How Blockchain Affects Our Coverage
We believe companies with moats from switching costs and intangible assets are less vulnerable.
Companies that engage in financial transactions, data management, and marketplace activity are seemingly at risk, as blockchain technology can potentially lower transaction costs as well as the costs of recordkeeping. However, we’ve identified some narrow- and wide-moat companies that should be less vulnerable to the blockchain threat. These companies provide value-added services along with simpler transaction processing and recordkeeping functions. For our previous articles in this series, see "Blockchain: Disruption by Decentralization?" and "How Is Blockchain Used?"
Decentralization via blockchain offers several potential benefits to customers. For example, bitcoin offers more privacy, greater transparency, more security, and lower costs (at least in theory) than centralized methods of payment processing.
On the other hand, these benefits are not always the sole factor in a customer’s decision process. The source of a company’s economic moat plays a role in its susceptibility to disruption by decentralization. We recognize five distinct moat sources: cost advantages, intangible assets, switching costs, network effects, and efficient scale. All five can prevent disruption by decentralization. It takes years to build the network effects and intangible assets possessed by some of the companies we cover. Switching costs--especially those related to complex technology--can be quite high. Decentralization does not always lead to lower costs--the redundancy inherent to initial blockchain applications adds time and expense over centralized solutions. Efficient scale moats are created when centralization is needed to attain satisfactory returns on capital and is often grounded in measures of proximity that blockchain does not address (railroads, airports, and so on).
Cost advantages often stem from production at scale--a clear benefit to centralization in most cases. Retailers such as Amazon (AMZN), Costco (COST), and Walmart (WMT) use scale to procure goods at low cost and to distribute them cheaply. Similarly, providers of cloud computing such as Amazon, Microsoft (MSFT), Alphabet (GOOG), and Alibaba (BABA) spread the large costs of hardware, software, power, and staffing across a broad base of customers.
Other moat sources are less dependent on the centralization of activity. Intangible assets confer competitive advantages via pricing power associated with brands, patents, proprietary technology, and/or regulation. The consumer sector abounds with strong brands, while healthcare companies rely on patent protection to sustain excess profits.
We believe companies benefiting from switching costs can also be well protected from the disruptive threat of decentralization. Switching costs sometimes arise because of accumulated customer expertise. Customers of technology companies such as Adobe Systems (ADBE), Autodesk (ADSK), and Microsoft (Office) would incur significant costs to retrain on a new software package. Similarly, doctors sometimes invest years learning the intricacies of a particular medical device. Enterprise software companies often benefit from high switching costs due to long product cycles, high costs of implementation and training, and difficulty of data migration. The mission-critical nature of many enterprise software applications--including financial, supply chain, and data management--increases switching costs, although these functions are some of the most mentioned use cases for blockchain technology.
Some of the widest moats are held by companies able to centralize economic activity. Network effects are a strong force driving activities toward centralization. Globally, we cover around 100 companies across various sectors with wide or narrow moats attributable to network effects. Within the consumer space, marketplaces such as Expedia (EXPE), Priceline (BKNG), eBay (EBAY), MercadoLibre (MELI), Alibaba, Rakuten, Ctrip (CTRP), and Amazon create value by coordinating the transactions of buyers and sellers. In financial services, American Express (AXP), Visa (V), Mastercard (MA), PayPal (PYPL), Western Union (WU), and Discover (DFS) transfer funds from consumers to merchants. Exchanges such as Deutsche Boerse, London Stock Exchange, Intercontinental Exchange (ICE), ASX, and Singapore Exchange allow customers to trade securities. Industrial companies such as Expeditors International (EXPD) and C.H. Robinson (CHRW) serve as marketplaces for transportation services. In the technology sector, social networks such as Facebook (FB) and Weibo (WB) coordinate social interactions, while platforms such as operating systems of Microsoft and Alphabet create networks of developers and customers.
These are all centralized networks, while blockchain enables decentralized networks. Network moats are powerful because they are often difficult to form and to disrupt, and because they rapidly lead to efficient scale in many cases. Forming a new network business typically requires large amounts of funding and a brilliant business plan. Blockchain tokens, however, provide a completely new way to encourage network participation. Network users and developers benefit from their participation as demand for services grows. This is the key mechanism by which current economic paradigms could be disrupted.
Here we examine how blockchain technology could be applied to a variety of economic activities and identify the moat sources that could allow incumbents to resist the threat of decentralized competition.
