#151. Taobao: Treasure hunting in China
May 30, 2025: 25 min read, long but hopefully profitable
[The Taobao app: a wormhole to a universe of alternate consumer prices]
My first role in finance was working in equity index arbitrage, buying a portfolio of stocks and simultaneously selling index futures with the same economic exposure, in order to capture the spread (usually a fraction of one percent) between these two essentially equal products. The idea of finding near identical products and trying to capture the spread between them was therefore a rather foundational part of my development, and I think I will always find it exciting: I am profiting even as I help make the world more efficient. And generally in finance, one of the basic skills is being able to determine the difference between fake, fair, and fabulous prices. My goal in this issue is to shine a light on the product arbitrages in China, and detail some of the astonishing spreads available (with a little work).
When I moved from the US to Hong Kong last year, I was still a prisoner of the Amazon.com mindset (high product convenience at reasonable prices). Hong Kong does not have a local Amazon but I discovered that I could access Amazon Japan, which offers many Japanese products not readily available in Hong Kong, so higher quality but less convenient. But recently I figured out how to use the Taobao app from Alibaba in English, and that has opened up a vast world of product arbitrage opportunities, because now I can access the wonderful and highly competitive supply chain from the Chinese mainland. From the US perspective, it’s a little like combining the pricing of Costco with the convenience of Amazon, except the price differences are an order of magnitude higher than Costco, and a much wider range of products are available.
A word about Alibaba (and Taobao, Tmall, Temu, Shein)
Taobao’s parent Alibaba is one of China’s largest listed companies and is one of the ecommerce leaders in China (albeit with significant domestic competition, unlike Amazon in the US). Taobao (which appropriately means ‘search for treasure’) is Alibaba’s C2C (consumer to consumer, but actually more like F2C [factory to consumer]) marketplace in China, AliExpress is its international C2C counterpart, Tmall is its B2C site for branded goods. The difference for Chinese consumers between Tmall and Taobao is that prices are higher because branded goods tend to be higher quality and use extensive marketing to create consumer awareness and affinity. Tmall also looks like it’s focused on selling to Chinese consumers (both domestic and international brands), rather than selling Chinese brands to a global consumer.
Temu is run by PDD, a different Chinese e-commerce giant which uses the small package de minimus exemption (where available) to allow Western consumers to access inexpensive Chinese products. From spending a little time on the Temu app, it feels like (stunningly) low prices are more central to the Temu value proposition than product quality (but which is still of secondary importance). The mindset of the Temu consumer appears to be “it’s so cheap, I can use it once and throw it away” or “if it does the job I’ll be thrilled” – rather than a place to buy core products which can be used for a while.
My cursory analysis of Temu suggests 1) a focus on clearance sales, which results in 2) less selection, and 3) lower price ranges on Temu than Taobao, implying either lower quality or lower desirability (seasonality, or style). Taobao selection seems much more extensive, with three rough price tiers: (Temu level) bargain basement (very low), Taobao standard (usually 2-3X higher), and branded goods (another 5-10X higher), with most prices in the Taobao standard tier. While both Temu and Taobao have access to many of the same Chinese suppliers, research I’ve read suggests many suppliers use Taobao as a primary sales channel, while favoring PDD/Temu for clearing inventory / overstocks, which dovetails with each distributor strategy.
Shein is a Temu competitor in the fast fashion space, which also uses the de minimus exemption channel; it is still a private company – its plans to go public this year have been scrambled by questions about China tariffs and the future of the de minimus exemption.
My working assumption is that global consumers associate “Made in China” with the Temu / Shein level, but are far less aware of the higher quality excellent value Taobao level, and have yet to fully embrace the high quality of the brands: TCL and Hisense in TVs, or BYD/Xiaomi in autos.
Same product, different price
In order for an arbitrage to work, the products have to be essentially the same, but the prices need to be different. For the consumer, a variant of this strategy is when the prices are the same, but one product is materially superior. With unbranded goods coming out of China, there is always the risk that the products are very different (e.g. poor) and thus the benefit of the product arbitrage is far less than you expect (or it is not an arbitrage at all). The other component of the arbitrage is the magnitude of the price differential, if the price spread is small, then the products have to be extremely similar (the same goes for financial arbitrage); if the price spread is very very wide (like many of the products on Taobao), then the consumer’s tolerance for product differences rises proportionally.
