Towards A Centralized B2B Trading Platform

For those of you who don’t want to read this whole post, here is a 9-point summary:

1. Technology has succeeded in disintermediating markets for books, film, and music. In markets for art, watches, jewelry and other collectibles, intermediation remains a source of value for consumers, especially for higher-end product. B2c intermediation (preferably branded) is necessary to validate aesthetic value, provenance, mechanical guarantees and authenticity.

2. Secondary markets are characterized by product dispersion, high fragmentation and information asymmetry. Traditionally, independent b2b traders gathered product from their local markets to supply b2c intermediaries in offline settings. Although this continues today, a growing portion of trade is conducted using technologies primarily designed for consumers.

3. Technology has exacerbated fragmentation, leading to islands of markets that stretch out horizontally, characterized by different fee structures and an evolving composition of participants with different trading profiles. Auctions, Chrono24, Artsy and chatgroups are all examples of islands. It is too difficult to track supply and demand (and hence price signals) coherently across these islands.

4. The price signals from each island conflict due to A. the heterogenous nature of the products and B. the diversified trading profiles of the participants in each island. For example, differences in pace of buying can lead to asynchronicity of wholesale and retail price and defective price discovery. Although A. is difficult to solve, B. can be mitigated by optimizing market structure.

5. An optimal market structure would consist of 2 tiers - an upper tier where b2c dealers specialize in intermediation to consumers and sourcing from a global range of vetted, professional dealers, and a lower tier where b2b dealers specialize in sourcing for sale to b2c dealers¹. This structure currently exists informally, but lacks infrastructure to support it.

6. These 2 tiers should be connected via a low cost, low tech, no frills, centralized trading platform with the broadest participation of b2b and b2c professionals. Or in other words, the biggest professional trading chatgroup ever.

7. This centralized trading platform would improve supply/demand tracking, price discovery, capital and labor efficiency, ameliorate supply constraints, reduce b2c platform dependence, and importantly, be the sourcing ground for authorized dealers participating in manufacturer-affiliated authentication programs such as the Rolex CPO, who are poised to become the most important b2c intermediaries² in the future. Centralizing global supply in one place could reduce price volatility by lowering labor capital and improving turnover probability.

8. Although human intermediation will remain important for the forseeable future, co-intermediation with AI may become reality sooner than expected. For example, instead of rigid menus and filters, consumers may instead discover their interests via a conversation with ChatGPT. All b2c platforms will be forced to adapt and evolve. However, a centralized b2b trading platform (database) will remain the foundational layer upon which future forms of AI-b2c human dealer co-intermediation will be based.

9. In the $23Bn pre-owned watches trade, Worldwatch.market is a dealer-founded startup that I think is the first to build such a centralized b2b platform. Contact Stan Duin (info@vintagemasters.eu or sales@worldwatch.market) if you are a b2b or b2c dealer. If you tell him I referred you, I get a discount, so thanks in advance.

   ---

(Please click on “View Entire Message” if reading from email)

In my last (serious) post I argued that over-reliance on retail and auction data leads to inefficient and defective price discovery, due to the asynchronicity of wholesale and retail prices, in addition to the more complex, heterogeneous nature of finished goods. In the absence of robust pricing models, I proposed an alternative, imperfect formula for understanding the relationship between wholesale and retail prices, but concluded that a centralized wholesale trading platform was necessary to fundamentally improve price discovery and ultimately, market efficiency. This seems to be applicable to all secondary markets with similar market dynamics, and will become increasingly important as a growing share of consumption shifts to secondary (pre-owned, grey, etc) product, whether art, vintage clothes, furniture, jewelry, watches, etc. It is especially applicable to markets where disintermediation³ has failed, and a 2-tier market structure is necessary.

In this post I write about how we got to where we are, a best practice from the real estate market, the key characteristics and benefits of a 2-tier market system supported by a centralized wholesale trading platform, how AI may change future intermediation, and introduce a rare startup that is taking on this challenge.

