Source: Ruizhe Insight (ID: ryzepartners) When discussing Hema, our principle is to delve into the industry's most leading and future-representative companies. This does not mean they have already succeeded; they also have many problems. But they dare to pave the way for the industry during periods of change, insisting on doing "extremely difficult but long-term correct things." Whether practitioners, partners, or investors, it is worth deeply understanding what they are doing and why. This article will cover the following discussion points:

  1. How are Hema's supply chain and warehouse distribution?
  2. How does the membership store drive Hema's growth?
  3. What is the difference between Hema Neighborhood and community group buying?
  4. What are the pros and cons of Hema operating as an independent company?
  5. What kind of retail model does Hema want to build?
  6. How to view Hema's current valuation level?
  7. How much impact does the economic downturn have on high-quality retail? ****Deepening the supply chain is not the best path When it comes to product supply chains, especially non-standard products such as fresh produce, fruits, milk, eggs, meat, and live seafood, Yonghui is often mentioned. Recently, multiple investors have asked how Yonghui's supply chain advantages compare to Hema's. Before answering, let's quickly sort out the operational logic of domestic non-standard product supply chains. Sources are generally divided into three types: direct sourcing from origin, (surrounding) regional direct sourcing, and wholesale market procurement (Figure 1). SKUs with daily shelf life, such as leafy vegetables, are mostly purchased from wholesale markets, including by large retailers. Because of high spoilage rates, except for China Resources and Wumart, which in high store density cities or partially adopt regional direct sourcing (i.e., driving ~200 km daily after morning procurement to deliver directly to hundreds of stores), most retailers' direct sourcing ratio does not exceed 25%. Root vegetables are better, with significantly longer shelf life, and large retailers have a higher direct sourcing ratio, and better payment terms, but prices are not always cheaper than buying from wholesale markets. Fruits and other SKUs with weekly shelf life have more diversified sourcing channels. Currently, more and more hit products and traffic drivers adopt direct sourcing from origin. In fact, the overall cost to store does not decrease significantly, but you can obtain better quality goods, and after rough processing and grading, distribute them to high, middle, and low-end markets. Meat is divided into frozen and chilled, often relying on local supply chains. On one hand, regional differences are small; on the other hand, cold chain transportation costs are high. The biggest challenge is live seafood, which is divided into river fresh and seafood. The latter has stricter requirements for transportation and temporary holding. Large seafood needs a period of temporary holding after sea transport, with oxygen supplementation, otherwise it looks dead. This is why the price difference for live seafood across platforms is not much, because the supply mostly comes from professional seafood wholesale markets, and why it is difficult to eat fresh large seafood in central and western regions. Figure 1: Upstream procurement channel distribution of large retailers. Saying this is not for science popularization. Retail is detail; superficial analysis has no value. The source of fresh supply chains is extremely complex, and here we can only give a rough idea. For upstream control, Yonghui is one type of representative, and also a minority. It handles everything from source to processing center, to logistics center, and even city warehouses, including origin inspections, quality inspections, price negotiations with small farms/cooperatives, repeated communication with first-tier wholesalers, logistics provider selection, packaging material supplier selection, processing center operation management and personnel attribution, and deep processing capabilities, mostly all in-house. Hema, Walmart, etc., are another type of representative, and also the mainstream. These processes are mostly outsourced to service providers, but they strictly control the important links, sending dedicated personnel for regular inspections and evaluations. Some processing centers are self-operated, but labor relations may be outsourced. So, Hema's supply chain is indeed not as heavy as Yonghui's, but it is also the mainstream choice, and there are significant operational differences across different categories. Then, analyzing level by level from source to store (Figure 2), whether Yonghui negotiates directly or service providers do, on large-scale procurement volumes, the cost price is almost the same. Differences arise in the details from source to store, including packaging material costs, processing fees, logistics and distribution fees, loading and shipping methods, etc. For example, for cherries of the same origin and grade, the source procurement price is 20 yuan/jin. If these processes are given to service providers, the price to store, i.e., cost plus ~15-16% (~24 yuan/jin), while Yonghui's own cost is 7-8% (i.e., 21.5 yuan/jin), with spoilage calculated separately, usually a discount on the procurement amount based