977-2016

Looking at a static current picture of where country ECI scores stand relative to their incomes per capita (Chart 19 below) the divergent story is clear.  The GCC countries are overwhelmingly in the top left quadrant (low ECI rankings, very high per capita PPP incomes), standing in clear contrast to a global trend of improving economic complexity that is driving a steady improvement in wealth. The GCC is joined by a few small population resource-reliant countries that have also likely suffered from the stubborn effects of Dutch disease (Australia, New Zealand and Kazakhstan) where growth since the 2007 peak in commodity prices has begun to decelerate. The positive quadrant (top right) outliers are worth highlighting (Singapore, Ireland, Hong Kong, Switzerland, Norway and Canada).  This group of small population countries have leveraged their move into higher value-added sectors to stimulate above average growth.  The US and Germany are the two large population countries in this quadrant that have pulled away meaningfully from the established curve. 

Chart 19 - Economic Complexity Index (ECI) vs. GDP Per Capita & Population Size – 2015

Finally on the subject of the efficacy of diversification strategies of the past decade, it is also useful to take stock of where the GCC countries currently rank in terms of export product/service concentration (Herfindahl-Hirschman Index or ‘HHI’) and the relative weight of high technology exports as a percent of total merchandise exports (see Chart 20 below).  GCC countries (red bars) continue to rank very high in terms of concentration with the likes of Kuwait, Saudi Arabia and Qatar scoring substantially worse than resource-reliant peers like Australia, New Zealand and even Canada.  In terms of high technology export intensity, all GCC members barely even register on the scale with weights ranging from a low of 1% for Bahrain to 4% for Kuwait. This is in stark contrast to MENA peers like Israel (14%) or the OECD (14%) and World (16%) averages.  

Once again, it may be premature to judge the efficacy of diversification strategies only 2-3 decades from when they begin. Nevertheless, the meager dividends from diversification so far go a long way to help explaining why productivity-led growth has been very disappointing, especially for the rich GCC countries.  

Chart 20 – Export Centration & High Technology Intensity of Exports - 2017

On the specific topic of productivity-led growth, we thought it might be useful to try and analyze the direct linkages between investments in logistics infrastructure (especially those aimed at diversifying exports) and productivity growth.  Do countries that invest heavily in enabling infrastructure like airports and seaports in turn generate incremental productivity gains over time? 

There is quite a bit of literature on this topic amongst academics and economists.  Before we get into this specific topic, it is worthwhile stepping back quickly and reinforcing one of the ‘holy grails’ of modern economic theory: namely that productivity is one of the most important drivers for sustainable economic growth.  

It is common knowledge that there are three main factors that drive economic growth: 1) accumulation of capital stock, 2) increases in labor inputs (workers or hours worked) and 3) productivity (or utilizing factor inputs more efficiently to generate greater output). Most of the levers available to countries and businesses that aim to drive productivity higher lie in the realm of technological advancement, or more simplistically – just doing things better and more efficiently.  

Once armed with this somewhat simplistic framework, it is intuitive to then decompose growth into these three main factor categories.  Since capital and labor inputs are by definition finite (in some cases they are shrinking in abundance: captive population growth eventually decelerates, capital is mobile and may migrate to better opportunities elsewhere), the only factor that is at least theoretically infinite is technological advancement. Most growth economists will claim that it has been robustly demonstrated, both theoretically and empirically, that productivity through technological advancement IS the main driver of long-run growth.  The simple logic is that holding other input factors constant, the additional output obtained when adding one incremental unit of input (capital or labor) will eventually decline (the law of diminishing returns).  As a result, a country or business cannot maintain its long-run growth by simply accumulating more resource inputs – you need to fundamentally alter the way you ‘process’ these inputs in order to generate sustainable incremental growth.  That’s the top-down economic way of explaining productivity.   

Since we are not economists but rather investors in businesses, we found the following quote from William Lewis, founding director of the McKinsey Global Institute, in the prologue to his book “Wealth, Poverty, and the Threat to Global Stability”, an equally compelling explanation of the potency of productivity in driving sustainable growth (emphasis is ours).  Mr. Lewis (and MGI) refreshingly look at productivity from a more practical perspective -through the sectoral and business-level lens.   

