Description

Delivery Hero

Delivery Hero is a leading global food delivery platform, delivering over three billion orders annually across 74 countries, which collectively contain over two billion people. 

Our general thesis posits that market leadership is a necessary condition for strong economic profits for a food delivery platform; leadership drives network effects and scale advantages, which together underpin the durable economic moat.  Market leaders offer consumers the greatest selection of restaurants and the densest delivery fleet, which in turn enables them to deliver food faster, hotter (better), and at a lower cost than competitors.  In 90% of its GMV, Delivery Hero is the largest, most advantaged market participant.  In 75% of its GMV, Delivery Hero is four times larger than its next largest competitor, implying a market share over 80%, with a commanding advantage.  

And yet, as the following chart shows, and without any additional information, the profit trajectory of Delivery Hero looks bleak and hardly worth its current €11.5 billion enterprise value (€7 billion of equity value, behind an additional €4.5 billion of net debt.)  Little in the chart below suggests significant forthcoming profits, never mind profits large enough to pay back €4.5 billion of debt.  As we will explain, these poor financial optics provide the grist for the market’s massive mispricing.  If that is correct, then when the financial optics improve dramatically, as we project, the mispricing should correct itself – hopefully violently.

With that as a backdrop, perhaps it is not surprising that the stock is on its lows, down 40% year to date, and 80%+ from its peak. 

The discussion that follows is organized into four subsections: the first two exploring each of the central tenets of the investment thesis, followed by an attempt to explain the market’s error, as we see it, and a brief discussion of what might go wrong:

• Margins.  Delivery Hero will earn 6.5% margins (EBITDA as a % of GMV .)

• Growth.  The business will more than double GMV in five years (i.e., 15%+ growth rate.)

• Market Reaction.  Our interpretation of why the market obviously does not believe the foregoing, and therefore why the stock has languished.

• Key Risks.  Opportunities to earn such high returns nearly always come with risks.

If the ‘Margins’ and ‘Growth’ assumptions above are realized, the value of the investment should increase tenfold from its current levels.  Here’s the very quick math: 2023’s GMV is expected to be €45 - €47 billion.  €100 billion of GMV at 6.5% margins = €6.5 billion of EBITDA.  At that point, we would expect it to be growing profits 15% - 20% annually, which would suggest that it is worth €100B+, or more than 10x today’s value of ~€7 billion.  The next two sections lay out the basis for these assumptions.

Margins

Management has a long-term margin target of 5% - 8%.  We are currently using the center of that range, 6.5%, as our Base Case.  At year-end, we estimate the run-rate margin to be over 1% (roughly consistent with current guidance), up from -6% in 2019.  

A. “The trend is your friend.”  Our Base Case projections do not call for an inflection, but rather a continuation of past trend (in fundamentals).

Note: For 2023, 0.6% is a blend of roughly ~0% in the first half and ~1% in the second half.

That is, management has historically improved margins by 100bps – 150bps annually, so the core bet is that they can maintain the prior pace of margin improvement.  

B. Disaggregated view.  A more sophisticated analysis disaggregating the business into its constituent parts/countries helps explain the mathematical mechanism driving the “trend.”  Delivery Hero’s business can be broken into the following four parts: (i) profitable countries, (ii) unprofitable leadership countries, (iii) unprofitable non-leadership countries, and (iv) Dmarts.  As the data presented herein shows, the profitable markets are growing profitability rapidly, at the same time as the unprofitable markets are inching toward profitability, quarter by quarter.  That mechanic drives the foregoing trend.

(i) Profitable countries (75% of the business)

Includes many of Delivery Hero’s most scaled and mature markets, and cuts across regions, regulatory regimes, and population density structures (e.g., Korea, Kuwait, and Sweden are all mature and highly profitable markets in this category)

Even some sophisticated market participants – looking at the consolidated P&L shown earlier – regard food delivery as a VC bet, with unproven unit economics.  That is simply wrong.  In our investment case, as discussed earlier, we are postulating that we believe that the company can get to 6.5% GMV margins.  If it can get margins merely to 3% (less than half of what we think), it would still be an exceptional investment at these levels – 4x+.  Profitable markets are already at 3.8% as of year-end 2023 and still improving rapidly!  Food delivery has already proven itself to be an attractive business in numerous countries around the world.  Not tomorrow, but today.

