Description

More a long update on this one from the note published last year (https://www.valueinvestorsclub.com/idea/Dogus_Otomotiv/4670110614), as since then including dividends paid the stock has returned over 150% in local currency (yet only about 50% in USD). With the stock trading on about 5x our revised-up 2022 earnings forecasts, and earnings expected to double again from there over 2023-24, we believe there’s still plenty of room to go, as on our revised-up 2022/23/24 (December y/e) EPS forecasts of TRY 13.1/18.2/26.1 the stock is still only trading on 5.0x/3.6x/2.5x P/E with still plenty of further growth over and beyond 2024 in our view.

Dogus posted TRY 11.8 EPS for FY 2021, up from TRY 5.2 in FY 2020. Stripping out one-off FX gains (which are however cash gains, stemming from the unprecedented ~eur220m customer advances the company was holding as of end-2021, reflecting ultra-long delivery times amid a scarcity of vehicles coupled with the wild lira depreciation in the last quarter of last year), we calculate underlying EPS stood at TRY 8.2, still comfortably ~60% up y/y, despite a severe shortage of vehicles which capped deliveries for the year at 95k, easily 30k units less than what could have been under normal supply conditions (Dogus sold 60k vehicles in H1 last year before supply shortages started to bite).

For the current year, all three listed Turkish car companies (Dogus, Tofas, and Ford Otosan) are presently guiding for broadly flat domestic sales volumes (Dogus’ present guidance is 97k units). Considering the extent of recent lira depreciation and ensuing deterioration of purchasing powers, plus the war in Ukraine delivering another blow to the Turkish economy, a flat market for this year may sound at first glance overly bullish. However, one has to consider that due to demographics and huge pent-up demand accumulated over years of policy-induced suppression, the Turkish car market is already pretty much near the floor of its potential. Any further undershooting can only be temporary in our view, and would only add to the already vast pent-up demand and long-term potential.

In addition, as of today due to policy (highest car taxes in the world, particularly for premium cars, and very tight vehicle loans limits) only about 15% of Turks can afford a new car, and only about 5% of Turks can afford a Dogus (VW Group) car. As the Turkish economy keeps on growing (Turkey has been in real and even more nominal terms the fastest growing G20 economy over the pandemic in 2020-21) and eventually incomes invariably catch up, so over time will the portion of the population that can afford a car and particularly a premium car (premiumization has been a ubiquitous feature of most car markets for quite some time). However, at present a car in Turkey is basically a luxury product, thus largely unaffected by low-income or even median-income purchasing power dynamics. Indeed, although light vehicle sales were down approximately 25% y/y in Q1 mainly due to lack of supply, they came above expectations and, once accounting for seasonality, broadly consistent with an outlook of flattish volumes for the full year, provided supply bottlenecks gradually ease over the course of the year, which remains the current consensus not only in Turkey but globally despite the war in Ukraine affecting the supply of some components (mainly wire harnesses). Notably, in the luxury segment (Audi, BMW, Mercedes and the likes) where a disproportionate % of profits are made by dealers (and especially by Dogus), there was hardly any contraction at all in Q1, despite the very high base and still severe supply shortages in Jan-Feb (in March they started to relax, particularly for importers like Dogus). For instance, for the Audi brand sales volumes even increased y/y in Q1, in the face of it all (lira collapse, as well as exceedingly challenging supply-side comps), as all carmakers including VW are deploying their limited stock of chips and other components to prioritize production of their premium brands and models. In our view, this also shows that it’s supply (or lack thereof) more than demand that is keeping the market down; but limited supply brings significant pricing and mix (hence margins) benefits.

As illustrated in the table below, we conservatively project gross profit per car to ease from eur2,824 in 2021 (and eur3,500 in Q4 2021) to eur2,555 in 2022, and eur2,350 in 2023-24. We believe it’s entirely appropriate to model ASPs and GPs/car in euro rather than lira because this is the reality of the Turkish car market really, even for domestic producers (which always have the alternative to export to Europe rather than selling domestically), with retail prices getting adjusted to the exchange rate almost on a weekly basis at times of high lira volatility, like at the end of last year. We don’t believe Dogus’ GP/car will ever revert to the ~eur1,500 average prior to 2018, because from 2018 onwards Dogus changed strategy, deliberately pursuing profitability vis-à-vis market share. We also project OPEX growth of 50% this year, 40% in 2023, and 20% in 2024, broadly in line with expected general inflation. We expect the TRY/EUR rate to keep deteriorating broadly in line with inflation differentials between Turkey and the Euro zone, although we do conservatively pencil in some real appreciation for the lira (negative for Dogus, as OPEX is virtually all in local currency). As far as Dogus sales volumes are concerned, we project 93k units this year (below guidance of 97k – this is wholesale sales volumes ex-Skoda where the dealer Yuce Auto is only 50%-owned by Dogus hence consolidated below-the-line as a subsidiary), growing to 150k by 2024. 150k is not pie in the sky, it’s still some 30k short of what Dogus was selling at peak in 2015-16, when the overall market (light vehicles) reached approximately 1 million units per annum (currently is 700k to 800k). One may argue that even 150k is unrealistic because Dogus is now pursuing profitability vs. volumes. However, to be fair the focus of the entire car industry globally has shifted to profitability since the pandemic, and after all we are projecting GP/car in 2023-24 to be only about eur800 higher than the pre-2018 (very low) average.