Bitcoin is the first and most obvious example of blockchain technology. The cryptocurrency is now worth tens of billions of dollars, with billions of dollars changing hands in (primarily) peer-to-peer transactions on a daily basis. With merchants clamoring to lower their cost of accepting payments, the consumer-to-business payment market (with its multiple highly profitable intermediaries) provides a straightforward use case for a decentralized ledger.
However, nearly 10 years after it was introduced, the consumer-to-business bitcoin ecosystem is still in its early stages, with numerous parties still in the process of building a network of consumers, merchants, miners, and bitcoin businesses. Unlike the traditional payment paradigm, there is no theoretical need for financial institutions acting as issuers and acquirers. In practice, though, companies like BitPay and Coinbase provide services to merchants, and a variety of companies provide storage in the form of bitcoin wallets and exchange other currencies for bitcoins.
To pose a threat to the current payment paradigm, the bitcoin ecosystem must establish a network effect comparable to those of Visa, Mastercard, PayPal, and other major players. This will require a critical mass of each of the participants. Just as Visa’s network’s value depends on its enormous base of cardholders, issuers, merchants, and acquirers more so than its technology, the bitcoin ecosystem’s value depends on the same.
The primary obstacle is that bitcoins do not have large enough base of users to compete with the network of Visa and Mastercard. The importance of the network effect to payment providers cannot be overestimated. In our view, the decentralized nature of bitcoins is a major disadvantage in building a network of users and merchants. Credit cards and debit cards rely on banks for distribution to customers. This provides the Visa and Mastercard brands with an enormous base of users. PayPal was able to build its network by focusing on eBay transactions and providing a valuable service to users and merchants selling online; it is still attempting to establish a physical presence. Sears (SHLD)--the largest U.S. retailer at the time--originally issued the Discover card, which has only recently expanded acceptance to a level competitive with Visa and Mastercard. At present, bitcoins essentially rely on word of mouth to gain consumer users, and only a few small startups are actively attempting to sign up merchants. Merchant acceptance is rumored to be falling, not growing. Thus the new payment paradigm has a steep hill to climb if it is ever to threaten Visa and Mastercard. We note that companies that have already established extensive user bases, such as Google, Apple (AAPL), and Facebook, have a major distribution advantage over bitcoins but have yet to introduce successful proprietary payment systems.
In addition to their distribution systems, the established payment networks put significant effort into marketing campaigns. Over the years, these expenditures have contributed to the establishment of valuable brand intangible assets and helped the companies build their customer bases. In our view, a decentralized system like bitcoin cannot hope to match these efforts on its own.
We think the relative lack of protections will be an issue for consumer adoption. Bitcoin transactions are designed to be nonreversible--a benefit for merchants who would no longer need to take responsibility for preventing fraud, but a major impediment to consumer adoption. Finally, the current payment paradigm passed benefits on to consumers in the form of rewards. In theory, the lower cost of bitcoin transactions would eventually be passed on to customers in the form of lower prices, but the initial loss of benefits is also likely to hinder consumer uptake.
On the merchant side of the equation, barriers to adoption also exist. First, it is costly and difficult to roll out new point-of-sale technology to physical locations on a large scale. Square has used iPads for small merchants, but PayPal has partnered with Discover in order to achieve acceptance, and the United States is still waiting for the EMV chip-and-PIN infrastructure that would significantly reduce fraud.
That said, bitcoin/blockchain technology may be able to provide cost savings to merchants. Traditional merchant acquirers charge discounts averaging around 2.3% of transaction value, which is split among issuers, networks, acquirers, and cardholders. Bitcoin appears at first glance to be a much cheaper option. Bitcoin’s lack of reversibility is an obvious source of cost advantage for merchants. We estimate that fraud costs account for about one fifth of a typical 2.3% merchant discount fee. The bitcoin peer-to-peer network, perhaps surprisingly, is not a major advantage to merchants. We estimate that network operation costs, including marketing functions, account for only a small portion of a typical discount fee, with network operating profits making up a slightly larger portion. Card issuer and acquirer operating costs make up almost half of the discount fee. At first glance, this seems like an obvious advantage for bitcoins. However, most merchants and cardholders will need to employ similar services to access the bitcoin network, keep records, and so on, and these newer companies will not enjoy the scale of large card issuers and acquirers. In our view, a bitcoin cost advantage depends primarily on shifting the burden of security away from banks and merchants to the buyer.
