After hitting record highs towards the end of 2021, the crypto market crashed in 2022. The drop in value was steep: from $2.9 trillion in value in November, 2021 to around $800 billion a year later. And as market value declined, leading crypto lenders and exchanges collapsed, too.
Now, many of these platforms are facing lawsuits and charges of serious wrongdoing: Celsius is being sued for fraud. Voyager Digital faces a class action suit for selling unregistered securities and misleading customers. BlockFi settled with the SEC and 32 states over similar claims, still faces a class-action suit, and ultimately filed for bankruptcy. The co-founder of Terraform Labs, the creator of the algorithmic stablecoin TerraUSD, is the target of two class-action lawsuits and an arrest warrant in South Korea. Now, the centralized exchange FTX, which just recently filed for bankruptcy, is accused of committing “serious fraud and mismanagement,” while its co-founder Sam Bankman-Fried is facing an immense amount of scrutiny from Congress, criminal prosecutors, and civil suits.
For some, this recent turbulence foretells the end of crypto. To them, the sequence of collapses seems to illustrate that crypto is just a house of cards.
But these collapses weren’t at the level of crypto’s foundation. Rather, they reflect a specific segment of the market that was driven by opaque, centralized institutions that made it difficult for their customers to understand the risks they were taking. And the relative ease with which customers could transfer their crypto assets between platforms contributed to particularly intense competition, driving platforms to offer apparently better and better deals with greater and greater degrees of hidden risk.
In this sense, this kind of meltdown isn’t really new — intense competition in markets with opaque products has led to financial collapses in many other contexts. And ironically, whereas making competition easier is typically a good thing in markets, here it may have intensified the problems.
At the same time, however, the centralized institutions that collapsed stand in sharp contrast to other segments of the crypto market — those based around decentralized protocols that operate with fixed algorithms and extreme transparency. These platforms were subject to the same market forces, yet performed relatively reliably during all the turmoil.
In this article, we explain some of the economic dynamics that fueled this chain of events. We then discuss how as decentralized finance matures, it may provide a path toward harnessing the competitive benefits of crypto infrastructure with better transparency and consumer protection.
Competition and Opacity in Web3
To understand how we got here, you have to grasp why competition in crypto/Web3 operates differently from many traditional business contexts.
In established Web 2.0 platforms, as well as in traditional financial institutions, data and assets are locked within “walled gardens,” and thus require significant cost and effort to move. It’s difficult to extract even the content you yourself have created on platforms like Twitter and Facebook, and once you’ve done so, it arrives in a file format that isn’t easily transferred onto some other platform. And meanwhile, people perceive significant switching costs in consumer banking, and as a result, very few seek out better banking products in practice.
By contrast, crypto makes it easy and intuitive to move a user’s assets from one place to another, because those assets are stored and managed on public, interoperable blockchains. This makes the cost of switching platforms incredibly low — users have control over their assets and often do not even need to notify a platform when they leave for a competitor. As a result, platforms have to compete fiercely not just to win users but also to keep users.
It’s almost a truism in economics that such competition should benefit consumers by resulting in lower platform prices and improved quality, as well as by more generally empowering — and protecting — the buying public. This is precisely why regulators and policymakers explicitly encourage competition as a mechanism to improve consumer welfare.
And indeed, we have seen such forces in action in various crypto markets. Leveraging users’ low switching costs, new platforms for trading both fungible and non-fungible crypto tokens have launched and successfully captured market share from incumbents by offering novel features and/or lower fees. Aggregators enable users to compare transaction opportunities frictionlessly across numerous platforms, and the platforms themselves often compete with each other by offering users greater shares of governance tokens and other rewards. “Yield farming” is yet another example; here, users rely on public information about different offerings and low switching costs to algorithmically and instantly reallocate their capital to the products with the highest yield.
There’s a problem, however: At present, some crypto products and services are opaque and possibly too complicated for the average user to understand. And when consumers don’t understand an offering — either because they lack the expertise or they aren’t being given important information — intense competition can lead to trouble. Specifically, competing firms can offer consumers increasingly attractive deals while obscuring potential pitfalls.