“Productivity, on the other hand, is the most important objective that businesses and their management try to improve all around the world. The reason, of course, is that productivity is very closely connected to profitability. Productivity is simply the ratio of the value of goods and services provided consumers to the amount of time worked and capital used to produce the goods and services. If a firm produces more goods and services for the same effort or it produces the same goods and services for less effort, its profitability increases. Such a firm is likely to invest the funds from its increased profits in building a bigger business. The firm then makes even more profits. This process continues until other firms note the success of the more productive firm and copy its more productive ways. The profitability of the innovative firm returns to normal. However, the productivity of all firms is increased. Since the productivity of a society is simply the average (weighted) of the productivity of all the firms operating within it, understanding the productivity of firms around the world must be important. Why are firms in some societies more productive than firms in other societies? The answer has got to help us understand where the world is going.  Societies with higher productivity have overcome to a greater degree whatever cultural, religious, ethnic, climatic, and political barriers have constrained productivity. These higher productivity societies have been successful in competition with lower productivity societies. They have been successful either through conquest or through simply surviving the hardships of nature.”

Chart 21 below is a current snapshot of labor productivity per worker vs. income per capita in 2017 and is illustrative of the recent empirical evidence supporting this relationship.  Since most countries within a specific socio-economic profile tend to have similar labor participation rates, they tend to cluster around one another in terms of labor productivity (GDP/labor count) relative to per capita incomes (GDP/total population).  The outliers are those that are either generating more value per labor input (above the curve) or less value per labor input (below the curve) – both relative to average incomes.  

The GCC countries cluster in the top right quadrant (high income, high labor productivity level).  In order to normalize for resource intensity in GDP (all natural resources, inclusive of hydrocarbons, coal, minerals and forestry) we adjusted the GDPs by deducting the resource rent component from them (see top left insert panel in Chart 21 which ranks the top resource intensity countries; for example, Kuwait generates 37% of its GDP directly from resource extraction and sale).  Here we are relying on the World Bank resource rent estimates which we believe may be grossly understated; neither the World Bank or IMF publish historical data series on non-resource GDP to our knowledge. We thus make the simplifying assumption that the portion of population and labor force that is directly employed by the resource sectors is minimal.  By haircutting GDP (the numerator for both axes) for resource intensity, we attempt to neutralize some of the implicit productivity derived from resource exports.  Saud Arabia remains a big positive outlier nonetheless, most likely due to its very low labor force participation rate (56% vs. 73%-87% for its GCC peers) which artificially flatters its labor productivity scoring.  Please refer to our post on demographics and the GCC Migrant Model for a full analysis on this demographic idiosyncrasy.  

Chart 21 – Labor Productivity & Per Capita Income (Non-Resource GDP) – 2017

Chart 22 plots labor productivity growth vs. GDP per capita growth for the 15-year period of 2002-17), once again using adjusted non-resource GDP as a numerator for all countries.  Here again, the relationship is intuitive and unsurprising – positive growth in productivity tends to have a positive relationship to higher income levels for most countries.  Countries below the regression line, and particularly those printing negative productivity growth rates encompass all of the GCC countries along with Yemen, Iraq, Zimbabwe, Greece, Italy and Jamaica.  

This loudly suggests that GDP growth over the last 15 years for these countries has been achieved despite very poor labor productivity performance and exclusively due to overwhelming factor input accumulation.  For the GCC, both factor categories (labor and capital) are likely culprits here with a large influx of foreign migrant labor on the one hand and huge hydrocarbon resource rents being deployed on the other (please refer to our previous posts on both topics).

Chart 22 - Labor Productivity Growth & GDP Per Capita Growth (Non-Resource GDP) – 2002-2017

When we look at how the GCC stacks up relative to other countries globally in terms of labor productivity during the past 15 years, the picture is quite bleak.  According to The Conference Board data, six of the bottom ten countries in terms of labor productivity growth between 2004-2019 are GCC countries (see Chart 23 below).  The only two countries that printed lower productivity growth were Yemen and Venezuela, while the super-majority of countries (developed and emerging) generated positive growth during this period.  

One common counterargument is that the starting point income levels for the GCC countries are high, resulting in a skewed base effect. The logic here is that these countries reached very high income levels very early in their overall factor development evolution, something that tends to go hand in hand with Dutch disease, one of the many variants of the natural resource curse.  There is quite a bit of economic literature that has focused on this issue (the effects of and linkages between high resource rents and factor productivity growth).  Most of the empirical evidence suggests that high natural resource rents, especially those that accrue early in a country’s stage of development, tend to have significant negative effects on productivity growth, whether through direct channels or through strong negative externalities.  The logic here is that the earlier the stage of development that this ‘curse’ takes hold on a country, the more rigid the disincentives are to reform or diversify sustainably. For a fantastic piece on this broader subject, please take a look at a 2011 piece published by our friends at Renaissance Capital titled “The Revolutionary Nature of Growth”.