We also think bitcoin’s current competitive advantages pale in comparison with the economic moats established by existing payment companies. For bitcoins to pose a real threat, the early-stage business models based on the technology must figure out how to increase benefits to consumers while developing powerful distribution models, trusted brands linked to--but separate from--bitcoins themselves, and keep costs of acceptance low. At this point, we believe the probability of a company succeeding across all these fronts is minimal. While this shouldn’t and won’t stop venture capitalists from investing in bitcoins’ potential, we think the excitement about the cryptocurrency may be overdone.
The final obstacle is that bitcoin processing alone can already be quite expensive. The average transaction fee is currently less than $1.0018, but fees have spiked well above that over time. In comparison with the average U.S. debit transaction, bitcoins are currently no bargain.
Blockchain promises to eliminate centralized control of transactions, reduce transaction costs, and eliminate redundant records. Financial markets are some of the most complex, costly, and redundant systems in the economy. For example, post-trade processes currently involve payment systems, securities settlement systems, central securities depositories, and central counterparties, often across multiple jurisdictions. The Federal Reserve has outlined a few key motivations leading financial companies to explore blockchain solutions, including reduced complexity, improved processing speed, decreased need for reconciliation, increased transparency, improved network resiliency, and reduced operational risk. In general, combining trading and post-trading activities into a single step offers immense increases in speed and decreases in costs. Accenture estimates that investment banks could save 70% on central financial reporting, up to 50% on compliance, and 50% on centralized client management activities and back-office operations (including clearing and settlement).
The Fed’s report on blockchain also highlighted the importance of network effects in payments, clearing, and settlement activities. Both market participants and regulators would probably have to join and participate in a blockchain-based system. It follows that an industry consortium may be the source of new technologies, rather than a disruptive, decentralized force from outside the industry, as it would be difficult for a disrupter to gain enough buy-in from the major financial industry players.
Switching costs are another factor. Banks are still using legacy systems dating back more than 30 years in some cases. In addition to the need for a new technology to be interoperable with these systems, the advanced age of some existing financial systems illustrates the slow speed at which the financial industry moves and the high switching costs for participants.
In the world of custody, switching costs are high. Client back-office infrastructures are often tightly connected to the infrastructures of a custodian. The services provided are mission-critical, with extremely high costs of failure. Additionally, large custodians are typically providing multiple value-added services in addition to basic custody and reporting, further increasing switching costs. The complex needs of institutional investor clients extend far beyond basic post-trade processing and recordkeeping.
While blockchain promises to decrease costs, the cost savings in post-trade and custody may not be the best opportunity for blockchain. The largest custody banks, exchanges, and industry utilities (such as Depository Trust & Clearing) are already well scaled: They are custodians of trillions in assets and process tens of millions of trades per day. Additionally, because of the scalable nature of many of these activities and either competitive forces or the nature of being an industry utility, cost of services is generally kept low.
We don’t see blockchain technology replacing many of the trading and clearing functions of the wide-moat futures exchanges that are based in intangible assets. While blockchain can be used in some aspects of trade processing, one of the main value-added aspects of trading at the futures exchanges is in their created or licensed intellectual property. While a new blockchain system could enable trading, it may not have the ability to trade products that belong to other entities, such as S&P 500 futures. It would also be difficult for any new entrant using blockchain to duplicate the physical infrastructure, such as warehouses, needed to trade and settle many physical commodities, such as metals and agriculture.
Important functions of clearinghouses may not be immediately replaced by blockchain. For the futures clearinghouses, their main function is the amelioration of counterparty credit risk. A blockchain entrant into futures clearing would have to gain the trust of many of the largest financial institutions and be able to transfer assets from their accounts when their trading positions are in a loss position to ensure that the holder of the other side of the trade that is in a gain position can be paid. For futures clearing, we currently view blockchain as a way to make existing processes more efficient, rather than a completely disruptive force. In fact, the Australian Securities Exchange is already planning to use blockchain technology for clearing and settlement purposes.
Companies affected: Amazon, Microsoft, Dropbox (DBX).
A blockchain is a form of distributed computing power, so it makes sense that technology companies entered the fray in short order, using blockchain to decentralize computing power. Ethereum itself is a distributed computing system. Filecoin, a decentralized storage network competing with Box (BOX) and Dropbox, raised more than $250 million in its initial coin offering. At first glance, it seems that decentralized, distributed computing resources could replace large corporate providers in short order.