Where We Are and How We Got Here

Secondary markets for heterogeneous products such as art, watches, jewelry, etc have not changed much structurally since their inception. Their unregulated nature has allowed the full brunt of technology to shorten distances and improve trust and payment options, but the fundamental structure is the same - whereas primary markets are a function of inputs, costs, margins, long term contracts with ADs and marketing, secondary markets are characterized by fragmentation, information asymmetry, heterogenous product and a “deal by deal” process that generally attempts to match a unique item to a unique end-consumer.⁴

The start of this process generally begins with the primary market retail end-consumer or authorized dealer relinquishing their item(s) to a wholesaler (hereon “b2b”) at some kind of discount to the secondary market retail price, and traded until it reaches its maximal price destination, which usually involves a retailer (hereon “b2c”) at the tail end. The problem as I discussed in my previous post is the unreliability of secondary retail prices, which leads to inefficient and sometimes defective price discovery.

And to make matters more complicated, the retail price is significantly influenced by the b2c retailer’s brand/intermediation, which itself confers aesthetic, provenance, mechanical guarantee and authenticity value on the item. Human intermediation adds rather than extracts value, in contrast to markets for books, film and music. Platforms such as Chrono24 or Artsy attempt to replicate Amazon for their respective markets, but are hampered by consumers’ need for human intermediation, especially for higher-value items. It is a fundamental flaw in their business models⁵. Direct sales between end-consumers exist, but they are the exception, not the norm, for the same reasons.

Traditionally, b2b enables b2c - there is a need for brand-less b2b (aka “wholesale” or “grey”) dealers, many of whom are participants in offline tradeshows such as the IWJG/WTA. B2b dealers operate in broad product at lower margins with quick turnover, supplying b2c dealers and auction houses that need inventory. Because there are no long term supply contracts between b2b and their b2c customers, pricing is generally negotiated in real time on a deal-by-deal basis. Self-sufficient, b2c retailers that source directly from consumers exist, but the fact that b2b dealers and tradeshows thrive is evidence of their value-add in the market; B2b specialize in sourcing and monitoring supply/demand, while b2c specialize in branding, intermediation and retail operations.

Although I cannot speak for all of us, I would estimate that the majority of b2b sourcing/trading is still conducted offline, whether via trade shows or in physical office/retail locations. However, a substantial and growing portion is conducted online, using technologies that were primarily designed for consumers or b2c in mind.

These b2c/consumer technologies have enabled many more islands of markets than in the past. Auctions, tradeshows, pop-ups, chatgroups, marketplaces and platforms such as Chrono24 are all islands that technology has allowed to stretch out horizontally across online and offline space. Unfortunately, it is too difficult to track supply and turnover across islands. While there are some obvious hierarchies of islands, generally they stretch out horizontally, preventing supply and demand tracking and leading to defective price discovery. The price signals emanating from each island conflict, primarily because the compositions of the participants is constantly evolving⁶ and to a lesser extent, because each island has its own cost/fee structure. Each participant has a different trading profile (for example differences in the pace of buying as I wrote in my previous post), obfuscating the price signals.

Imagine a treasure hunt where you hop around islands a, b and c while another person hops around c, d and e. Both of you will not have enough time or stamina to hop to islands f and g, where others are hopping around. There are lots of different people (retailers, consumers, b2b dealers, semi-professional dealers, etc.) hopping and doing the same thing, leading to significantly higher labor capital and lots of confusion and anecdotes regarding price, complicated by different fees and heterogeneity of the products themselves. Some data is shared, others totally different. The prices from other islands may also be at odds with the prices in your islands.

Categorizing islands based on speed and safety is a quick way to evaluate their fit with a participant’s trading profile. For example, auctions can be categorized as slower/risky, while b2c dealer websites and marketplaces such as Chrono24 (with escrow services) are fast/safe. Ironically, one of the most popular types of islands today is the fast/risky chatgroup⁷, where diverse participants trade quickly based on photos. Many consumers who for various reasons are unable to be served by b2c dealers participate in these chatgroups, changing the nature of the price signals.