Some of the more recent economic research (e.g. ‘Economic Freedom and Productivity Growth in Resource Rich Countries’ - Farhadi, 2015) attempts to study some of the important nuances in the empirical data.  Why have some rich, resource intensive countries succeeded to overcome this resource curse while others have failed?  One of our favorite books on this important topic is Daron Acemoglu & James Robinson’s book titled, ‘Why Nations Fail’ – 2012).  Both the Farhadi and Acemoglu studies suggest that institutional strength and the resultant economic freedom that is derived from them are key contributors to overcoming this curse.  The more market friendly (i.e. private sector supportive) resource rich countries tend to experience significantly higher productivity-led growth than less market-oriented ones.  The Farhadi study, which uses a very comprehensive empirical model based on data for close to 200 countries spanning 40 years (1970-2010), concludes that the impact of natural resource rents on economic growth improves meaningfully as economic freedom increases, while “the quality of institutions ultimately determines whether natural resource abundance is a blessing or a curse….Countries with greater economic freedom may set up transparent, accountable, and forward-looking institutions that facilitate more efficient utilization of their resources and control rent-seeking motives by reducing the rate of return on unproductive economic activities or grabbing.”

In order to analyze this phenomenon for the GCC, we looked at long term growth trends in labor productivity going back to the 1950s.  Charts 24 and 25 below look at GCC productivity growth trends relative to select country comparables through two distinct time-envelopes: 2000-2019 (the ‘short-run’ – 19 years) and 1950-2019 (the ‘long-run’ – 69 years) – both data sets were sourced from The Conference Board.  The short run trends are anemic at best, with productivity performance negative across all six GCC states (Saudi Arabia -1.2%, UAE -3.3%, Qatar -2.3%, Kuwait -1%, Oman -3.5% and Bahrain-1.8%).  The trend in the UAE and Bahrain began to improve, at the margin, beginning in 2009 post the global financial crisis when capital became scarce and expensive (UAE +1.2%, Bahrain +0.5% from 2009-2019).  The other four GCC states witnessed either persistent deterioration at worst (e.g. Oman -4.6%) or stagnation at best (Saudi Arabia -1.5%, Kuwait -2.7%, Qatar -0.8%) during this time frame.  In contrast, the trend amongst comparable countries (small and large population countries, developed and emerging, resource-intensive and resource-poor) showed a different trend completely.  During 2000-2019, all of the comparator countries showed consistent improvements in productivity growth trends, including those states disproportionately exposed to globalization that experienced a meaningful ‘speed bump’ in 2008/9 (e.g. Singapore and Ireland).

The long-run trends (Chart 25, 1950-2019) are even more sobering, exhibiting the same relative dynamics as those in the short-run trend charts (Chart 24), but with greater amplitude for the outliers.  GCC countries such as the UAE (-2.2%), Qatar (-2.5%) and Kuwait (-2.2%) stand in sharp contrast relative to leaders in the non-GCC peer group (Ireland +3.3%, Singapore +3.1%, Hong Kong +3.6%, Norway +2.3% and China +4.3%).  Even the likes of Russia, which emerged from the break-up of the USSR as the poster child of resource dependency (resource rents peaked at above 40% for the USSR in the 1980s, they are sub-15% now), while facing several existential crisis episodes during the interim period, managed to squeak by with +1.6% productivity growth during most of this timespan (1960-2019).  

Finally, if productivity really is in fact the ‘end-all, be-all’ driver to growth and prosperity, the natural question then is – why aren’t resources being directed effectively to productivity-enhancing or enabling investments?  Shouldn’t investments that enable greater connectivity and export diversification lead to higher productivity-led growth?  If this is in fact an empirically observed relationship, why hasn’t the GCC reaped these rewards?   Why have the GCC countries failed to generate positive productivity growth during the last 69 years, despite such herculean efforts to diversify and invest?  

Could at least part of the answer lie in the realm of their chosen development models, the ensuing quality of their institutions and the resulting low levels of true economic freedom?  Has their development model in turn stymied market forces in allocating resources to the private sector?  Why have these bottlenecks persisted for such an extended period of time?  These remain burning questions for us, with no obvious or simple answers. 

Chart 23 – Labor Productivity Growth: Country Ranking – 2004-2019

Chart 24 – GCC Countries Labor Productivity Trends vs. Comparables – 2000-2019 

Chart 25 – GCC Countries Labor Productivity Trends vs. Comparables – 1950-2019 

These questions, though unanswered, do however bring us back to the topic at hand – logistics - and specifically the large investments made in the aviation sector (airports and airlines) in the past 15 years. 