This is not a venture bet.  Importantly, Delivery Hero’s best markets are already generating the targeted 5% - 7% Adj. EBITDA margin (% of GMV).  The entire business is trading for less than 10x the current run-rate EBITDA of profitable markets.  Not bad for that profit trajectory shown in the chart, with the remainder of the business free.

(ii) Unprofitable leadership countries (15% of the business).  

Includes markets like Egypt, Spain, Romania, and the Philippines, where Delivery Hero has leading market position, but the market is relatively higher-growth and earlier-stage than in the currently profitable geographies.

Importantly, as shown in the following chart, the unprofitable markets have a higher gross margin profile than the profitable markets but are unprofitable at the EBITDA level because they are spending significantly more on marketing and growth (6.4% in unprofitable countries vs. 1.0% in profitable countries) and are relatively subscale (6.9% opex in unprofitable countries vs. 2.3% in profitable countries), both of which will naturally normalize as the markets grow, just as the profitable markets did.  Marketing expense is higher, in large measure because the company is investing, at high ROIs, to drive supranormal growth.  Opex is higher because these markets currently have a smaller GMV base over which to amortize fixed operating expenses.  That is, there is nothing structurally unattractive about the unprofitable markets; if there were, the gross margins would be lower.  They are simply earlier-in-their-lifecycle, faster-growing versions of the profitable markets.

Supporting this idea, the total unprofitable portfolio (which includes the non-leadership markets) is improving quickly.  Here again, the trend (of the fundamentals) is your friend.

Therefore, there is every reason to believe that these leadership markets will develop into profitable markets.  Of course, that outcome is not certain, but the trajectory, and the margin structure shown above strongly support the contention, or at least appear to skew the odds heavily in our favor.

(iii) Unprofitable non-leadership countries (5% of the business)

Without market leadership, strong profitability is essentially unattainable.  The #4, the #3, or even the distant #2 are fundamentally disadvantaged, outside the very moat you want to be inside.  Accordingly, the 5% of Delivery Hero’s footprint where it does not hold market leadership has little prospect of achieving its margin targets.  As a result, we regard GMV from these countries as far less valuable than GMV from market-leading geographies.

But while these countries are less valuable standalone, they have very high “takeout value” to a potential in-market competitor.  If a #2 player can leapfrog to #1 by buying #3, that’s incredibly valuable; likewise, if a #1 can grow to become a more dominant #1, that’s even more valuable.   Consistent with this framing, management publicly confirmed a recent rumor that they are in talks to divest seven Southeast Asian markets, the region that contains the majority of unprofitable, non-leadership GMV.  We believe a deal makes sense for all parties and would, if consummated, prove to be a near-term catalyst for the stock.  

(iv) Dmarts (unprofitable) (5% of the business)

In 2019, Delivery Hero launched its “Dmart” business, through which it offers fast (sub-30-minute) delivery of grocery items to its customers. Unlike in its food delivery business, where Delivery Hero is simply a platform, in its Dmart business the company owns and operates nearly 1,000 micro-fulfillment centers. Operating stores markedly improves the customer experience compared to simply partnering with a traditional grocer: Delivery Hero can optimize Dmart layouts for picking (rather than in-store shopping), resulting in faster delivery times, lower costs, and fewer unexpected stock-outs.

Over time, owning the stores ought to also confer greater profit potential than grocery partnerships, as Delivery Hero can capture the full merchandise margin on each item, in addition to the structural cost efficiencies around picking and inventory management.  However, in the initial phases, taking an owned approach results in greater losses.  Each new Dmart initially loses money until it achieves sufficient order volume to break even on fixed costs.