Unfortunately, the fast-paced, upward scrolling and siloed nature of the chatgroup makes it difficult to gain an accurate pulse on broad market supply and turnover. And the endless 24-hour notifications of new offers demands significant labor to filter through, degrading the quality of life of participants⁸.

Why did we end up like this?

We Deserve the Systems We Have

Here is an illuminating quote by

Arnold Kling

, from this post

I could not help but notice the correlation: tight organizations had well-engineered computer systems; loose organizations had messy data. It struck me that

Organizations get the computer systems they deserve.

When you have an organization that is improvisational and fragmented, its computer support will not be well-engineered. Moreover, top management cannot simply order up a well-engineered computer system for a loose organization. You cannot use information technology to organize the disorganized. The software engineers cannot drive the business from the back seat. That was a difficult lesson for Freddie Mac’s senior management to absorb.

Instead, if you desire a well-engineered computer system, you have to tighten up the business process. Managers have to be able to articulate business rules, write them down, and stick to them.

I think this also applies to whole markets like ours. Participants need to be rules/standards based and somewhat altruistic to develop a common protocol. The technologies that participants in our markets use has exacerbated the inherent fragmentation and resulted in islands of disconnected markets such as the chatgroup.

Now let’s look at a market that has overcome fragmentation by successfully implementing a centralized trading platform.

Best Practice from US Real Estate: The MLS

In the US real estate market, most selling agents are required to list their properties on the multiple listing service (MLS). The MLS is a low cost, low tech, centralized repository of listings which benefits buyers/consumers by decreasing information asymmetry in what would otherwise be an opaque and fragmented market. It provides equal sourcing opportunity and improved price coherence by taking power away equally from the selling agents, who in years past could extract more concessions by controlling pocket listings and decide who to sell what to. The MLS does not extract sales commission fees, but instead charges a predictable, monthly membership fee to ensure broad participation. B2c real estate platforms such as Zillow and Redfin are built on top of the MLS, and have to innovate to get buyers, because the source MLS data layer is all the same⁹. Consumers therefore only need subscribe to one b2c platform, not different islands of platforms. Importantly, even a novice buyer is well protected by specialized ancillary services by 3rd parties such as escrow, title and engineering due diligence, in addition to standardized contracts. These services reduce the overall risk premia associated with buying real estate.

In the unregulated secondary market for watches, art, jewelry and other collectibles, the majority of sellers sell their own inventory as owner, without licensed agents. Hence they (rightfully) retain all rights to the sale, including where to list and to whom their inventory will be sold. Whereas licensed real estate agents could be required to list on the MLS, naturally there is no such obligation for sellers of art, watches and jewelry. Sellers are incentivized to maximize margins independently.

Cohesion requires some level of standardization and altruistic relinquishing of independence from its members, but this is difficult to coordinate due to the globally fragmented nature of our market. Hence why we have what we have now - islands of horizontal markets with an evolving mix of different trading profiles, resulting in defective price discovery and inefficient investment and labor capital.

A 2 Tier System Supported by A Centralized B2B Trading Platform

Whereas markets for books, film and music have succeeded in disintermediation, buyers of higher priced art, watches, jewelry etc still value intermediation due to the need for validation of aesthetic value, provenance, mechanical guarantee and authenticity. Consumers look to branded b2c intermediaries to provide this value. Dealer websites/social media, dealer-sponsored platforms such as Pushers.io, auction houses both traditional (Christie’s) and new (LoupeThis, etc), and even marketplaces such as Chrono24 are all examples of intermediaries providing all or some parts of this important value¹⁰, in return for a margin or a percentage on final value. And there are many new types of intermediaries emerging, mostly marketing themselves with keywords such as “collector-first”, “community”, “trust”, etc. Ironically, the most powerful of the new intermediators are the traditional brick and mortar authorized dealers (AD) participating in the recently introduced manufacturer-affiliated authentication programs such as the Rolex CPO¹¹.