The next two charts attempt to investigate the relationship between labor productivity growth and growth in aviation-enabling infrastructure capacity (airport departures and airline passengers) – see Charts 26 & 27 below.  Both charts paint the same picture for most countries along the capacity investment curve – the greater the capacity growth (horizontal axis) and scale achieved (size of the bubble = size of departures or passengers) tends to be positively linked to higher productivity uplift (vertical axis).  The more mature/saturated markets (US, developed Europe, Oceania) tend to cluster in the flatter part of the curves (bottom left) with much smaller uplift in productivity from an incremental dollar invested in capacity due in large part to the relatively high penetration of their domestic aviation sectors.  Countries on the steeper end of the curve (top right, above the steep part of the curve) are high productivity countries that are, at least in part, generating positive growth from top decile growth in aviation connectivity.  

Chart 26 – Labor Productivity Growth & Airport Departures Growth – 2002-2017

Chart 27 – Labor Productivity Growth & Airline Passenger Growth – 2002-2017

It is no coincidence here that this same small group of countries also rank in the top five aviation volume growers (China, India, Turkey, Indonesia, Ireland, Philippines, Thailand).  The big negative outliers here again are the GCC states (bottom left) with some of the highest aviation capacity growth rates.  The UAE & Qatar have printed 12%-15% compounded average growth rates (CAGR) in airport departures while the UAE is by far the biggest grower of airport capacity of material size since 2002 globally. These two countries have also ranked in the top 5 in terms of airline passenger growth during the last 15 years, with 15%-16.5% average annual growth.  

Despite this impressive growth in connectivity through aviation capacity, these countries in particular have very little to show for it in terms of incremental economic productivity.  The acute and growing variance in airport departure capacity and actual inbound tourism arrivals speaks volumes to how much of this growth has in fact not led to any meaningful multiplier effects domestically.   A case in point is the UAE, where aggregate airport passenger figures (arrivals & departures) topped over 100 million in 2014 and today are estimated to be north of 130 million (with globally 3rd-ranked Dubai International Airport alone handling over 90 mil passengers in 2018 or 71% of the national total). Compare this to a total inbound tourism arrival figure of just over 20 million for the UAE and 15.8 million for Dubai, and the implicit transit ratio is as high as 31% (assumes arrivals=departures in any given year).  

Chart 28 below plots the relationship between tourist arrivals growth and passenger growth in the past 15 years.  By driving passenger growth so aggressively, and despite starting with quite a small endogenous travel market (tourist arrivals in 2002 were 5.5 million for the UAE and just 590,000 for Qatar) these two Gulf city-states, along with Turkey, have positioned themselves as dominant transit aviation hubs (top right cluster).  In doing so, city states such as Dubai have managed to translate at least a portion of incremental passenger traffic into inbound tourist arrivals.  As a result, Dubai has earned a coveted spot on the map of the top 20 global tourist destinations, garnering tourism receipts of over $21 billion in 2017 and eclipsing regional rivals such as Egypt (2017 arrivals of 8.2 million and receipts of $8.6 billion) in a very short period of time. As the saying goes, “Build it and they (some of them) will come (and stay!)”.

Chart 28 – Tourism Arrivals Growth & Airline Passenger Growth – 2002-2017

Finally, in order to close the loop on the topic of productivity, it is important to look at an important driver of labor productivity – the quality of human capital.  The World Bank tabulates and publishes an index for countries named the Human Capital Index (HCI) which calculates the contribution of health and education to worker productivity by benchmarking prevalent health quality and education quality scores relative to “full health and complete education”.  We assume that a high/improving HCI score is not only assumed to be a by-product of greater productivity, but also a contributor to it, creating a virtuous cycle (a healthier and more educated citizen is likely a more productive one).  

This relationship is plotted in Chart 29 below and produces a familiar and intuitive picture.  Countries above the curve on the left (low HCIs) have neglected to improve their human capital statures despite generating relatively decent labor productivity in the past 15 years.  Several of the countries in this quadrant are resource reliant (Iraq, South Africa, Algeria, Kuwait, Saudi, Qatar), while quite a few suffer from high unemployment rates and low youth and female labor participation rates (most of MENA, the rest of the GCC) thereby overstating their real labor productivity.  On the opposite side of the spectrum (high HCIs, high labor productivity) we find countries like Singapore, Hong Kong, Ireland, Norway and the USA.  Most of these positive outliers are fully exploiting this virtuous cycle.  We would go even further and posit that many are likely generating a disproportionate amount of incremental productivity as a result of having enhanced the quality of their human capital. 