This dynamic isn’t unique to crypto. It happens in markets for everything from used cars and wealth management to medical care and insurance. But shrouding potential problems can be especially possible on some crypto platforms because the assets they deal with are novel and sometimes intrinsically complex. The nascent status of regulatory frameworks for disclosure and consumer protection add another layer of risk.
In cases like this, as competition intensifies, companies may tend more and more to offer attractive product features while hiding increasing underlying costs and risks.
When Good Economic Forces Go Bad
The cocktail of intense competition, nontransparent financial products, and unsustainable yields has significant precedent.
For example, deregulation in the banking sector in the 1970s and 1980s increased competition, lowered profit margins, and magnified banks’ risk-taking activities and bank failures. More recently, researchers have shown that when competition intensifies, banks tend to design financial products for households in a way that increases the headline interest rate while increasing the products’ complexity and risk. And of course, the 2008 financial crisis was marked by widespread fraud in the mortgage securitization industry, and this behavior was driven to a large extent by the increased competition in the market, as originators lowered credit standards and engaged in predatory lending to shore up profits.
Thus while competition can and should help consumers, past events have illustrated repeatedly how intense competition in financial markets creates conditions for risky and complicated products to thrive — however briefly — by enticing users with impossibly good deals that inevitably fall apart.
The recent collapse of the crypto market shows this snowball dynamic yet again in the context of a number of centralized platforms. Celsius and Voyager Digital, two major centralized, retail-facing lending platforms now in bankruptcy, offered incredibly attractive yields on customer deposits without fully disclosing the extreme leverage and risk they undertook to achieve those returns. Anchor, a lending platform built and controlled by Terraform Labs, offered consumers unsustainable yields on deposits in an attempt to increase demand for Terra’s stablecoin, TerraUSD, which collapsed later in the year. Meanwhile, FTX, one of the largest centralized crypto exchanges, recruited users by promising discounts on crypto asset trades but then reportedly secretly channeled billions of dollars of customer funds to subsidize the losses of its sister company, the hedge fund, Alameda Research.
In a highly competitive industry, when customers can easily switch financial providers, small changes in promised yields can cause large capital movements. Such movements magnify incentives to improve customer terms artificially, particularly when the financial products themselves are opaque and complex.
Competitive markets can and should drive unproductive companies out of the market. But the lack of transparency in centralized crypto finance allowed unproductive companies to compete by offering products that appeared attractive in the short run but were unsustainable in the long run. Unfortunately, as we’ve seen, such companies can amass significant assets before their business models eventually unravel.
What’s Different About DeFi
Over the past several years, a natural experiment has been running: Alongside centralized crypto platforms such as Celsius, Voyager Digital, and FTX, various decentralized finance (DeFi) protocols have emerged, and in the midst of the broader crypto market turmoil, DeFi has held up.
DeFi platforms manage crypto and other assets through protocols which are stored simultaneously on thousands of computers and use the blockchain to agree on the same results each time the code runs. These protocols are automatically executed through software called “smart contracts,” which ensure that they will operate as promised. DeFi protocols enable making markets, settling trades, processing payments, and providing loans similarly to centralized platforms, but in a disintermediated manner. Because these platforms run on infrastructure that is public, open, and decentralized, their workings are consistent and fully auditable, and their transactions are all observable.
For example, Aave and Compound are decentralized protocols that facilitate over-collateralized borrowing and lending. These protocols set interest rates automatically based on real-time supply and demand to reflect current market conditions. They are effectively in the same business as, for example, Celsius — i.e., crypto lending — but their operations can be publicly verified at any time. Moreover, both Aave and Compound function autonomously. There is no person who takes control of assets and decides what to invest in — all borrowing, lending, and interest-earning happens automatically, so there is no way for humans to introduce additional risk-taking. And indeed, while Celsius boasted with interest rates of 12% for stablecoin deposits (which were substantially higher than the loans they provided, starting at 4.95%), such rates were not feasible for Aave and Compound because they set interest rates mechanically based on market supply and demand. As a result, Aave and Compound were able to operate reliably even as Celsius imploded.