However, this use case illustrates some of the tensions inherent to the sharing economy. The most successful companies in this business, such as Uber and Airbnb, are essentially a means of monetizing excess or underutilized capacity rather than a means of organizing all such capacity. Indeed, Uber is already embarking on a program to maintain its own fleet of automated vehicles, rather than merely match supply and demand. Similarly, computing power is increasingly being centralized due to the massive cost advantages that accrue with scale. Corporations are increasingly outsourcing the capital-intensive, high-fixed-cost functions of computing power and data storage supply to a handful of large vendors rather than running these functions in house and attempting to sell varying amounts of excess capacity on the open market.
Technical considerations are an obstacle to some blockchain-based solutions. Centralized control is actually necessary for certain types of processing, including serial computations and others that are difficult to run in parallel. Decentralized systems work well for tasks that can easily be broken into smaller pieces and when a high degree of redundancy/resiliency is required. IBM’s (IBM) Watson, for instance, is made up of a cluster of processors, providing enhanced communication speed and flexibility (ideal for Jeopardy), while the BOINC system running SETI@Home and other similar projects runs on thousands of smaller computers around the world (ideal for searching large amounts of data for alien signals).
Once again, security and data sovereignty issues also favor the incumbents over a distributed solution. Sensitive data is often required to be stored in a specific jurisdiction. This is an issue for all providers of cloud-based solutions, not just blockchain.
Cloud computing can be broken down into three segments: infrastructure as a service (outsourced virtualization, servers, storage, and networking), platform as a service (outsourced infrastructure, operating system, database, middleware, and runtime), and software as a service (outsourced infrastructure and platforms plus applications and data). Blockchain-based distributed systems are initially offering primarily infrastructure as a service and platform as a service.
Cost advantages stemming from scale favor the incumbents in the IaaS market. Vendors must be able to generate scale efficiencies on the input costs that are included with running a data center: technology hardware, infrastructure software, land, power, cooling, networking, and IT staffers to manage the infrastructure. This requires significant initial investment to build out a large enough network of data centers to meet customer service-level agreements, clear governance, data sovereignty, and regulatory hurdles and provide enough supply in the event of outages or spikes in demand. Again, the services provided--above and beyond commoditized computing power--are the more valuable part of the IaaS business.
We believe competitive advantages in platform as a service will primarily come by way of switching costs. Most PaaS offerings create a natural lock-in for customers, as applications are developed on a fixed software stack that often features proprietary components such as programming languages, data sets, and other software layers. At the same time, large tech companies such as Microsoft, IBM, and Amazon are adding blockchain as a service to their current PaaS offerings.
Credit reporting agencies demonstrate the risks associated with centralized control of ostensibly private data. Major credit bureaus have suffered breaches in recent years, including Equifax’s loss of millions of customers’ personal information in September 2017. Other private data, such as healthcare records, may be equally vulnerable. A secured, shared distributed ledger offers the promise of greater security and privacy for customers. Ventures intending to produce such a system have already launched. Bloom, for instance, allows users to create an identity, allows others to verify that identity, securely stores debt repayment data, and produces a score without revealing the personal data it uses.
However, disruption of these business lines also seems unlikely. The credit bureaus’ moats come from valuable intangible assets stemming from data collected over years; established credit bureaus have a long head start over newcomers.
Logistics functions involve the coordination of information among various parties in the supply chain; the storage and transportation of goods often requires multiple intermediaries. Communication between parties in trucking remains quite manual; sometimes it takes brokers and asset-based truckers hours to complete a transaction, given the need for multiple phone calls and emails including scheduling, rate negotiation, status updates, and so on. Transportation markets seem like a good fit for simplification via distributed ledgers.
In fact, transportation markets are already beginning to evolve. Digital freight markets such as Uber Freight now match shippers and truckers on demand, using algorithms to replace the intermediaries that have historically brokered such transactions. We think the key differences between DFMs and traditional brokers such as C.H. Robinson, Echo, and Landstar involve the freight apps’ digitized, head-count-light infrastructure with no (or very few) salespeople or commissions and limited carrier sourcing personnel. The large traditional brokers use a model combining technology, back-office head count, and highly proactive salespeople. The digital freight matching model depends on the development of a broad network of small carriers and lower-cost, lower-priced service. Freight apps are likely to gain some traction in the years ahead, and they could prove to be a first step toward the application of blockchain technology in logistics longer term, but we don’t think this spells disaster for moatworthy brokers.