on the situation after store inventory. This asset-light operation is, in our view, more beneficial than harmful for Hema. The core of the supply chain is product selection and quality control. As long as the source price is relatively market-oriented and stable, whether it is necessary to personally invest in and deeply operate the entire chain, saving every penny, this approach of deepening the supply chain may have advantages, but it may also carry higher risks when supply and demand are unbalanced or when facing operational difficulties. It is said that Yonghui's supply chain is strong, but recently some quality issues have still appeared. Figure 2: Supply chain and warehouse distribution model from source to store. One additional point: there is no universal buyout in the industry. Buyouts require large amounts of cash, which does not conform to the retail logic of deferred payment. However, buyouts of a few SKUs based on strategic layout are feasible. Whether it is Hema, Yonghui, or regional leaders like Jiajiayue, no one can buy out China's fresh supply chain; that would require multiple times the cash flow. To enhance product supply chain capabilities, especially for non-standard products, deepening the supply chain is not the optimal path. The future competition in the supply chain will ultimately return to being people-centric and data-driven. Nielsen, the world's number one consumer research agency, can tell you which standard products sell best within a 1-kilometer radius in large cities and what the average retail price is. That is not enough. In the future hypermarket, you also need to know the daily consumption structure of all users within a 1-kilometer radius, especially your members: what they often buy, what they have been looking at recently, and what price range they are in? Today, Hema has this capability because it relies on the user profiles of ~800 million DAU from Taobao, and in terms of ARPU for comprehensive e-commerce, Tmall is second to none. The granularity and volume of data determine the effectiveness of profile-based marketing. There are no fewer than 50 retailers with more than 5 million registered members, but no one in the company knows what members are looking at or like. They can only let procurement drive the product side, and if goods don't sell, they blame each other. This is the norm for most retail giants today. Traditional retailers also face many baggage issues, such as having elderly aunties and uncles as core members, which is not the best user structure for retail, and internal procurement-led innovation deficiency. ****X Membership Store has first-mover advantage, but ultimately it's about product strength Speaking of membership stores, interestingly, both the world's largest Costco and Germany's largest Aldi are jokingly called "poor people's supermarkets." This is, of course, consumer banter. Costco covers almost the vast majority of American households, with over 100 million members globally (~45% are supplementary cards, ~62 million active paying members) and ~90% membership renewal rate. Costco went public decades ago, and there are many articles analyzing its model, so we won't expand here. Let's briefly sort out the characteristics of its historical development and the core indicators of the three major U.S. membership stores (Figure 3). Costco's predecessor was Price Club, founded in 1976, which, as the name implies, was a "membership discount store." The social e-commerce, community group buying, and reverse customization that have risen in China in recent years are all concepts others have already played with. Price Club went public in 1980, and in 1983, the first store after Costco merged with Price Club opened. In 1985, it entered overseas markets, and it wasn't until 20 years later, in 1995, that it established its own brand Kirkland. By 2000, Costco's annual revenue and membership were $31.6 billion and 15 million, with SKUs ~3-4K. In the nearly 10 years after the financial crisis, Costco's CAGR has been ~3 times that of Walmart, reaching ~$192 billion in global revenue in fiscal 2021, about 40% of Walmart's scale. Figure 4 shows the differences between Costco and traditional hypermarkets in core operational indicators. Combining these, the key points are:
  • Started with C2M model, relying on membership stickiness to penetrate the "low-price, selected" positioning
  • A global quality product supply chain system honed over nearly 50 years
  • Costco is also the number one auto retailer in the U.S., and other number ones
  • Over the past 20 years, global CAGR >9%, outperforming U.S. GDP CAGR ~3.3%
  • Single-store revenue ~1.2 billion RMB/year, member per capita contribution ~20,000 RMB/year