What caught our attention when we first looked at the World Bank’s HCI scores for the GCC countries is the big gap between relatively low HCI scores (measuring exclusively the quality of healthcare and education in human capital stock) and the high global competitiveness scores (GCI) attributed to them by the World Economic Forum.  The WEF GCI scores tend to flatter most of the GCC countries relative to global peers (see Chart 30 below), mainly because they encompasses a far broader scope of attributes (12 ‘pillars’ that include not only health and education, but also everything from infrastructure quality to macroeconomic stability to such soft attributes as institutional quality, ‘business dynamism’ and ‘innovation capability’).  For example, the UAE is ranked 27th globally in terms of GCI, Qatar 30th, Saudi Arabia 39th, Oman 47th, Bahrain 50th and Kuwait 54th), placing the two front-runners (UAE and Qatar) in the top 20% of countries.  On an HCI ranking basis, the UAE ranks 49th while Qatar ranks 60th, placing both in the fourth decile, with countries like Kuwait (77th ranked) and Saudi Arabia (73rd ranked) scoring below the 50% rank level. According to the WEF 2018 report on Arab Competitiveness, “Of the 12 pillars of the GCI, infrastructure and technological readiness are the areas where the Arab world has made the most significant progress over the past decade relative to the OECD countries, a result of heavy investment in transport and information and communication technologies (ICT) connectivity. However, these improvements have not led to gains (relative to OECD countries) in innovation.” The report points out that the gap has actually widened the most in terms of two specific pillars- macroeconomic environment and labor market efficiency.  The latter confirms the acute underperformance in overall productivity.  

Chart 29 – Labor Productivity and The Human Capital Index – 2017/18 

Chart 30 – Global Competitiveness Ranking & Human Capital Index Ranking – 2017/18

This top-down assessment leaves us with several key observations and a few important unanswered questions.  The GCC has made big strategic investments in connectivity in the past 15 years, but have (thus far) very little top-down dividends to show for it in terms of productivity growth or human capital uplift, deviating materially from a relationship experienced by other successful developed and emerging market countries during this time span. Several have been rewarded with successfully establishing new and scalable new industries as a result (tourism, world class seaports and airports, world leading airlines). Yet, despite this, they have not been able to translate these accomplishments into substantive improvements in growth.  This remains a nagging conundrum.

We now shift our focus to the micro – looking at the sector-level and company-level dynamics within the aviation space to try to diagnose these issues from a different perspective.  We do this in the hope that this approach may lead to some alternative explanations for this conundrum.

Before we dive head into the GCC aviation sector dynamics (which we do in Part 2), it is probably useful to take another step back and look at several of the overarching and persistent drivers for the airline sector globally.  These ‘big drivers’ include the following three long-run dynamics: 1) competitive intensity, 2) cyclicality & capacity utilization and 3) investment efficiency relative to capital intensity (or ROIC).  

Let us first look at competitive intensity – which we believe, at the end of the day, is THE most important driver for sustained profitability, not only measured in the magnitude of dollars earned by a franchise or sector, but much more importantly, the returns generated relative to the resources deployed (ROIC). 

The folks at IATA, with the help of Professor Michael Porter of the Harvard Business School, have mapped competitive dynamics quite comprehensively and elegantly in the 2011 seminal report they co-authored entitled “IATA Vision 2050”.  Chart 31 summarizes the broad competitive framework itself (Porter’s ‘Five Forces’) and characterizes, both textually and via assigned degrees of competitive intensity (high/rising is negative, low/falling is positive), how the global airline industry is currently positioned.  

It is worth noting that very little has changed in this broad picture since the 1980s-1990s when most of the regional airline regulatory regimes undertook aggressive deregulation steps.  We won’t delve too deeply here as this subject has been exhaustively vetted and analyzed by industry experts over the years (even though the IATA report is 70+ pages, we would highly recommend anyone keenly interested in airlines read this superb report front-to-back).  What is worth highlighting here is that the global airline industry is unique amongst almost all global industries in that it scores extremely poorly on all five forces (rivalry among existing competitors, bargaining power of customers, bargaining power of suppliers, threat of new entrants and threat of substitutes).  

Chart 31 – Competition - Five Forces in the Airline Industry

At the margin however, and especially more recently, one aspect (rivalry among existing players) has shown some improvement, particularly in the US and European markets (much less so in other regions).  Most of this improvement occurred in the aftermath of the global financial crisis in 2008/9 and the ensuing raft of regional consolidation that has taken place within the US and European legacy airline segments.  This has translated into an effective easing of ‘exit barriers’, better scale economies and improved capacity utilization as the cycle improved and demand growth accelerated.  Since 2009, most of the legacy airlines have been quite conservative in expanding capacity. 