As Delivery Hero has scaled volumes across its still-young store base, margins have inflected rapidly.  Gross profit (inclusive of merchandise, picking, and delivery costs) has increased over 1,000bps year-over-year and reached breakeven over the summer.

On an EBITDA level, the Dmart business as a whole likely will not reach breakeven until 2024 or perhaps 2025. But across seven best-in-class countries, the Dmart business is already EBITDA-positive (and has been for over a year), evidencing proof-of-concept that the economics can be very healthy at sufficient order scale. 

Disaggregating the business into four separate parts clarifies that the launch of Dmarts and losses in the unprofitable markets mask underlying improvement in every segment, except the portion of the business – (iii): unprofitable non-leadership – that they are actively divesting or exiting.

C. Drivers of the improvement.  The thrust of the argument for extrapolating trends for the various segments of the business rests on the premise that these improvements will continue – the profitable markets will become much more profitable, and the unprofitable markets will continue their upward trajectory through breakeven to target profitability.  But how do we know, or why are we so confident that those trends will endure?

Numerous durable and predictable operational developments have driven these financial trends (for both the profitable markets and the unprofitable markets) and will continue to drive the segment-level financials over the next several years because they are patently incomplete.  By understanding these at a granular level, at the level of the humans ordering food and humans delivering food, we can make reasonable forecasts about their effect on the near future.  That is, we are not relying on the trend itself, but rather on the driver of the trend.  The following chart details management’s bridge to long-term margins:

Source: Delivery Hero management

The bridge’s conservatism is hidden in plain sight in two crucial ways (“pink bars”):

(i) “Marketing & opex” expenses are collectively already closer to 3% in mature, profitable markets compared to 6% in the bridge; we therefore expect substantially more leverage than they show here.  As the business grows, it will continue to get leverage on these lines.  As we noted earlier, the unprofitable (high growth) markets are currently at 13% across these two line items, but that, remember, is a deliberate choice (with respect to marketing spend).

(ii) Management includes 1.6% -4.6% of “Risk/margin for error,” which is a negative fudge factor, arbitrarily high in our estimation, especially at the high end of the range.

We also expound on a few key elements of the bridge below (“green bars”):

• AdTech.  Delivery Hero derives revenue by charging restaurants for “top of page” placement, a business model invented by Google, but copied well in a variety of contexts.  The revenue is nearly pure profit.  In e-commerce, AdTech is well-trodden ground (companies like Alibaba, Etsy, and Amazon generate substantial ad revenue, 4% - 7% of GMV).  Effectively functioning as price-discrimination, AdTech revenue allows platforms to increase realized take rates on merchants for whom traffic is most valuable.  Delivery Hero itself has already been quite successful with this operational lever, having achieved management’s near-term target of 3% - 5% of GMV in select counties.  Other global peers have had similar success.  

To date, Delivery Hero has increased AdTech revenue from 0.7% in Q1 2021 to 1.7% of GMV in the most recent quarter; even more noteworthy, excluding Korea, they are already at 2.7%, because they just launched Korea several months ago.  Importantly, building revenue by developing robustness to the ‘placement’ auction takes some time – recruiting restaurants and then demonstrating a sufficient Return on Ad Spend (ROAS) so that restaurants ramp spending.  With ~80% share in Korea, and an unmatched selection of restaurants, and lower-than-average take rates, Korea’s AdTech revenue should be structurally higher than average, even though it is currently weighing down the reported average substantially.  Some countries have already hit their LT target of 3% - 5%.  If 1.7% of Adtech revenue were to increase to 4%, that alone would represent a 2.3% margin expansion opportunity, enough, as a single profit lever, to deliver a phenomenal return, especially if they hit the high end of the range, as the CEO believes they will:

“Best-practice markets today are advertising ~3.5% of GMV.  We don’t think we have any markets where we won’t be able to get to at least that level over time.  Then the question will be whether that 3.5% level can go to 5, 7, 8% over time in some markets.”  