  • Profit ~65-70% comes from membership fees, gross margin only ~13%, compressing all operating costs Figure 3: Analysis of indicators for the three major U.S. warehouse membership stores. Figure 4: Analysis of operational indicators: Costco VS traditional hypermarkets. Costco's rise is driven by the collective procurement demand of ordinary Americans for cheap but quality high-frequency consumer goods, with household coverage of over 90% and renewal rate of ~90%. This is completely different from the customer base and demands of membership stores in China. Whether at Sam's or Hema X Membership Store, unit prices are not lower than brand products in hypermarkets, let alone compared with brand goods on domestic discount stores, community group buying, or Pinduoduo. Currently, the million-level membership of various stores is limited to first-tier cities, accounting for ~2‰ of the national population. In the end, it's a gap in per capita economic contribution and purchasing power. U.S. per capita GDP is still ~6 times that of China, and Washington D.C.'s per capita GDP is ~$200,000, ~8.3 times that of Shanghai. What is "low-price selected" for Americans is merely "affordable quality" for China's affluent population. But the popularity and rigid demand of membership stores among the top 5-10% income families in China is undeniable. Sam's Nanshan store has been the global number one for 20 consecutive years, with single-store annual revenue of ~2 billion yuan, and Hema X Membership Store's first-day revenue was ~6 million yuan. Many of my friends and relatives in Shanghai are families with annual disposable income of 150,000-250,000 yuan, and almost all have Sam's memberships, with average annual household spending of 6,000-8,000 yuan. That is, they are willing to spend ~4-5% of their annual disposable income to improve their daily living standards. The collective procurement demand for better quality, novel, and consistently stable quality goods after people have a certain purchasing power is universal and irreversible worldwide. So Hema's entry into warehouse membership stores is a rigid demand, incremental market worth looking forward to. How big can the X Membership Store become? Let's first look at some household disposable income data from MBB (McKinsey, Bain, BCG). Figure 5 shows data from one MBB combined with Ruizhe's forecast. By 2020, there were ~100 million households with annual disposable income above 200,000 yuan. Another MBB's white paper on membership stores estimated the potential number of households at ~70-80 million, also with the standard of annual household income above 200,000 yuan. If we conservatively estimate a CAGR of 15% for this population over the next 5 years (10-18 CAGR ~30%), and ~10% penetration rate, the market size by 2025 would be ~112 billion yuan (160 million * 10% * 0.7 million yuan). Another calculation: currently, the domestic market size, based on Sam's, Costco, Metro, Hema X stores, and others, had a combined revenue of ~45 billion yuan in 2020. The National Bureau of Statistics gave an unreliable forecast of ~15%, but the actual should be higher. At a CAGR of 15%, by 2025 it would be ~90 billion yuan, slightly lower than the population-based estimate. Hema currently has only 6-7 X membership stores, most newly opened within about half a year. Based on single-store revenue of ~500 million yuan, benchmarking Sam's, in 3 years, if Hema captures 1/4-1/5 of the market share, i.e., 18-22 billion yuan, a ~6-fold growth space during this period is expected. Figure 5: Analysis of operational indicators: Costco VS traditional hypermarkets. Hema's strongest competitors are undoubtedly Sam's, Costco, and Metro. Others like Carrefour Club and Aldi are not yet a threat. In terms of global supply chain, Hema must humbly learn from these three. The global product capabilities built over decades cannot be broken in a few years, but following the same old path as these three will never catch up. After all, there are differences in Chinese and foreign consumer tastes, and oversized packaging may not be universally suitable for Chinese families. Moreover, young Chinese families may be the pickiest consumers globally. In the short term, the three giants' flagship products are indeed stronger than Hema's, but the long-term decisive point cannot be just a few dozen SKUs to satisfy Chinese consumers. This is Hema's long-term opportunity, not to mention the 800 million MAU user base behind Hema. Overall, facing this wave of incremental demand of at least ~150 billion yuan in 10 years, Hema X Membership Store has a very good first-mover position. Currently, there is basically no competitor in the domestic retail industry. What needs to be considered is how to have a sufficiently differentiated competitive strategy against global giants, with product strength always being the foundation to continuously consolidate. ****Neighborhood is an important supplement to large stores, and will not compete with community group buying Let's first talk about community group buying. We have thoroughly studied this model. If it weren't for the crazy price wars, it would have been a very good retail channel model innovation. All operating costs after arriving at the store are saved: the last 500 meters delivery cost, front-end marketing, backend customer service and return/exchange costs, store rent and utilities, etc., all are handed over to the group leader, who gets ~10% commission. According to various public data estimates, the transaction volume of community group buying in 2021 was ~120 billion yuan, but after accounting for returns, it is definitely less than 100 billion. The financing amount in this track is no less than 10 billion USD. A good new model, with transaction volume entirely burned by financing, making it difficult to improve service and quality or raise prices. Neighborhood (Figure 6) ultimately wants to do what, Hema should also be exploring, but our view is that it will not follow the old path of