The recent Deepseek moment in the world of AI was, in many ways, kind of an arbitrage moment. Some pundits called it a Sputnik moment, which is marginally appropriate in the sense of ‘wake up realization’ but I think the better analogy is a Moneyball moment, because it captures the idea of solving a problem using a different analytic toolkit, and thus achieving similar results for far less money. Riffing off Marc Andreesen’s mantra that ‘software is eating the world,” I would argue that “China is Deepseeking the world” – they have already, or will shortly, develop similar / better versions of many of the most advanced technologies in the West. I’ve already gone into their lead in automobiles in issue #145.
Example #1: the aluminum pilot’s case epiphany
I’ve always wanted a Rimowa pilot’s case, partly because I like the idea of a case which opens on the top, and partly because there’s a certain romance to it, and they’re relatively rare (compared to the normal carry-on roller boards). In particular, they can carry six bottles of wine upright, with packing material, which is a far better method to carry wine than with clamshell designs. But at retail price of US$1550, I seems hard to justify, especially when I have plenty of less appropriate but serviceable pieces of luggage, in various states of entropy.
So when I found a remarkably similar item on Taobao for the princely sum of HKD500 (US$64), or a 95.8% discount versus the original — I checked the reviews and took a flyer on it. The case was delivered to me a week later, and my arbitrage sensor went into overdrive. Although it was 1/20th of the price of the ‘original’, this unbranded case was better in three ways, and ‘worse’ in one. The ‘worse’ quality was that the original (which I examined at leisure at the Rimowa store), has slightly thicker aluminum panels, which makes it ‘better’ although it also means the Rimowa case is 0.7 kg or 15% heavier (which isn’t necessarily better).
Ideally, the best thickness is the one that protects your gear from damage, while also weighing less than the armored doors of a Presidential limousine. The places where the Taobao version was better: 1) it had hinges on both sides of the lid whereas the Rimowa only had a one-sided hinge; 2) the sprung handle was soft-closing where the Rimowa was not, and 3) it came with a custom plastic + velcro cover to protect the case finish. The wheel structure, build quality, and solidity feels the same to me — but we shall see. To be fair, the Rimowa comes with a lifetime guarantee, but since it is possible to buy 20 of the Taobao versions for the price of one original, that seems like a reasonable tradeoff. Some people want to show off the Rimowa branding, but for many like me, the branding is a negative (because now it might mean ‘I paid a stupid price because I was too lazy to search on Taobao’).
A non-branded anodized aluminum pilot’s case
This particular experiment was a resounding success, because not only was the price spread kind of ridiculous, but the less expensive product was better (to my mind) than the more expensive one. This epiphany has sent my arbitrage brain passionately scouring Taobao for other areas where similar one-sided (because I’m only doing the long side of the trade) arbitrages might be achieved. I mean, what other products might you want to own, if their prices were 95% lower than your current expectations?
Knock it off with the knock-offs
The typical way to dismiss Made in China products is to profess an aversion to ‘cheap knock-offs’ — which implies an enormous quality difference, and suggests a blanket assumption that anything made in China would never be close to the same quality, much less better than, the original, at any price. Characterizing Chinese products as ‘cheap’ rather than ‘inexpensive’ or ‘economical’ or ‘great value’ is another way of projecting the same biased assumptions. In the past, there may have been ample reasons to justify this assumption, but it cannot be applied in general any more, even if they still produce knock-offs as well.
Of course ‘Made in Japan’, ‘Made in Taiwan’, and ‘Made in Korea’ all went through their transformations, based on a slow drip of empirical evidence, morphing from cheap knock-off to quality product status in roughly a generation. Perhaps the marquee product is the automobile space, where the origin of a car is typically associated with the brand of the buyer — it was probably decades before Japanese cars were viewed as equal or better to American cars (how ridiculous that seems now!), and similar issues were experienced by the Koreans until recently. I’ve mentioned before that my Genesis GV70 (‘Bubbles’), made by Hyundai, was perhaps my favorite car (other than Esmerelda). The difference with the coming transformation of the ‘Made in China’ image is that the sheer cornucopia of products that it covers will dwarf any of its ‘Made in’ predecessors. And despite the increasingly frosty Cold War between the US and China, even many Americans believe Chinese car quality is already world class.