More importantly, we believe that there’s plenty of growth potential in the Turkish car market, after nearly a decade of artificial demand suppression due to policy, namely the highest car taxes in the world, particularly for premium cars. Even if policy never changes, we believe that the market should over the next 2-3 years be able to recover and surpass its prior 1 million units peak just by virtue of economic/demographic growth coupled with huge pent-up demand. After all, as witnessed by the experience of other high-tax car markets, in the long run taxes only slow down the convergence to the ‘normal’ level of car penetration, largely dictated by income levels, demographics, and geography (large vs. small countries). We believe that Turkey’s ~15% LV penetration is way too low for a vast, young, and growing (including demographically) middle-income country. The country’s existing (tiny) car fleet is already very old (median age over 13 years), and imports of second-hand cars are prohibited to protect local manufacturers, hence in order for penetration to go up, new car sales must go up materially from the present level of only 700k-800k a year. Germany, a car-saturated ageing country with the same population as Turkey, typically buys 4 million new light vehicles a year. Eventually, we believe Turkey should be able to reach about half Germany’s level.

We believe that policy reform will most likely come next year if, as polls overwhelmingly suggest, a change in government were to occur. Car policy is of course only one aspect of Erdogan’s extravagant economic policies; the bigger picture is even more dystopian. The united opposition has already made clear that economic policy normalisation will be the top priority for the new government. First and foremost, interest rates will be hiked to a level closer to inflation. The policy rate in Turkey was lowered in H2 last year from 19% to (present) 14%, thus prompting yet another lira collapse and (via higher import prices) a dramatic spike in the CPI from under 20% to over 60%. Such insanity reflects more Erdogan’s religious beliefs (‘fight against usury’) than economics of course, and perversely resulted in a tightening of financial conditions, as medium to long term interest rates actually increased as a result of soaring risk premia, thus hurting the economy for virtually no one’s benefit really. Worth noticing, despite a strong fiscal position (debt to GDP trending flat to down, and under 40%), Turkey’s local-currency 10-year government bond yield of ~22% is almost level with Ukraine, a country that this year will probably see its GDP sadly halve, whereas for Turkey the IMF still projects 3.3% real GDP growth this year (consistent with >50% nominal growth), in the face of it all (macroeconomic policy insanity, hyper-inflation, war in Ukraine, ongoing pandemic), which is a testament to the country’s strong structural growth drivers (a young and growing population, an increasingly more sophisticated manufacturing and export hub, already virtually part of the EU common market by virtue of its customs union, thus poised to benefit dramatically from the shortening of supply chains).

In our view, monetary policy normalisation will actually benefit the economy even in the short term (via lower risk premia hence lower medium to long term interest rates), and at the same time also fix at once the perennial problem of financing the current account deficit, because in our view, once pursuing an orthodox policy amenable to global investors, funding a low single-digit % CA deficit should be very easy for a country with such strong fundamentals. Basically, foreign investors at present shun Turkey only due to Erdogan. The Turkish stock market is on less than 6x P/E, by some measure the cheapest major equity market in the world, and foreigners’ holdings of Turkish stocks and bonds keep hitting all-time lows. Remove Erdogan, and the sky is the limit in our view, for the economy and for asset prices.

The next presidential elections are due in June 2023, and the united opposition is expected to win by a large margin with virtually any of the three possible candidates (https://en.wikipedia.org/wiki/Opinion_polling_for_the_2023_Turkish_presidential_election). We don’t believe a scenario of postponing/cancelling the elections is realistic, as feared by some foreign investors. The army, which as recently as 2016 attempted a coup to oust Erdogan, wouldn’t allow it in our view. And anyway Turkey, an open economy (both exports and imports are over 30% of GDP) with no significant natural resource, fully integrated into the EU common market and supply chains, and a NATO member, couldn’t possibly survive as a global outcast. To be fair even Erdogan, for all his faults, has always made a point of drawing his legitimacy from the ballot box. True, he did force a re-run of the Istanbul mayoral election in 2019 on bogus claims of electoral fraud, however after the opposition won for the second time (by an even wider margin) he respected the result. Just a few days ago, Erdogan criticized the Tunisian president’ decree dissolving parliament as a “smearing of democracy” and a blow to the will of the Tunisian people; this doesn’t sound like someone about to suspend parliament and cancel elections.