The network effect, which supports the economic moats of several truck brokers Morningstar covers, provides context for framing the threat of digital freight matching. At its heart, the network effect implies that the more parties (suppliers and customers) that use a logistics provider’s network, the more powerful the network becomes and the harder it is to replicate. Large traditional brokers such as C.H. Robinson, Echo Global Logistics, and Landstar have amassed an immense customer base of shippers that aggregates sufficient demand to attract substantial truckload capacity while bestowing robust buying power relative to small and midsize brokers and shippers. Along these lines, vast capacity relationships and deep lane experience have allowed C.H. Robinson and Echo to win highly service-intensive price-committed business among large shippers over the years. Also, these companies are no slouches on the IT front; heavy investment has allowed them to monetize their deep reservoirs of market transaction data, and we expect that to persist.
Sourcing capacity in the incredibly fragmented truckload market first requires freight density and numerous carrier-facing sales personnel to be in constant contact with small fleets, which make up most of the carrier base. The more that capacity tightens, the more a broker must work the phones to secure trucks to get freight moved. C.H. Robinson employs roughly 1,300 carrier-sourcing reps, while mobile app-based marketplace Convoy had around 120 total employees as of mid-2017. In theory, a blockchain-based rival would rely on a more organic (and most likely slower) method of expanding the carrying capacity of the network.
Sales and service activity plays a role in establishing network effects and generating intangible assets in the form of customer goodwill. To date, trucking activity requires human support, especially for manual oversight of location tracking, appointment scheduling, and exceptions like billing errors, cargo damage, and late arrivals. It will be some time before all of these activities can be automated.
All types of marketplaces can, in theory, be moved to a blockchain. Amazon and Alibaba have disrupted dozens of industries by connecting buyers directly to sellers. However, these companies have done so by taking on the role of “trusted third party.” Blockchain technology and smart contracts--by solving the trust problem--can remove the need for these newer corporate intermediaries.
In general, the more commoditized the product or service, the more amenable to a move to the blockchain. Israel’s LaZooz, for instance, already provides a decentralized transportation service like Uber’s. OpenBazaar is ostensibly a rival to marketplaces such as eBay or Etsy. Even for companies that provide only a marketplace, existing network effects are difficult to break, and new networks are difficult to establish.
Morningstar covers a wide variety of industrial distributors, most of which provide value-added services in addition to serving as a marketplace. These supply chain and inventory management services include customer-managed inventory programs, vendor-managed inventory programs, industrial vending, and a host of related solutions. These solutions can eliminate the need for in-house procurement professionals and reduce product consumption and leakage. Many distributors have technical expertise and offer operational reviews that can save customers time and money. Many maintain inventory and benefit from buying power produced by economies of scale. Economies of scale are even more prevalent in the pharmaceutical distribution industry: Three companies control 90% of the buying power, while end providers are relatively small and fragmented. Pharmaceutical distributors provide franchising/network infrastructure, inventory management, reimbursement/payment management services, and group purchasing discounts to their customers. While smaller pharmacies rely on these services for essential operational needs, branded manufacturers also use AmerisourceBergen (ABC), Cardinal Health (CAH), and McKesson as their distribution and inventory management divisions. Instead of building internal logistical operations, manufacturers leverage the expertise, warehouse networks, and retail pharmacy relationships of the major distributors.
Morningstar recently conducted an assessment of distributor vulnerability to digital disruption, and we believe the results of this research are applicable to the blockchain threat. Companies that sell specialized products, use their scale to increase customer buying power, and provide value-added services should be resistant to the threat of new digital intermediaries, as well as the complete disintermediation of their businesses via blockchain.
Amazon itself has had more difficulty penetrating the business-to-business market, as specialization and service (intangible assets) are arguably as important as network effects. In the realm of office supplies (with fairly commoditized products), the Federal Trade Commission found that Amazon’s limited request for proposal experience, lack of customer-specific pricing ability, and lack of control over third-party pricing and delivery hurt the retail giant’s ability to compete with Office Depot (ODP) and Staples. Business-to-business distributors typically provide logistics expertise at least on par with Amazon’s; a blockchain-based solution would be at a huge disadvantage on this front.
Jim Sinegal has a position in the following securities mentioned above: GOOG. Find out about Morningstar’s editorial policies.