community group buying. The population of community group buying is extremely price-sensitive. I once participated in a community group buying platform in Beijing with a daily order volume of about 50,000. Over 70% of users were outside the 5th ring. In summer, if the individual group leader model doesn't have a large refrigerator, the already low-grade leafy vegetables would be wilted by evening, and there were frequent stockouts. If one SKU is missing, a dish can't be cooked. For a family in a third-tier urban area, if they use community group buying for groceries every day, they could save about 1,000 yuan per month in living expenses, but this money is exchanged for generally poor quality, unstable quality, basically no service, and occasional problems. Even so, this is still a rigid demand model under the purchasing power of the sinking market, but it is not a retail format suitable for Hema. Figure 6: Hema Neighborhood model, storefront photo. For first-tier suburbs and sinking markets, Hema's advantage lies in doing dimension-reduction output, saving ~8% of home delivery fulfillment costs, plus product strategy adjustments and neighborhood store operation strategies. Hema's large store quality and guaranteed products can be offered at ~85% of the price in Neighborhood, some at 70%, but you need to order a day in advance and pick up downstairs yourself. Today, even in third-tier cities, there are still ~10% of middle and high-income families. Who will meet their demand for high-quality, reasonably priced goods? In recent years, in first-tier cities, those who used to be regulars at hypermarkets and wet markets have turned to Hema. In third-tier cities, Hema products are also 20% off. This is the market and population that Neighborhood should target, completely different from community group buying. Moreover, in the short term, Hema does not allow Neighborhood franchisees to run small grocery stores, fully serving Hema members. Neighborhood has also opened courier stations. In the future, there may be opportunities to divert the demand for storing and picking up billions of packages annually from the Cainiao system to Neighborhood, driving offline traffic and saving marketing costs. This long-term imagination space is huge. From the UE model perspective, we interviewed several Neighborhood store owners. With monthly rent of 7,000-8,000 yuan, daily orders of ~120, and average order value >50 yuan, they can be profitable. Compared with community group buying, Neighborhood's biggest advantage is the purchasing power of its customer base and the corresponding average order value. For community group buying, including previous community fresh food stores like Yipin Fresh, the customer positioning and demand characteristics determine the ceiling of average order value. This cannot be solved by simply introducing some 3C products. Selling 3C and other high-ticket items in community group buying is a mismatch of scenarios. Secondly, can prices beat Pinduoduo? Can service beat JD.com? Hema's potential customer base has a much higher price tolerance for fresh produce and daily necessities than community group buying customers. Opening Neighborhood in first- and second-tier cities with sufficient density is an important supplement to Hema's large stores, but sinking to tail-end second-tier or third-tier cities is indeed difficult. Whether it's store-warehouse integration, front warehouses, or small stores, sufficient order density is needed to support supply chain and logistics warehousing costs. We interviewed several senior supply chain and logistics experts in retail. If the daily procurement amount in a single second-tier city is less than 10 million yuan, it is difficult to support the reasonable operating cost of a logistics center, equivalent to opening 30 traditional retail large stores or 150 front warehouses. This is why front warehouses face huge challenges in second-tier cities: high customer acquisition costs, lower average order value, lower retention than first-tier users, and second-tier users are more price-sensitive. Although the two cannot be directly compared, using a single warehouse model, if daily orders per warehouse are less than 1,000, average order value is less than 90 yuan, or density is insufficient, any one of these conditions is a huge challenge. Hema has had its ups and downs and suffered a lot along the way, but if there is some adjustment and optimization this year, front-end profitability is certain, while front warehouses still see no end. ****Internal operations management is Hema's challenge, but independence may not be a bad thing Hema is a startup, and an internet-gened company that loves to tinker. Over the past few years, it has continuously tried and erred with more than 10 new models. It's really very Alibaba. Many senior friends at Alibaba have told me this is called rapid trial and error, small-step iteration, with new projects changing every six months. We believe that as long as it is within financial tolerance, innovation should always be encouraged. Once funds are not abundant, temporarily closing these trial-and-error projects can significantly improve losses, as Yonghui has done. Furthermore, half of Hema's staff come from traditional retail, with a rhythm of daily meticulous work, while the other half are typical Alibaba people. When people from