What happens to cars under a new political regime? Worth noticing, the tiny Turkish car market is at present largely a cash market, as policy strongly discourages car loans especially for premium cars. Even after a recent loosening of limits, the max allowed loan-to-value and loan maturity for cars with a retail price of max TRY 400k/800k/1.2m/2m (approximately eur25k/50k/75k/125k) is respectively 70%/50%/30%/20% and 48/36/24/12 months, while no loan at all is allowed for cars of over eur125k retail price. This all makes car financing barely viable only for small cars really. Thus, we believe that, even if car policy is left untouched, macroeconomic policy normalisation would already provide a net benefit at least to the premium segment, as stronger economic growth and a stabilising lira/inflation (boosting purchasing powers and affordability) is likely to more than offset the negative impact of a higher policy rate on vehicle loans that for premium cars are either not there at all or are anyway largely irrelevant to purchasing decisions.

In addition to that we believe that the virtual overnight fixing of the current account ‘problem’ via macroeconomic policy normalisation would also allow in short order a loosening of the extraordinary policy restrictions, not only on car loans but more importantly on the special consumption tax (SCT). At present, Turkey levies the highest taxes in the world on new car (passenger vehicles) purchases, escalating as a function of the pre-SCT retail price (but inclusive of 18% VAT, meaning the SCT is applied even on the VAT!). The SCT ranges from 50% for cars costing less than TRY 200k (eur12,500) all to way to over 200% for luxury cars. The dystopian world of the Turkish car market, made of ultra-old vehicles on the road and brand-new Porsche Cayennes costing US$600k (vs. US$150k in the US for the same model) is well explained in this article here: https://jalopnik.com/why-this-150-000-porsche-costs-600-000-in-turkey-1847572799.

Turkey’s seemingly insane car policy has one justification: reducing the current account deficit, a perennial problem for the Erdogan regime as foreign investors shun the country, largely due to Erdogan’s extravagant economic policies. Turkey has a fairly robust domestic car industry, with the likes of Ford, Renault, and Fiat/Stellantis having long picked the country as one of their key manufacturing hubs. In 2021 Turkey produced 1.276m light vehicles (783k PVs and 493k CVs), of which 937k were exported, virtually all to Europe. By keeping car taxes so high, the government manages to suppress domestic demand (773k last year) hence imports (426k, still 55% of the market in unit terms and much more in $ terms), which makes the automotive industry a key positive contributor to the current account along with tourism (energy and gold imports being instead the biggest drags). Import duties would be of course a more efficient way to achieve the same objective, but they are incompatible with the Turkey-EU customs union, on which the Turkish economy, and particularly the auto sector, sorely depends.

Turkey’s SCT policy on cars has been there for over a decade, but it’s only over the past 7-8 years that the SCT has been progressively hiked and steepened (i.e. made increasingly progressive) to the present truly exceptional levels, which goes a long way to explaining the market’ and particularly Dogus’ poor sales volumes performance from 2017 onwards. To be fair, in the absence of supply constraints, last year the market should have definitely recovered its prior peak of one million units. But even one million is in our view probably only half of Turkey’s potential. Hampered mobility is increasingly hindering economic growth, which makes the policy unsustainable in the long run in our view.  

Besides, the policy is increasingly proving self-defeating. True, Turkey has managed to reach a strong position in commercial vehicles (Ford Otosan is the champion there). However, by suppressing domestic demand and diverting it as much as possible towards small/cheap petrol/diesel cars, unsurprisingly in PVs Turkey has managed to attract manufacturing of only small/cheap petrol/diesel cars (less than 1.6L engine). Although Turkey is part of the EU car market thanks to the customs union, it’s not part of the EU proper, so for an OEM to put up a plant in Turkey vis-à-vis say Slovakia or Bulgaria the key attraction must be to have a sizeable enough domestic market to support it. There’s a reason why for instance Tesla has decided to set up shop in Germany rather than Turkey, where it could have benefited from significantly more competitive costs; Germany provides the key attraction of a domestic market 5x larger than Turkey. As OEMs increasingly focus on their premium brands and models and on EVs, Turkey’s PV manufacturers look more and more doomed. The government has tried to address that by setting up a domestic premium EV brand, TOGG, which should start producing in 2023 (at present EV penetration in new car sales is less than 1%, mainly due to lack of EV-specific incentives, uncompetitive electricity prices vis-à-vis gasoline and diesel, and inadequate charging infrastructure). To support TOGG, from next year the government may start handing out significant incentives to consumers for the purchase of EVs, thus benefiting first and foremost Dogus in our view, as VW is the EV leader in Europe and Dogus/VW is the EV leader at present in Turkey (any consumer-side incentive will have to be for all EVs not just TOGG, again not to infringe on the common market rules). However, TOGG faces many challenges. Although on paper a private enterprise owned by a few Erdogan-friendly business groups, in reality TOGG is 90%+ funded by crony-lending from the state-owned banks (which by next year will have grown into billions of dollars), and like most other EV hopefuls is unlikely to turn profitable for several years. Most likely, a new government will just cut its losses and fold it. Even if TOGG does come to light, with only 27,500 vehicles targeted to be produced next year, it won’t move the needle for the country’s automotive industry. Ultimately, in order to attract new international OEMs to the country, or even just to support the migration of existing plants to models that actually have a future (higher-end and electric), car policy will have to be reformed, with the SCT substantially reduced and made much less progressive, at least for EVs, with Dogus being the primary beneficiary.