Typically knock-offs try to create value by imitating the look and feel of the original, but usually save money with poor internal quality, limited functionality, and rough finishing. In the world of wristwatches, where imitations have proliferated since the 1970s, Rolex knockoffs are famous, both for their ubiquity as well as how sophisticated some of them have become, where it takes an experienced watch maker to tell the difference between the real and the non-original version. If the market is deep enough, super-fakes will arise, hoping to find buyers at a higher level or perhaps even get passed off as an original (here’s a video detailing an Audemars-Piguet Royal Oak high quality replica, at a 99.6% discount to the original). My preference is to buy similar items which do not try to imitate the branding / label – they should be judged on their merits of functionality and quality.
One excellent way to demonstrate better value is through similar or better functionality at a superior price, and that is what eventually changes the perception. If I try to imitate a more famous product, then implicitly I’m still admitting that the original is better (and definitely easier to sell). But once my skills are similar, it makes more sense to forego the imitations and make something similar but with clearly better functionality, at a far better price. Once I make that shift, I don’t want to pretend I’m the same brand as the original, I want credit for my innovation, I want my own brand to be on that product, to emphasize the source of the value added.
Example #2: Zerotime vs Richard Mille
One way to deal with the quality issues is to use a trusted intermediary, a specialist who can vouch for production standards. This is a fertile space, because the price differences can be so large for items of similar quality, there is plenty of margin for an intermediary, who might also provide value as the designer in addition to guaranteeing quality control.
In that vein, I will discuss exhibit #2 below, my ZeroTime T4 skeletonized tourbillon in a gold (plated) tonneau case, which is reminiscent of, but does not attempt to replicate, the ultra bling Richard Mille brand aesthetic (or budget requirements). To me, it feels like a cross between the RM 035 gold skeleton Toro, and the RM74-02 skeletonized tourbillon, both of which dance around the US$400-500,000 price point. But since design and final assembly were done in Japan, with most of the components and some assembly coming from China, the ZeroTime T4 is roughly 1/200th of the cost of its inspirational competitor. As an extra benefit, I was able to see it in person before I bought it in Japan, to verify the craftsmanship and the functionality. Of course my ‘RM homage’ timepiece is not solid gold (and actually the RM 74-02 is carbon fiber laced with gold), but it also means I don’t obsess about it, and yet I still very much appreciate its appearance and functionality.
[The Zerotime T4 skeleton tourbillon]
Although my Zerotime watch does not share the same expensive materials as the RMs, it also does not have the screw pock marks on the case, which I have an urge to scratch off; I find the slightly wider case more pleasing and balanced, and compared to both RM models, it has side/top/bottom sapphire windows to allow more light into the skeletonized movement. A number of my friends thought it might have been an RM, because it’s an attractive gold tonneau case watch with evident complications, but then they wondered whether I had lost my mind spending a half-mil on a midlife crisis bling watch. There is a purely Chinese brand called CronusArt which has less expensive tonneau tourbillons (as well as more expensive Richard Mille inspired watches) – the brand looks interesting, but it’s purely Chinese and I haven’t found anyone online to vouch for its quality yet – it’s not inspiring that most of their YouTube shorts are from “Quick Fashion Store.”
The arbitrage game: so fun, and yet so disturbing
Searching the Taobao website is a little like panning for gold in a foreign country — it requires some work to find the nuggets, but they are large and numerous enough that the effort is rather enjoyable. Results vary greatly based on search terms — if you put in “Rolex watch” you’ll get a slew of Rolex-style watches but nothing with the Rolex name — “OLEVS” comes up a lot, but the watch indeed resembles a Rolex from about 30 feet. I’ve learned it helps to search for the functions or materials you are looking for, rather than the brand, since that is my preferred dimension of evaluation in any case.