Apart from this improvement, very little else has changed.  Aircraft suppliers remain concentrated oligopolies while large urban airports continue to act as local monopolies, both commanding resilient pricing power.   The revolution in greater ticket price transparency introduced by the explosive rise of internet penetration and online travel aggregators has in turn ratcheted up the bargaining power of customers, while their switching costs have remained very low.  Even though ticket prices on balance have fallen over the past 15 years in real terms, they still represent a meaningful share of discretionary spending for most travelers.  

On the general topic of competitive intensity, we thought the below time-line of public quotes made by Warren Buffet on the subject of the airline industry might be insightful in understanding how the industry has evolved since 2007 – at least through the unique perspective of one of the most successful investors of our time.  It is important to note that Buffet began to invest in the US airline stocks sometime in late 2016 (so after the second quote we listed below) and has since deployed over $10 billion of Berkshire Hathaway’s capital in the sector through investments in the top four dominant US legacy airline stocks (Delta Airlines, American Airlines, United-Continental Airlines and Southwest Airlines).  The final quote is from Ryanair’s CEO (Michael O’Leary) on his opinion of the likelihood of Buffet expanding his holdings to the European airline space.

“The airline industry’s demand for capital ever since that first flight has been insatiable. Investors have poured money into a bottomless pit, attracted by growth when they should have been repelled by it. And I, to my shame, participated in this foolishness when I had Berkshire buy U.S. Air preferred stock in 1989. As the ink was drying on our check, the company went into a tailspin, and before long our preferred dividend was no longer being paid. But we then got very lucky. In one of the recurrent, but always misguided, bursts of optimism for airlines, we were actually able to sell our shares in 1998 for a hefty gain. In the decade following our sale, the company went bankrupt. Twice. To sum up, think of three types of 'savings accounts.' The great one pays an extraordinarily high interest rate that will rise as the years pass. The good one pays an attractive rate of interest that will be earned also on deposits that are added. Finally, the gruesome account both pays an inadequate interest rate and requires you to keep adding money at those disappointing returns.” Warren Buffet 2007 Letter to Investors

“It [the airline industry] is a labor intensive, capital intensive, largely commodity type business and it’s been a death trap for investors ever since Orville took off…..Investors have poured money into [airlines and aircraft manufacturers] now for a hundred years plus, with terrible results.  If it ever gets down to where there was one airline, and there is no regulation, it will be a wonderful business.  Then the question will be whether having [survived through many bankruptcies] a relatively few [airlines] who are doing a relatively high percentage of the seat miles, whether it is a good business yet. I don’t know the answer to this, but I am skeptical. Warren Buffet - 2013 Berkshire Hathaway Annual Meeting

"It's true that the airlines had a bad 20th century. They're like the Chicago Cubs. And they got that bad century out of the way, I hope...It’s been a disaster for capital.  It’s got glamour to it, so you can always get some guy to put up money for an airline…..It’s a very tough business, because the marginal cost of the seat is practically nothing.  You have these huge fixed costs, and yet, if you take one more person on, there is virtually no cost to it.  So, you are very tempted to sell that last seat too cheap…..Unless the airlines operate in the well over 80% capacity - what kills you is when they have way too many airlines around - going to marginal cost causes you to go broke in the airlines business   The hope is they will keep orders in reasonable relationship to potential demand." Warren Buffet – February 2017 CNBC Interview

“It’s a business that is always subject to someone doing something very dumb competitively. They’ve done it allot in the past.  There was more chance of them doing it when there was seven of them than four….The industry was suicidally competitive for decades, they net lost money while they were growing like crazy in units….It can turn into fierce competitive battles and wipe out earnings, or it can be a business that is more decent, but still subject to lots of competition.  It’s really hard to know for sure how it will develop.  It is not risk-free of competition at all [unlike the railroad business].” Warren Buffet – February 2018 CNBC Interview

“What do you think it would take for Warren Buffet to invest in the airlines in Europe?”- CNBC anchor asks Michael O’Leary (CEO of Ryanair).