- Niklas Östberg, CEO, Delivery Hero

• Order Stacking.  The company today leverages stacking – when a rider delivers multiple orders on the same route – on approximately 20% of its volume, up from virtually no stacking just a few years ago.  Delivering multiple orders, especially in dense cities, reduces cost per order meaningfully without extending delivery times.  If stacking penetration were to reach 40% over time – well below best-in-class markets (which exceed 60% today and are still increasing) – we estimate that would, by itself, drive more than 500bps of gross margin improvement (the bridge above depicts 100 – 300bps).  In other words, order stacking alone also has enough potential to reach the Base Case margins.

• Pricing.  The bridge includes delivery fees of 0.5% - 1.0%.  With an average order value of ~ €14, a 1% fee increase equates to €0.14/order, hardly enough to influence consumers negatively.  Importantly, Delivery Hero has already rolled out small service fees across all but a handful of its markets, and so has already breached the “psychological” consumer barrier of introducing a new fee.  From here, simply increasing service fees slightly (particularly for consumers with a higher willingness to pay) is far less likely to influence behavior.

• Basket Size Increase.  Delivery Hero generates revenues as a percentage of food value, while it pays its couriers a (roughly) fixed per-order fee.  Any increase in commissions from increased basket size therefore falls to the bottom line.  As a result, when the company takes actions like selectively increasing minimum order values or nudging consumers towards marginally higher-priced restaurants, or items, margins mathematically expand.

Growth

Since 2019, the business is roughly 6x bigger (largely organically!) If, with a jeweler’s loupe, you carefully study the chart below showing Delivery Hero’s GMV over the last several years, it appears that the company properly qualifies as a growth business.

While our forecasts embed far less growth in the future than in the past (~15% per year going forward as compared to more than 50% per year over the last four years), we are still expecting Delivery Hero to deliver meaningfully more (~2x) food in 2028 than in 2023.  What is driving this growth?  First, global benchmarking suggests that some countries are much further on the penetration curve for food delivery apps than others.  In South Korea, Kuwait, and China, for instance, there are 15x - 25x orders per capita per year compared to an average of 9x in the United States, and only ~1.5x across Delivery Hero’s emerging markets estate.  With much of its footprint in regions like Latin America, Eastern Europe, and the Middle East, Deliver Hero’s secular growth will be fueled by increased internet penetration and adoption in these higher-growth economies.  Moreover, once people use the platform, they tend to have fairly predictable usage patterns, where the amount of food that each cohort orders increases modestly each year, driven, no doubt, by the continually improving product offering: more restaurants, faster delivery times, and better UX on the app, etc.  So, each year, the company continues to add new cohorts of customers (spending €1 – €2 billion annually on customer acquisition), while old cohorts all grow GMV modestly – and this growth continues even in some of the most-penetrated food delivery markets in the world.

In the Market Reaction section of this discussion that follows, we address why growth has slowed since 2021 (shown in the green box above), which is central to the bear case and, we believe, to the market’s reaction.  We should note, though, that at these valuation levels, absolutely no growth is required to deliver exceptional returns.  At 6.5% EBITDA margins (which is achievable without growth ), at €47B of GMV 2023, that’s EBITDA of €3 billion.  That should be worth at least €20 billion, and probably at least €25 billion, even if growth slowed to GDP levels.  Thus, we would submit that the investment debate should revolve around margins, much more than around growth.

Market Reaction

And yet, the stock is down nearly ~40% this year, even though everything that we have explained is in the public domain.  We can find three primary narratives explaining the stock performance:

1. Fatigue with losses.  However carefully we just painted the picture, the losses and diminutive 2023 EBITDA reported at the top of this section dominate the narrative.  The Company has invested billions (in opex and capex) and is only now reaching cash flow breakeven.  Delivery Hero is also believed to be a venture bet, with an unproven business model.  Because it is listed in Germany, it is widely believed to be a European business, even though only ~15% of GMV is European. 