The game I’ve been playing: using the camera search function of the Taobao app to scan items (from a Japanese menswear magazine, for instance), see how closely Taobao can find a similar item (at a highly dissimilar price). So far, the price differentials for very similar items range from modest (items like branded Chinese household goods, which are already inexpensive) to very large (80-98% discounts for branded luxury goods). Some readers may think I am exaggerating the 98% number but actually I was rounding down, I found a carry-on very similar to a Globe-Trotter rollerboard which was 98.6% lower than the original (HKD376 vs HKD27700). [I’m not counting the watch example discount above of 99.6% because the difference in materials (18K gold vs gold plating) are so different.] The extremely high discounts are a function of the astronomical markups for luxury goods, in addition to the super competitive Chinese supplier market. In the case of the Globetrotter carry-on, made from layered vulcanized fiberboard (i.e. multilayered paper) using Victorian era tools, it doesn’t take much imagination to think that their experienced workforce at their factory in Hertfordshire might have far higher production costs than the state of the art in China, and Globe-Trotter has to cover their physical store network, in addition to marketing the silly narrative of ‘an elephant can stand on it’ to prospective customers. A carry-on sized piece of luggage made from treated compressed paper with a variety of polished hardware and leather straps probably shouldn’t cost the equivalent of US$3500, which is what enables the discount of 98.6%.
By comparison, my pilot’s case carried a discount of 95.8% so maybe I didn’t get as much of a bargain there as I thought (anodized aluminum probably has more of a lower limit than vulcanized paper). At these extreme numbers, however, the math gets a little wonky: the item with a 95.8% discount is three times more expensive than an item with a 98.6% discount.
Feeling the pain
Arbitrage can work in both directions, it is also possible to sell low and buy high for a loss (or a case of regret). My Taobao feed showed me an item identical to a foaming soap dispenser I had previously bought on Amazon Japan, but which was 80% less — so I bought it to confirm… and they were exactly the same. At least now my average cost is lower. With soap dispensers and many low value items, we don’t have a built-in estimate for what they actually cost, but there is a product that we see every day where relative pricing is reasonably well-known: automobiles.
I’ve been to Shenzhen three times this year, and I love to observe cars when I am there. The people who bought Range Rovers or Maseratis once thought they were pretty cool, paying top dollar for an imported car with associated duties to get the status of a foreign marque. But with Chinese automobiles surpassing Western autos in functionality (we can argue about looks), and far surpassing them in price economics, I cannot help but think that there is buyer’s regret for these show-offs, because now the narrative has become “look, I’m so wealthy I bought a car with less functionality / luxury for multiples of what I could have bought a better domestic vehicle” which is another way to say “I don’t mind wasting money” or “I don’t think Chinese autos are any good” or simply “I’m kind of an economic moron” as their shiny domestic counterparts zoom by them.
Another case in point: a new Rolls Royce Phantom in China is CNY4.5 million, or US$625K (or double that for a Chinese version with dragons). The new Huawei Maextro 800 is expected to be about CNY1-1.5 million or about ⅓ of the price, with substantially more functionality. Similar things could be said for the Maserati GT2 Stradale with 631hp at CNY4.1 million vs the (BYD) Yangwang U9 jumping hypercar with 1287hp at CNY1.68 mn, also about 60% less. Over time, the landscape in China should look more like South Korea, where domestic Genesis offerings like the G90 have taken significant market share from BMW and Mercedes in the luxury segment.
Perspective change
A little like my rabbit hole exploration into Chinese Extended-Range EVs, the Taobao epiphany feels like I know something that most of the world doesn’t know, but the scope and applicability of the Taobao knowledge are far greater than the EREVs. The videos on YouTube/TikTok purportedly showing how “luxury goods are all made in China” are hard to take at face value, I’d argue there is some version of the truth there, with exceptions on both ends. But what I don’t think is up for debate is this: for many classes of products, Chinese manufacturers can make goods of similar or better quality without trying to pass them off as original brands, for far less than the makers of branded goods. And branded Chinese goods like TCL and Hisense in the premium television market have made enormous improvements in quality.
I agree with the idea that the “Made in” labels are misleading heuristic meta-brands – there is a spectrum of quality for any country – but it takes a while for consumer perception to catch up with changes in the average quality for that country. And the average quality of Chinese goods, as a result of a highly trained workforce, economies of scale, and fierce competition, is higher than most people’s perception, which is probably formed more by Temu/Shein rather than BYD or Xiaomi.