All the above intermediaries make up the upper tier of the market.

The lower tier is composed of participants that spend most of their time sourcing and trading among themselves and/or supplying b2c intermediaries in the upper tier. It is important to note that the roles are fluid and not mutually exclusive - many b2c retailers source from consumers and/or liquidate to b2b dealers, and some b2b dealers have consumer clients - but they will generally fit into one or the other tier.

Today, we informally have such a structure, but there is no infrastructure that enables it, only islands of markets that are generally agnostic to trading profiles. What is necessary is the development of a common platform on a global scale. I envision such a platform to have the following 10 characteristics, highly influenced by the MLS:

1. Widest global participation of b2b and b2c dealers with strong participation from the high volume US market

2. Low and predictable monthly fees (no sales commissions)

3. Some integration with inventory management

4. No ancillary services (customer service, escrow, etc a paid or outsourced option)

5. No frills (expensive market analysis a paid or outsourced option)

6. No SEO, no branding, no marketing

7. Some listing standardization (photos, descriptions, etc) required

8. Members of tradeshow groups and guilds are required to list

9. Consumers can request their consignments to be listed

10. Authorized dealers purchasing for manufacturer-affiliated authentication programs such as the CPO require that the their potential acquisitions be listed

As a trading platform for professionals, there would be no costly handholding, with trust established outside the platform.¹²

Being required to list may infringe on seller’s rights, but the more rules-based we can get, the more we improve the overall trading process, supply/demand tracking, price discovery and ultimately the efficiency of the market. With the world’s supply all in one centralized location, pricing will better reflect intermediation value, rather than a perceived local supply constraint. Turnover probability will improve, as b2c dealers will actually be able to take and fulfill orders from consumers in a reasonable amount of time, something that is difficult in any secondary market. I think the overall effect of this would be to reduce sudden retail price volatility.

B2c dealers will find that they can spend less time on sourcing and more time on intermediation, something that consumers/collectors value. There would still be pricing errors due to heterogeneity (wear, patina, etc) but the supply/demand tracking errors attributable to islands would decrease, and bid-asks would narrow.

The platform would also be a key sourcing ground for authorized dealers participating in manufacturer-affiliated authentication programs such as the Rolex CPO. In fact, AD’s should require that all offers are listed to lower their sourcing labor capital!

B2c dealers would be able to source and liquidate inventory in a closed global b2b system, improving their capital efficiency without hurting their local reputation.

Although not directly related, developing a rules-based culture could also have positive second order effects on overall business ethics, dis-incentivising consignments fraud and trade of stolen goods.

Like the relationship between Zillow/Redfin and the MLS, b2c intermediaries such as Chrono24 and others would effectively be tethered to a foundational, centralized b2b trading platform. Obviously many b2c dealers will source directly from consumers, but the majority will be from the platform. It would also level the power balance with b2c platforms such as Chrono24 by demonstrating tangible b2b dealer unity.

AI/B2C Co-Intermediation?

If future forms of AI could deliver intermediation value, then the arts, watches and jewelry markets may succeed in human disintermediation. Unfortunately this is not the case today. Perhaps it will be possible in the nearer future for 2-dimensional fine art, but 3 dimensional products such as jewelry, art, wine, are orders of magnitude harder to train AIs on due to the higher number of components and unique forms of patina, not to mention the lack of a centralized training database.

Although human intermediation will remain important in the forseeable future, co-intermediation with AI may become reality sooner than expected. Consumers may prefer using natural language to identify potential purchases, rather than using a platform’s rigid idea of how they should navigate menus to filter and discover new interests. To quote

Arnold Kling

again from this post,

With the advent of large language models, I expect to see menus die. As you know I think of LLMs as the next evolution of the human/computer interface.