“I don’t know.  I suspect the same kind of environment as he’s seen here in the US. I think once he begins to see, as in North America, consolidation has played out, the airlines have more control over pricing power, more control over their costs, and you are into a much more sensible industry, which ultimately Europe is moving towards.  We are already about 10 years behind North America…then I think lots of investors will want to invest in this industry.” Ryanair CEO Michael O'Leary – May 2019 CNBC Interview

The deeply cyclical nature of the airline industry is also worth highlighting, even though it is not unique for globally integrated sectors.  What is somewhat unique here is the demonstrated predictability of these cycles despite the consistency of long-run demand growth (see Charts 32 & 33 below).  Chart 32 shows global industry operating margin trends from 1948-2017, spanning seven full cycles (down- and up-cycles).  Invariably, on average, most cycles have a duration of around 9 years (measured from peak-peak or trough-trough), with the most recent cycle still playing out (10+ years with no trough in sight...yet).  Passenger volumes follow a similar cyclical trend (Chart 33), with very short intermittent demand decelerations occurring every 9-10 years (on average), typically caused by a global systemic event, but always recover swiftly with incredible resilience to resume a sharply upward-sloping growth curve.  

Chart 32 – Global Airline Operating Margins Through the Cycles - 1948-2017

Chart 33 – Global Passenger Kilometers Flown – 1950-2018

From a capacity utilization standpoint (measured through an industry-specific indicator named a passenger ‘load factor’ – or the percent of available seat kilometers occupied by paying passengers) the cyclicality became less pronounced in the years post 1980 as the annual supply response relative to recovering demand began to decline (Chart 34).  Looking more closely at down-cycles over the past five decades (Chart 35) the pass-through relationship between capacity utilization and operating margins is also evident – margins tend to peak with load factors ahead of a downturn and trough once load factors have bottomed.  The exception to this long-standing relationship occurred in 2012, when high fuel prices disproportionately impacted margins, but recovered swiftly when oil prices crashed in 2015.  During the overall period (2010-2017), utilization remained very high and has continued to trend up, in what is playing out as one of the longest and most resilient air traffic cycles on record. 

Chart 34 – World Capacity Growth & Passenger Load Factors – 1950-2017

Chart 35 – Global Airline Down-Cycles: Passenger Load Factors & Operating Margins – 1970-2017

Just how strong this current cycle has been is illustrated in Chart 36 below.   While capacity in total aircraft kilometers (passenger KMs and freight KMs) has grown in line with past cycles (5.1%), traffic has grown at a faster clip relative to capacity growth (7.7% for RPKs, 6.5% for FTKs) – driving both higher load factors as well as supporting historically high operating and net margins.  The tailwind of lower fuel prices since the peak in 2012 has also been a big factor supporting profitability since fuel tends to make up between 25%-30% of total operating expenses.  The current cycle’s subdued capacity growth along with lower fuel expenses (particularly since 2014) has translated into a boon for airline profitability, so much so that beginning in 2010, the global sector once again became cumulatively profitable on a net profit basis (measured since 1965, see Chart 37 below).  By 2013, profitability had eclipsed the past record of cumulative profits last set in 2000 ($40 billion) – a real break-out from what was a volatile and stagnant long-run trend.  In the past three years alone (2015-2017), the sector generated over $113 billion in net profits, almost 6.3x the annual peak profitability it recorded in 2010.

Chart 36 – Down Cycles: Traffic Growth Metrics – 1978-2017

Chart 37 – World Airline Net Profits History – 1965-2017

The performance on profitability since 2013 is also remarkable relative to other indicators that track global cyclicality over the long-run.  Chart 38 below from IATA shows operating margins relative to global GDP growth from 1970 to 2018.  Up until 2013, margins tracked GDP growth almost in lockstep fashion, reacting to global recessions and demand shocks sharply within 12 months.  The trend since 2013 has been one of acute divergence, where operating margins have re-rated meaningfully above a sideways GDP growth trend.  Pundits like IATA point primarily to lower fuel prices and incredibly high load factors and have begun to pound the table, saying “it’s different this time”; (the title of Chart 38 sort of says it all). But is it ever “different this time” when it comes to persistently cyclical industries?

Chart 38 – Global GDP Growth & Average Airline Operating Margin – 1970-2018

This is a good lead-in to the third and final big dynamic at play – investment efficiency and returns on capital– what many believe is the true measure of sustainable value creation for the sector.  

Before we look specifically at global sector ROIC evolution, it is important to first understand yet another secular driver that has influenced profitability most acutely since the 1950s – pricing power (or the lack thereof).  Chart 39 (again from our friends at IATA; they have fantastic charts!) plots the unit cost of an airline ton kilometer (or ATK; in constant 2013 US$ prices) relative to the average yield on that kilometer’s revenue (RTK).  The long run relationship is clear and a pretty powerful illustration of where all the real value that has been created by the airline sector ends up going to – the consuming traveler.  