Until there is meaningful consolidated cash flow, there is no true valuation support.  But if our projections are correct, that should change as soon as 2024, which isn’t that far away.  Crucially, we believe that management understands the importance of, and is highly aligned to drive, profitability improvements.  A divestiture that accelerates profitability improvements is far more likely than any further step-up in investments.

2. Slowing growth.  While the business is 6x larger than it was in 2019, nearly all that growth occurred in 2020 and 2021, aided by the COVID consumer’s affinity for food delivery (order frequency rose when restaurants were closed and lockdowns in effect).  That benefit is even more clear with hindsight.  Most of these gains were permanent, as new consumers who tried food delivery because of lockdowns have stuck with the service post-COVID, but some did not, and reversed once restaurants re-opened.

In the first half of 2023, GMV grew a paltry 5%. Underneath that headline, GMV grew 17%  on average in three of its four regions (Middle East, Americas, and Europe), and shrank 2% in Asia, basically because in Korea and Taiwan , COVID lockdowns extended through April and May 2022, respectively; as 2023 lapped these non-durable benefits, growth slowed.  That deceleration reflected COVID dynamics more than secular trends, a contention evidenced by Korea’s recent return to growth in the second half of the year now that the COVID comparisons have eased.  Despite Korea’s high penetration, we do not believe growth has “run out.”  In our Base Case, we still project growth of 7% - 8% annually as many Koreans are still not food delivery users (we estimate the industry has perhaps 60% - 65% of internet-connected, non-rural users), and historical cohort behavior suggests that existing users continue to increase order frequency over time.  In addition, even in developed markets like Korea and the United States, general e-commerce continues to grow steadily at 8% - 9% per year – driven by increasing frequency.  We see no obvious reason why food delivery ought to do markedly worse than this over time.

Moreover, the rest of the company is largely on the other end of the penetration spectrum, highly unpenetrated with respect to food delivery; penetration will tend toward that of more mature markets, driving substantial growth.  Excluding Asia, the business grew 18% (again constant currency) in the most recent quarter – in a tough economic climate and while prioritizing profitability over growth.  Nonetheless, there is a view, which we think is dead wrong, that the business can’t improve profitability and grow simultaneously.  Indeed, Delivery Hero is doing just that in many countries, but again, the consolidated figures obfuscate that reality.  

Key Risks

Below, we present key risks to the thesis, any of which may also be bearing on the stock this year.

• Competition.  By far, the predominant risk to our thesis is competition.  As discussed, Delivery Hero has more than enough levers to increase profit margins to target levels.  Our analysis implicitly assumes that these benefits will accrue to the company and will not be competed away.  Capitalism generally works hard to reduce such value-capture, as many an airline investor has learned the hard way.  In this case, we believe that scale benefits and network effects discussed create a wide moat that insulates food delivery market leaders from such competitive pressures.  

This view is deeply rooted in empirical experience: Delivery Hero’s margins in more mature, profitable markets are already high (and growing rapidly), as are Meituan’s (China), iFood’s (Brazil), and Just Eat’s (Germany and Netherlands) – all of whom have demonstrated consistent and high profitability through market leadership.  As importantly, whenever competitors have attempted to attack a dominant market leader via price promotion and discounting, burning hundreds of millions (and sometimes billions), they failed.  No (plausible) amount of discounting could compensate for the twin disadvantages of a weak value proposition to consumers and restaurants and an uncompetitive cost structure.  

Even with low-cost capital deployed indiscriminately, none of the following attacks against a well-run market leader (with a solid first-party offering ) succeeded:

- Amazon (in the US)

- Alibaba (in China)

- Noon (in the Middle East)

- Uber (in the Middle East, India, and Korea) 

- Sea (in Southeast Asia)

- Coupang (in Kofundrea)

We have watched smaller competitors launch competitive attacks against Delivery Hero in its leading markets numerous times and have observed a consistent pattern:

• The competitor launches aggressive discounts (often 50% or more).