two worlds come together to change traditional retail, and are also constrained by the group's rules and regulations, the only way to avoid mental exhaustion is to muddle along. Now that front-end profitability is imminent, independence may not be a bad thing. After calculating, the mid-to-back office costs are about a loss of ~7-8%. Temporarily closing unnecessary trial-and-error projects, adjusting group-related operational personnel, and sending large-scale module developers back to Alibaba temporarily. A Hema P8 developer costs as much as 4-5 senior headquarters employees in traditional retail. Additionally, there are some unnecessary administrative expenses. With such simple adjustments, saving 3-4% of mid-to-back office costs is not a problem. Closing stores is also an effective strategy to recover losses and the most reasonable adjustment in retail. The figure below (Figure 7) shows the store closures of typical leading hypermarkets in 2021. Hema's closure rate is far lower than several peers. It may be that it is more famous and easily attracts attention. Figure 7: Store closure rates of major leading retail enterprises in 2021 ****Hema's losses are valuable, and the underlying logic is becoming clearer Traditional retail hypermarkets have reached a period of historical transformation. In the next 10 years, not changing is waiting to die, and changing recklessly is sudden death. The conservative strategy is to continuously shrink scale. But in fact, demand is still there; only the level of demand, channels, and preferences have changed. On the road to exploring new models, those who dare to innovate are often questioned. When Bianlifeng first came out, traditional convenience stores were almost all pessimistic about it. Now? It has opened more than 2,000 stores. Although it has also stopped and started, it has not raised funds in the past two years during its expansion. If it weren't continuously improving, how could it afford the losses? In-depth research on leading innovative companies does not mean they have already succeeded, but that on the path of trying, these companies have run the farthest, accumulated the deepest, and encountered the most pitfalls. This is the value of our analysis of Hema. Losses don't matter; the key is whether it can be profitable in the long run and whether it is viable. Amazon also lost money for N years before becoming fully profitable, and now has a market value of $1.4 trillion. Of course, in times of cyclical change, there are always lagging enterprises that are replaced. It's not that competitors deliberately try to kill you, but that customers/users gradually abandon you. This is cruel, but for innovators, this is also the reward under the high risk of innovation. Hema has lost money for these years, but its supply chain is becoming more mature, front and back office losses are narrowing, and it has stepped on many pitfalls, ultimately converging into the entire operating system shown in Figure 8 below. Figure 8: Analysis of Hema's omni-channel operation model, 2022.02. The operating system shown in this figure has undergone several changes in the past 5 years, such as Hema Cai Shi in the format matrix, Hema Xiao Ma in channel complementarity with RT-Mart, local fresh food brands in product layout, and Hema Gongfang in specialized formats. If one day Hema reaches a market value of 100 billion USD, it will not be an easy success. How could it grow without stepping on these various pitfalls? The future of retail must be omni-channel operation. Those without data capabilities will not keep up with changes, those with lagging digital technology will suffer, product operation capability is the basic skill, but the most fundamental competitiveness is user operation. These extremely difficult but correct things ensure that Hema, as of today, can maintain a long-term leading advantage. For example, how to grade a batch of goods to squeeze out profits is becoming clearer in terms of formats; how to form a differentiated advantage in the supply chain is certainly not by repeating the old path; How to leverage the powerful force of digitalization? RexOS has invested billions over the years, and it will naturally accumulate and burst forth. Hema Bakery can also be considered a chain bakery brand in the top 5 nationally. Based on Ganso's market value of 7-8 billion USD, it still has independent brand value, unlike some new bakery brands with a single-store valuation of 100 million (ridiculous). Also, the popular Hema Flowers has opened independent stores; Not to mention that once Hema gains a firm foothold within Alibaba, the group's resources will be fully empowered. With 800 million MAU users and 8 trillion yuan/year in consumption data, how can this value be compared? In fact, Hema is quite harsh on suppliers, but they are still willing to bind with Hema. Why? Suppliers of all sizes have their own judgments about the retail situation in the next 10 years. ****How much is Hema worth? Money for long-term value is suitable for investing in Hema From the secondary market perspective, Chinese concept stocks are completely disconnected from fundamentals, so we won't reference them for now. We have compiled representative retail enterprises at home and abroad (Figure 9). Domestic enterprises generally suffered losses in 2021, so we adopt 2020 data. Roughly, the