Invariably over the long run (almost 70 years, and pretty much since commercial air travel has been alive) any real cost savings the airlines have achieved due to lower fuel prices or better technology has been passed through swiftly via aggressive drops in pricing.  This was partly due to deregulation and the ensuing increase in competitive intensity, but is also testament to the strategy of pursuing growth in volumes at the cost of lower structural profitability.  We don’t want to oversimplify this phenomenon as there are obviously other things at play here too – e.g. cheap financing offered by aircraft vendors and supportive host governments, strategies focused primarily on volume penetration growth at lower price points (LCCs) and a generally weak culture of expenditure discipline at precisely the wrong points in the cycle.  Nevertheless, the complete absence of any real pricing power says volumes about how the industry’s returns on capital have been debased for an extended period of time. 

Chart 39 – Unit Cost and the Price of Air Transport – 1950-2018

This chronic debasement of capital shows up prominently in all the multi-decade ‘league tables’ of long run sector ROICs.  Chart 40 below, taken from a report published by IATA and McKinsey in 2013, titled “Profitability and the Air Transport Value Chain”, is one such ranking, showing average sector ROICs for 29 global industries from 1965-2007. The airline industry then ranked dead last (sub 7%) with the lowest returns of any industry globally and meaningfully below any reasonable estimate for the cost of capital.  Since then, performance has improved somewhat.  

Chart 41 updates this league table for the period of 2004-2013, just one year before profitability re-rated in the sector.  At that time, average ROICs had actually declined further (to 4% and still the league table worst), but soon began a recovery to a 2015-2018 average of 9.7%, well above its 42-year average ROIC. What is not surprising is that top-line growth for the airline industry during this same time span (2004-13) was 3rd ranked amongst global industries at 6.1%, outgrowing all other industries except for healthcare equipment and railroads.  The re-rating in ROIC for the industry in 2015 above what many believe to be the minimum cost of capital (7%-8%) was in fact trend setting as it was the first time since the mid-1970s that airlines collectively generated positive economic value or ‘EVA’ (note: EVA=ROIC-CoC or the cost of capital) where average ROICs were now printing in excess of the prevalent cost of capital (Chart 42 below).  

Chart 40 – Industry Returns on Invested Capital (ROIC) Ranking – 1965-2007

Chart 41 – Revenue Growth & ROICs – Global Industry Ranking – 2004-2013

Chart 42 – Return on Capital Invested in Airlines and Their Cost of Capital – 1993-2019

Once again, driving this sharp improvement in ROIC since the beginning of the upcycle in 2009, was as much an uplift in operating margins as it was due to higher capital productivity (Chart 43 below). Operating margins have however weakened since 2016, when fuel prices rebounded, and look set to track lower, purely based on where we are in the duration of the cycle.  2018 has proven to be a much weaker year for both unit revenue growth and unit cost growth (Chart 44 below) and may prove to have been a cyclical peak (at least in operating margins) if these dynamics persist into 2019.  

Nevertheless, capital productivity, the lynchpin to less volatile ROICs since 2009, continues to be very strong, anchored by strong and persistent capacity growth discipline – especially amongst the big legacy airlines globally.  What is however most concerning is the sharp deterioration in the global airline industry load factors (passengers, but also especially freight) since the end of 2018 (Chart 45 below).  Passenger load factors look to have peaked in February of 2018 at an elevated record level of 80.7%, but have held up relatively well for most of the 1Q of 2019. Freight load factors, which tend to be much more sensitive to global trade growth, have trended sharply downwards on a seasonally adjusted basis, and are currently at levels not seen since the period that followed the global financial crisis.  Was this due mainly to the trepidation brought about by the China-US trade disputes of late or is this a potential leading indicator of strong impending cyclical headwinds?  Time will tell.  

Chart 43 – Components of Returns on Capital – 2000-2018

Chart 44 – Airline Industry Growth in Unit Costs and Unit Revenues – 2011-2018

Chart 45 – Sharp Deterioration in Airline Load Factors in 2019

We are not, by any means, in a position to prognosticate on the global health of the airline industry nor are we attempting to predict the top of the current cycle (we will leave that to the likes of IATA). We are however very interested in analyzing what role the Middle East airline carriers have played in this past period of capacity growth. This is especially important since it may shed some light on how the global demand and supply dynamics may shake out through the next downcycle and who might be the winners/survivors when the cycle troughs. 

Continued in Part 2 (separate post - see the following Link).

Note to readers: The charts in the post are much better viewed enlarged in a separate window. To do so, please right click the chart and open image in a new panel or window.