In the near future, I will not have to figure out the WSJ’s menu. Instead, I will simply type in or say “I want to submit a letter to the editor,” or “I am going out of town next week, so don’t deliver the paper then.”

The LLM interface will have two advantages. A relatively minor advantage is that it saves users the trouble of learning menus.

The major advantage is that the user will not be limited by the “features” as specified by menus. Users will be able to make up features as they go.

From the WSJ, I might ask for a comparison of stories from the first week of Trump 1.0 to the first week of Trump 2.0. Or I might ask for a list of predictions made by various columnists over the past five years and how those predictions turned out.

I do not picture the WSJ providing me with a large language model. I think that I will use something generic, like ChatGPT or Claude today. The WSJ will get out of the business of software design altogether, other than setting up a database.

And AI’s impact will not be relegated to just interfaces - LLMs and/or their derivatives will help us to research, source, curate, compare and discover value much more efficiently and intuitively - but that is a topic for a different post.

Online b2c platforms will have to adapt and evolve to this new reality. Platforms such as Chrono24 may find that it is no longer sufficient to attract consumer attention by relying on traditional SEO for Google search. Agile b2c intermediaries will ensure that their publications (whether blog posts, social media videos or podcasts) are used to train AI’s, in order to redirect attention to them. AIs may also amplify the importance of handling and comparing product in large showrooms with abundant inventory.

Regardless of these inevitable changes, a centralized b2b platform will remain the foundational database upon which future forms of b2c intermediation will be based.

B2B Trading Startups

Younger people may not recall, but about 10-15 years ago, there was an investment boom by VC’s in b2b marketplace startups, which mostly failed. High sales commissions, sales leakage, out-of-platform trust structures, overbuilding for unnecessary services, difficulty in scaling, and the rise of influencer/social media trading were some of the factors that led to their demise.

From approximately the mid to late 2010’s, trading shifted to Facebook groups, Instagram, etc as all participants regardless of trading profile attempted to influence and intermediate to capture high margins from selling directly to consumers. Interestingly, in the $23Bn secondary watches market, Chrono24 was inadvertently a kind of online b2b trading platform, with reasonable advertising fees and clear and low sales commissions.¹³ Chrono24 could have continued its role as a centralized b2b and b2c platform, but they have now chosen to specialize in the latter. And most recent startups in the secondary art, jewelry, watches and collectibles markets also play in this consumer-facing b2c space. Who would want to play in the low-margin b2b space anyways?

So I was pleasantly surprised when Worldwatch.market showed up on my Instagram feed. It is the only platform I know that is solely for b2b and b2c dealers, kind of like an online version of the IWJG/WTA show, with zero sales commissions. Although the site is not live yet, Stan Duin¹⁴, the founder, showed me some initial mock-ups which were quite impressive. When we spoke, I felt that he understood all the pain points of professionals in our field, as he is an active vintage watch dealer, not just a tech bro. A feature I really liked was allowing members to list on various b2c platforms with a couple clicks, a huge time saver. Like the MLS, I hope that Worldwatch.market becomes a best practice applicable to other growing secondary markets, including art, jewelry, wine, clothes, furniture, home decor and other collectibles.

1. The roles are fluid and not mutually exclusive - most b2c do some liquidation to b2b, and some b2b have consumer clients. A centralized platform will not change this.

2. Authorized Dealers (AD’s) such as WatchesofSwitzerland source secondary Rolex watches which are then sent to Rolex to be authenticated under the Rolex Certified Pre-Owned Program. As is evident from the premiums that CPO watches sold by AD’s command over non-CPO watches sold by grey dealers, authentication by manufacturer is a source of value for consumers. WatchesofSwitzerland has also acquired vintage watch dealer AnalogShift and watch publication Hodinkee. Both add to WatchesofSwitzerland’s intermediation value, especially for manufacturers that lack a CPO program, and is the primary reason for their acquisition.