average PE is 20-50 times, and average revenue growth is 10%-20%. Setting aside Hema's future CAGR for now, if Hema can achieve full profitability within 2-3 years, with pre-tax profit of ~3-5%, and revenue of ~50 billion yuan by 2023, even using these traditional retail targets in the secondary market for evaluation, a valuation of ~10 billion USD is reasonable. Last year, many new consumer brands that have not yet gone public have seen their primary market valuations halved, let alone evaluated at secondary market fair value. On this point, Hema really hasn't overpriced. Whether it can be profitable and when, both the previous article and this one have analyzed. The fundamentals of the retail model are viable, and profitability is not difficult, with many methods. Secondly, in the nearly 1 month since the previous article was published, I have talked to many investors about Hema. The general view is that apart from the high valuation, there is nothing wrong. In contrast, the 10 billion USD in community group buying has gone down the drain, and investments in front warehouses seem to be failing. They still owe suppliers money, and cash flow is insufficient. Fortunately, a bank recently extended a line of credit. Interestingly, the bank issued a statement saying it provided the credit to support anti-epidemic efforts. Read into that. Of course, these entrepreneurs are admirable, but from an investment perspective, Hema may be the only reliable consumer retail target that will be fully profitable and has a strong backer. Figure 9: P/E ratios and growth rates of representative retail enterprises at home and abroad, 2021. Let's talk about Hema's expected CAGR. It would be a stretch to say 40-50% growth in the next 3 years, but 20-30% is expected. In the longer term, once front warehouses can't hold on, that market share will be 100% given up, and traditional retail will slowly give up good locations. This is directly beneficial to Hema. From a 5-8 year perspective, if Hema can both be profitable and maintain around 20% growth, that's actually good. It's just that different funds have different strategies, styles, and demands. This kind of target, which is about to be profitable in the short term, with annualized growth of ~25%, no significant premium compared to the secondary market, has the potential to become a new retail giant in the long run, has a strong backer, and is highly likely to exit, is quite suitable for funds seeking long-term stability. If you want to make big money in 3-5 years, with a 5-10x return mentality, then Hema is not a good target. That's all about valuation. As Alibaba's core strategic layout, with all trends improving, the organization is still Hema's strongest backing. Alibaba, no matter how difficult it is now, is a giant with 100 billion in profits, with 440 billion in cash and equivalents on its books, and China's strongest technology middle platform. Alibaba Cloud is far ahead. Having done a hard business for over 20 years (compared to Tencent and ByteDance), it has instead built the most solid business foundation. In this regard, Tencent is not comparable. Hema's intention to come out this time is just that the timing and conditions are mature. Independent operation is good for both the organization and the business division. With such system support, they are not really that eager for money. If everything goes as analyzed, even if it takes 1-2 years longer, investing then will not be at this price. I can't analyze further, or Ruizhe might as well become a financial advisor! ****Overcoming difficulties together: either create value or discover value As mentioned in the previous article "How Today's Venture Capitalists and Entrepreneurs Cross Cycles," China's miraculous 30 years of economic development has now entered a new cycle. This cycle is indeed the norm in history, and all venture capitalists and entrepreneurs need some time to adapt. How much impact does the economic downturn have on high-quality retail? Saying it has no impact is false, but saying it has a big impact is also an exaggeration. Hema is not the kind of premium independent supermarket that sells at ridiculously high prices. The unit price of PB products is even cheaper than brand products. Hema's target customer base has some tolerance, so there won't be much impact, including on X Membership Store. After all, daily household expenses are at most 10% of middle-class income, and once consumption habits are formed, they are hard to change. When the economy is bad, for them, eating out less at good restaurants a few times makes up for it. What really suffers is the business of those good restaurants. In the face of the epidemic, domestic retail giants have shown responsibility. Whether it's Hema, Wumart, or Yonghui, we have no disrespect for Yonghui. To be honest, Yonghui's innovative attempts in recent years are particularly admirable. This is the responsibility of leading enterprises. The leader should act like a leader and pave the way for the industry. Those who say they were led into a pit by new retail are irresponsible. It's easy to criticize others. Some media are truly irresponsible. There are few articles worth reading; they either expose trivial matters or are incoherent. Either create value or discover value; this is the proper attitude to overcome difficulties together. ** -END-