• They garner perhaps 5 - 10 percentage points of market share largely from low-quality, highly price-sensitive consumers.

• But they fail to gain material share beyond that and do not attain market leadership.

• Delivery Hero experiences a moderate impact to reported GMV growth (due to the 5 - 10 percentage point share shift) but little-to-no impact to EBITDA (those customers were generally low-profitability and price-sensitive anyway).

• The competitor eventually pulls back, as the large losses become unsustainable.

• Once unsustainable discounts are suspended, even the small number of price-sensitive customers who shifted ultimately return to Delivery Hero, because it is structurally less expensive than competitor offerings.

Most recently at the end of 2022, in Saudi Arabia, Delivery Hero’s second-largest market, Jahez, the #2 player, launched one such aggressive discounting campaign.  While this “succeeded” in boosting their growth rate briefly (at the cost of a substantial hit to their profit margins), it barely impacted Delivery Hero’s Saudi Arabia business, which continued to grow GMV and EBITDA through the entirety of the discount campaign period.  This year, Jahez reduced promotional behavior, and Delivery Heros’ market share and EBITDA reached record highs – at levels higher than pre-promotion.

Relatedly, we are carefully monitoring the competitive landscape in South Korea, Delivery Hero’s largest market.  In Korea, Delivery Hero’s platform, Baemin, has approximately 75% - 80% market share, and competes with two players who each have around 10% market share: Yogiyo and Coupang Eats.  PE-backed and levered, Yogiyo is run maximize cash flow and is not a significant competitive threat.  Coupang Eats, owned by Coupang (the largest Korean e-commerce platform, similar to Amazon), has historically been more aggressive.

In late 2021, Coupang launched a typical “attack” that employed “usual” promotional campaign tactics; notably, we estimate that they spent more than half a billion dollars on food discounts for consumers, coupled with aggressively discounted restaurant commissions across Seoul.  At its peak, Coupang’s market share reached ~15%, up from 8% - 10% pre-promotion.  Throughout this campaign, Delivery Hero’s Baemin continued to grow rapidly and remained highly profitable.  In 2022, when Coupang subsequently pulled back to reduce losses, its market share fell back to pre-campaign levels.  In other words, Coupang shareholders were left substantially poorer with little gain to show for it.

This year, beginning in Q2, Coupang took another run at Baemin, this time structured slightly differently.  Rather than direct promotions, they have included a blanket 10% discount on Coupang Eats orders for anyone who is a member of its “WOW” membership program (similar to Amazon Prime).  So far, this appears to follow the pattern of past attacks: Coupang has gained about five percentage points of market share nationally, with data we track indicating that market share has stabilized sequentially and they aren’t making material further gains.  In the short-term, this has had a small, but insignificant impact on Baemin this year. The modest share shift has slowed growth slightly, and EBITDA has faced a small headwind from targeted promotional activity to match Coupang’s discounts.  Importantly, though, Coupang Eats only competes in a fraction of Baemin’s business (first-party deliveries within Greater Seoul), so even fully matching Coupang’s discounting in perpetuity would amount to an estimated 100bps - 150bps margin headwind (% of GMV) in the Korea business.  Because Baemin can target its own discounts to the most price-sensitive customers, the actual expected impact is substantially lower.  To that point, Baemin has continued to grow EBITDA materially this year, even inclusive of these headwinds.  

Longer-term, we expect this attack to play out like previous ones: Coupang earns insufficient return on its spending (we estimate that this promotion carries a negative gross margin and will cost them $200 million annually) and that it will pull back on its discounting and that market share will revert to the pre-campaign state.  Of course, this time might be different: if, for instance, the food delivery promo can function as a lynchpin that increases WOW membership, or WOW member spending, perhaps Coupang could earn enough incremental e-commerce revenues and margins to offset the losses in food delivery.  Our analyses suggest otherwise, but it is difficult to know for sure from the outside.  