People who think the luxury brand names are everything should by all means keep buying the original pieces at their astronomical markups, but they should be aware that luxury reputations change (remember Cadillac?). But for many people who favor quality at a reasonable price, it’s clear that the range of Chinese made products that can satisfy those criteria is absolutely enormous. But this knowledge is relatively rare, or kept secret by the retailers who are happy to keep making these astronomical mark-ups.
Brands are heuristics
The importance of brands is a response to the consumer dilemma of figuring out who to trust – over time, and via direct and indirect marketing, the brand is a heuristic which embodies a level of quality, attention to detail, and commitment to service. For luxury brands, they often imply a desirable lifestyle, or an image of impeccable taste, for instance. As brands become more popular, they can (and do) charge increasing markups for their wares, especially because in the luxury segment, the demonstration of wealth is often an important objective. But as the profit margin grows, so too does the temptation for others to make and sell low quality fakes (‘knockoffs’) and high quality imitations (‘superfakes’). Having dabbled in the entertainment business, I’m a pretty staunch defender of intellectual property (IP) rights, so both of these cases seem over the line. But the case of similar but unbranded products with equal or better functionality seems like fair competition, assuming no copyrights or patents have been violated. And it is this class of products which eventually changes consumer sentiment, especially as they establish their own paths of design and functionality, while maintaining large price differentials.
The low cost paradox
We are conditioned to believe that in general, paying more for something is more likely to result in a higher quality product. If I’m looking at two polo shirts and one is $10 and one is $30, I will assume, like most people, that the $30 shirt is better made, and the $20 extra isn’t consequential to me. But if it turns out that the “right price” – any number of manufacturers could make it – for that quality shirt is actually $10, given the cost of making it and a reasonable mark-up, then the triple-price shirt might be a worse deal, and actually takes advantage of the fact that consumers naturally assume that “more expensive is better.”
Ideally, we’d have someone rate all the options, and show price point and relative quality, so we could have some confidence that our purchase price fairly reflected the quality we were getting, but it’s still the Wild West, so it’s more hit-or-miss for the time being. That said, I’ve had nearly a 100% hit ratio in my exploratory purchases, where the worst outcome was that it wasn’t the quality of the (high priced) original item, but it still represented excellent value at the price I paid; and the best outcomes are like the pilot’s case, where the quality exceeds the original, and the price is 95% lower (and I tell all my friends).
Access: Taobao versus AliExpress
Two questions I often get are: is Taobao available in English (yes: from the home screen of the Taobao app which starts in Chinese, touch the “my account” icon in the lower right corner, then hit the gear icon at the top right, and then go to the fifth line (country/region/language/currency) and select your region and language. The other question: what are the options if they don’t have direct access to Taobao. Taobao ships to Hong Kong, Macau, Taiwan, Malaysia, Singapore, Canada, Australia, New Zealand, South Korea, Japan, Vietnam, and a variety of the ex-SSRs on good terms with China. The USA has recently been removed, for obvious reasons. The app is clearly meant to operate in Mandarin, and sometimes it won’t let you buy certain things if it’s not set back to Mandarin; overall it works well enough in English, but it’s definitely not perfect.
AliExpress is the international version of Taobao, available in English and 7 other languages, with roughly the same look and feel, but the prices are materially higher (although still at large discounts to mainstream items). The same GlobeTrotter-like case I found on Taobao for the equivalent of US$48 in HK was selling for US$254 delivered to the US, US$152 to Canada, $157 to Switzerland, $146 to Hong Kong (!), and $154 to South Korea – basically a further markup of 200% compared to Taobao to Hong Kong, and a 67% implied tariff to the USA. Even then, the 3X ‘inflated’ $152 price still represents a 95% discount compared to the $3500 original GlobeTrotter bag.
My experience has been almost universally satisfactory so far, and with the Hong Kong / Taobao prices generally far below international / AliExpress prices, I’m happy to buy/try a variety of products in the US$30-60 range without worrying about being disappointed, but I’d probably be more circumspect if the prices were $100-150. So far the most expensive item I’ve purchased on Taobao was the pilot’s case at US$64.