3. Books, music and film are markets where disintermediation via technology has succeeded - today, anybody can become a creator, and the supply of creators has increased, albeit disproportionately to attention. In contrast, disintermediation in the art, jewelry, watches and other markets has failed because consumers require intermediaries to ensure aesthetic value, provenance, mechanical guarantee and authenticity. And in many markets, consumers expect intermediaries to assist in resale. There is also the controversial subject of whose intermediation value is higher? which I will not get into here. For more on disintermediation, see here and here.

4. Central to how we view watches, jewelry and many other collectibles is the dichotomy of the consumer durable vs art mindset. Newer product are consumer durables, but the majority of secondary stock is probably purchased as art.

5. Another flaw is that major manufacturers such as Rolex will not allow Rolex CPO watches to be sold on marketplaces such as Chrono24.

6. Let’s take auction prices as an example. If an auctioneer can be considered a seller whose markup is capped at the buyer’s premium, is the hammer price (excluding premium) equal to a wholesale price? The answer to this question is it depends. As I wrote in my previous post, a price change does not give us enough information to understand its signal. Were the bidders wholesalers? Are they bidding from a location where there are supply constraints to obtaining the auctioned item? Were there semi-pro dealers or retailers involved? Were consumers and collectors involved, as seems to be more and more the case? How many participants? Does the auction house’s prestige (or lack thereof) affect the price? Due diligence and participation constraints? Empirically, small, unknown auction houses have achieved record prices because of their obscurity, not despite it? Among other unanswerable questions.

7. There are many types of chatgroups - some are just an extension of retail. Here I’m talking about the more transaction-based chatgroups that a diversified mix of b2b, b2c and even consumers are increasingly utilizing. They are characterized by:

1. Fierce competition among buyers

2. Different expectations of risk and handholding

3. Minimal due diligence

4. Incessant 24-hour notifications

5. Scrolling upwards prevents accurate pulse on supply and turnover

6. Lack of listing standardization

7. Perceived exclusivity

8. Evolving participation

9. Blind to other chatgroups (and duplicate listings)

10. Tendency to go stale after a while

11. Other issues which I will not list here…

I’m not anti-chatgroup - they have their place. But chatgroups demand tremendous labor capital with minimal information on actual supply/turnover in the broader market. Extracting all the transaction data from various chatgroup islands and building a ‘Chatgroup Price Index’ would be a great benefit for all participants, if access can be obtained from existing and newly formed chatgroups. The data would also have to be cleaned cost-effectively to control for the non-trivial amount of consumers involved, duplicate listings, lack of listing standardization, etc. But still, chatgroups with their incessant notifications and scrolling are a real grind.

8. I think that art, watch and jewelry dealers get the most number of notifications on their phones than any other profession.

9. There are attempts to circumvent and dismantle the MLS. (ie the “coming soon” listings.)

10. Different platforms provide different levels of authentication, provenance, mechanical guarantee and authenticity value. Chrono24 has recently started an authenticity guarantee program, but it is not clear whether this service is valued on par with manufacturer-affiliated programs or authentication by a well-known dealer.

11. The Rolex CPO program is WatchesofSwitzerland’s 2nd largest brand and has the potential to deliver 20% of their overall revenue. CPO command significant premiums vs non-CPO.

12. Costs associated with handholding - customer service, fraud protection, authentication, etc - are significant for Chrono24 and a major reason for their relatively low operating margin.

13. This is the proverbial elephant in the room in the recent Chrono24 controversy, and is invisible to non-professionals.

14. You can contact Stan Duin (info@vintagemasters.eu or sales@worldwatch.market) via Worldwatch.market if you are a b2b or b2c dealer. Although the site is not live yet, you can sign up for a slot by filling in a questionnaire. If you tell him I introduced you, I get a discount, so thanks in advance.