Regardless, we believe Baemin will be fine and should hit its long-term margin target in any event.  Even if it were to absorb an additional ~100bps of ongoing promotional pressure (to match Coupang), there is substantially enough latent opportunity to reach 6.5% EBITDA margins (if not higher), given Baemin’s massive scale and market share dominance.  Korea already has EBITDA margins above 3% despite AdTech revenues below 1% (of GMV) and little-to-no order stacking.  Each of those represents a 300bps - 500bps margin opportunity that should (alongside continued operating leverage) ensure Korea continues to grow EBITDA in the coming years.

More broadly, any competitive attack, including this one, may prove more durable or more far-reaching than we anticipate.  These attacks are endemic to the business, indeed to nearly every business, and require vigilant attention.  So far, the carnage in the tech markets has generally reduced access to capital and has drastically lessened global competitive intensity, relative to, say, 2021, but that détente is neither permanent nor absolute (e.g., Coupang in Korea).

• Anti-Trust.  Regulators may try to stifle long-term profits, especially since Delivery Hero’s market share exceeds 80% in a number of markets.  We believe that the Company will enjoy economically attractive margins well before regulators attempt adverse remedies.  Today’s issue is that the market is worried that the Company will not earn enough, that losses will persist indefinitely, rather than worrying that it will earn too much.

• Gig Economy.  Certain jurisdictions, especially Western Europe, and especially Spain (3% of GMV), pursue aggressive legal tactics which challenge “gig economy” business models.  Already, Delivery Hero has provisioned for lawsuits it faces in Spain, although it remains resolute that its current business practices are fully legal .  Moreover, Delivery Hero’s emerging markets tend to over-index to jurisdictions less concerned about these matters.  Ultimately, while eliminating gig work would increase prices to consumers (and therefore possibly reduce the size of the market), it is unlikely to harm Delivery Hero’s long-term margins, as the company will end up passing through the higher costs of delivery to consumers and/or restaurants.  To this point, Delivery Hero fully employs its Swedish rider fleet today, and yet Sweden is one of its highest-margin countries.

• Capital Allocation.  As we have discussed, Delivery Hero contains a collection of businesses across different markets, each with different market structures, growth rates, and current profit margins.  As such, the business requires sophisticated capital allocation, both organic and inorganic.  Senior management must determine which countries receive more (or less) resources, and what markets to acquire or divest.  After numerous conversations, we regard Niklas as a supremely disciplined capital allocator at the helm and are highly confident in our assessment that he will not run unprofitable markets or business lines indefinitely.  In other words, we are implicitly asserting that Niklas will ruthlessly achieve profitability or will jettison them (via divestiture or outright shutdown), as he has historically.  He has been savvy about acquiring high-quality and market-consolidating assets at attractive prices (for example, paying ~€2.8 billion for Baemin, which currently generates >€20 billion in annual GMV and >€600 million in annual EBITDA), and, equally important, has been disciplined about selling or exiting markets where the prospective returns were unfavorable.  Specifically, he sold the German business in 2018 at an attractive price when it became clear Delivery Hero could earn better returns elsewhere.  He exited Japan in 2021 after the investments did not earn an adequate ROI.  In short, we believe Niklas’ capital allocation talent is a great strength, but this is again a contrarian perspective.

We expend substantial intellectual energy carefully monitoring changes in these key areas.  If our thesis were to break, the cause – with near certainty – would be competition or regulatory pressure.  If such risks were to materialize, the favorable development we see in the segmented financials would falter.  To date, that has clearly not occurred, and we remain sanguine about the company’s prospects!  We love betting on the thesis that a business with 80% market share can actually earn attractive margins when the world seems to think otherwise!

Catalyst

We expect Delivery Hero to re-rate in the next 12-18 months as EBITDA margins continue to improve sequentially and the path to achieving management's long-term EBITDA/GMV margin guidance becomes increasingly clear.