Grand unified pricing theory
As I collect all of these thoughts, it occurred to me that the final price to the consumer of any good might be broken down into discrete components as follows, assuming a uniform level of quality:
Globally competitive manufacturing price: for many goods, this will be the Chinese manufacturer price, both because of the advanced supply chain, and intense competition. Note that labor costs are far lower in places like Vietnam, but that often does not make a large enough difference to displace the China price. Manufacturing is not just about low labor costs, especially for higher value added products. And it’s shocking that this segment often accounts for only 1-5% of the final price.
Non-competitive price spread: making something in a country which is not on the efficient frontier for that product means there will be a price spread to account for the higher production costs. Making a T-shirt in the US so a ‘Made in the USA’ label can be slapped on it will have a higher production cost than the same item made in China where larger volumes can be produced and competition is higher.
Awareness premium: distribution costs: “make it and they will come” – i.e. producing a quality product by itself is not enough to drive sales, there needs to be some way to reach potential buyers. At the low end, this would be the operating budget for online market places, including the cost to drive clients to online marketplaces, whether this is what Temu spends on Super Bowl ads or influencer channels. In the physical world, it would be the cost of building a building and maintaining a consumer marketplace like Walmart, Target, or even an outlet mall. At the high end, the cost of a global network of physical Hermes stores (which don’t actually have Kelly or Birkin bags to sell) represents an expense which needs to be amortized through product sales.
Luxury brand markup: once the expenses above are covered, usually an appropriate profit margin is added depending on the competitive landscape. For luxury goods, which operate in a rarified economic universe where higher prices might actually drive more demand (depending on the marketing and positioning strategy) there might be great reasons to increase the markup to appear more premium. One of the more famous YouTube leather experts suggests markups of 3-5X of actual production cost seems like appropriate pricing, but many luxury brands are using markups of well over 10X.
Shipping / regulatory / tax expenses: This category now includes tariffs.
This framework allows us to theoretically dissect the reasons for the low prices on Taobao: since many of the Chinese producers are on the efficient production frontier, there is little to no production price spread, the awareness premium is mostly the net cost of running and promoting the Taobao app (one estimate is $8-12bn compared to revenue of $32-40bn annually, from the 0.5-1.0% fee they charge per transaction). Most of their ‘advertisement’ consists of user reviews on Taobao, which are free (unless they pay for fakes). There is little to no influencer cost for Taobao sellers (this is not true on the increasingly popular Douyin and Kuaishou video marketing channels), and obviously no high end marketing or distribution costs. Outside of production costs, the only other primary costs are shipping (negligible in the China/HK region), and return processing. Taobao prices are so competitive because the actual production costs are often only 1-5% of the final price, and so many of the indirect distribution costs of traditional retailers have essentially been bypassed, especially in the area of luxury goods (so that is where the largest arbitrages can be found).
By comparison, Globe-Trotter only had revenues of GBP10.5 million, and despite (or maybe because of) its extreme pricing, had an operating margin of -16.9% versus the industry average of +5.7%, suggesting that the combination of their above market subscale Hertfordshire production costs, in addition to the cost of running their marketing and physical stores, falls far short of compensating for their luxury markup, given their low volumes. Personally I’d rather get an unbranded but genuine carbon fiber carry-on from Taobao for $218 (selling at simplycarbonfiber.com for $640) than anything from Globe-Trotter.
Final thoughts
I’ve spent the last few weeks pointing out that China is no longer just making cheap knockoffs, they are also making very high quality products at eye-watering prices (usually with a little show-and-tell to get my point across). My recommendation is to try a few things out, even on AliExpress if you don’t have Taobao access, after checking the reviews and supplier ratings. Mostly, my recommendation is to have an open mind about products made in China, it’s not all just Temu trash, and in many cases products of similar or better quality can be obtained for small fractions of what a Western branded item would cost.
I suppose the new Cold War accounts for the reputation of Made in China remaining in ‘low quality mode’, a function both of slanted reportage, the high profile of Temu, and limited access to the China supply chain. But in a world of perfect information, these large consumer deals would be arbitraged away, so in the highly inefficient meantime, as the French would say, profitez-en, because there are many opportunities from which to profitez.