All of the above to say that our forecast of 150k new car sales by Dogus by 2024, consistent in our view with a market size of 1 to 1.2 million units, is in our view not outlandish at all, and actually leaves plenty of room for further growth over and beyond 2024. In addition, we are also very conservative in our view on spare parts/repairs and second-hand, projecting their contribution to keep at around 20% of Dogus’ total gross profit in 2023-24. In mature markets, dealers earn the majority of their gross profit from repairs and second hand. As virtually no car ever sold in Turkey over the past 20 years has been retired (unless severely accidented), we believe there’s plenty of growth potential in Dogus’ spare parts and repair business, which (as always the case all over the world) has significantly better margins than the new car sales business, and typically is also much less cyclical. Last but not least, second-hand contributes insignificantly to Dogus’ present profitability, as the Turkish second-hand car market, albeit very large (as of last year approximately 8x larger than the new car market in volume terms), remains for now mostly informal. However, this is starting to change, and Dogus is at the forefront of such evolution, with its second-hand brand DOD recently partnering with n11.com (Dogus parent’ e-commerce company, one of the largest in Turkey, and which may soon IPO) to deliver a potentially all-online sales experience supported by AI and with the car delivered to the buyer’s home, Carvana-style.

For all the reasons highlighted above, we believe Dogus present valuation of around 5x current-year P/E (on a realistic set of assumptions in our view, below-guidance) is extremely attractive given the secular growth potential. Looking 5 years down the road, with growing incomes, macroeconomic normalisation, likely car policy reform, and strong visible growth in second hand and repairs, the stock may be sitting on as low as 1x to 2x P/E, in our view. Mind Dogus is a small-cap (less than US$1bn market cap), but not a small company. With a dealer network of 400+ stores, it has a footprint comparable to the likes of Zhongsheng in China (Bloomberg 881:HK), another world-class enterprise (like Dogus is, in our view) and a former holding of ours, which boasts a market cap of well over US$15bn.

Whilst the war in Ukraine, if protracted as it seems, may hamper both demand and supply of cars including in Turkey (Turkey is a significant energy and grains importer, much of it coming from Russia and Ukraine, hence clearly a net economic loser from the conflict), on the other hand one must give Erdogan credit to have thus far deftly played all sides of the conflict (US/NATO, Ukraine, and Russia), using the war as an opportunity to raise Turkey’s profile in all quarters and emerging as the only credible mediator for any ceasefire and eventually hopefully peace. This has allowed Turkey to avoid imposing any sanction on Russia, thus benefiting for instance from much-discounted Urals imports, keeping open skies and flights with Russia etc, which should help mitigating the impact of the conflict on the Turkish economy, all the while without upsetting the US and Ukraine, which remains dependent and grateful to Turkey for supplying its lethal Bayraktar drones.

Last but not least, we like Dogus’ consistent free cash flow generation (ex working capital fluctuations, free cash flow broadly matches net earnings, as the company has recently completed a major capex cycle) and reliable dividend policy (50%+ payout), resulting in approximately 20% dividend yield at current price if one looks two years down the road (~9% yield on the actual 2021 dividend just paid). As Dogus parent (Dogus Group, one of Turkey’s largest and oldest conglomerates), holding 75% of Dogus Oto, is highly leveraged (no recourse to Dogus Oto, which in turn is modestly leveraged), the dividend looks highly secure. Stock liquidity is an issue (only 15% free float), but the stock does trade a few million dollars a day on average. Besides, one day the company may tender the treasury shares (10%), which may go a long way to improve liquidity; however, selling the treasuries has been a moving target for years now, with the company understandably getting cold feet every time the issue resurfaces, ostensibly due to the low valuation. 

Catalyst

earnings growth + potential macroeconomic stabilisation and policy reform