It’s not that DeFi isn’t subject to the same competitive forces as the centralized platforms — if anything, it’s even easier for users to move assets across DeFi lenders than it is to move them in and out of centralized crypto finance platforms. But DeFi by design prevents responding to that competition by undertaking the sort of opaque risks that have doomed Celsius and other centralized platforms.
More broadly, decentralized finance promises an extreme form of transparency heretofore unseen in the financial system. On DeFi platforms, users can observe the publicly verifiable transaction information on the blockchain ledger and confirm their balances, the orderly execution of transactions, and other state changes. Moreover, users can examine the precise software rules that govern the platforms’ operation and assets, as these rules are explicitly written into open-source code. And because platform operations are governed by software, their behavior is predictable even when market conditions change.
This draws a sharp contrast with much existing financial infrastructure — both inside and outside crypto markets — which rely on private and unobserved ledgers. Many centralized finance platforms make decisions and moves hidden from public view in a way that may give rise to risky but unobservable changes in their financial products. For example, whereas FTX reportedly loaned out customers’ deposits without their knowledge and permission — whereas this wouldn’t be possible in DeFi exchanges because with such exchanges, there’s no human control of deposits; rather, all transactions are managed mechanically by a pre-specified algorithm.
That said, DeFi requires much more work before it can live up to its promise. Currently, not all DeFi protocols are based on open-source software — and even among those that are, the code and underlying data are typically hard for anyone other than sophisticated power users to access and parse. Additional tools are needed to ensure frequent software audits, real-time evaluation of on-chain data, and intuitive, human-readable disclosures that will allow average consumers to assess the robustness of these products in a manner that is not feasible today.
And of course, the underlying software must be shored up as well. While DeFi protocols — by design — always function as implemented, this makes them vulnerable to hacks that exploit subtle economic or security flaws in their algorithms. And while decentralized governance helps to align platforms’ incentives with their users’ best interests, it can sometimes open them up to attacks as well, whereby a small number of users collect governance tokens and then try to force system changes – a bit like in a hostile takeover of a publicly traded company.
Finally, we need to incentivize people to build and use DeFi. In heavily competitive markets like crypto finance, as we’ve already seen, radical transparency may be a difficult strategy in the short run precisely because it prevents platforms from offering seemingly attractive but unsustainable deals. But in the long run, transparency is advantageous, and there’s a real sense in which DeFi platforms could be more valuable for everyone, including their creators. Indeed, especially to the extent that these platforms share some degree of ownership with their users, the value proposition for consumers can be higher, which has the potential to lead to more growth down the line.
It’s thus possible that, at scale, DeFi could be bigger and more value-creating than even the most successful centralized platforms. Nevertheless, decentralized platforms still can’t outcompete outright fraud, which means we need to constrain the ability of centralized actors in the space to operate in non-transparent ways.
More broadly, regulation can be a significant complement to DeFi: It can introduce transparency standards that level the playing field vis-à-vis centralized platforms, and it can also provide guidance and structure around which types of protocols are ideal. Likewise, data and modeling platforms can help DeFi protocols make optimal decisions — Gauntlet, for example, performs real-time analysis to advise DeFi governance bodies on how to adjust collateral and collateralization ratios to deal with changing market risk.
• • •
The past year has been full of hard lessons for crypto businesses — and users. But the repeated stress tests have helped show what does work. As DeFi matures, everyday users will have more opportunities to take advantage of the extreme transparency it offers. Hopefully, that will also leave them better equipped to benefit maximally and sustainably from the intense platform competition that blockchain architecture enables.
Acknowledgments: The authors thank Miles Jennings and Scott Walker for helpful comments.
Disclosures: Kominers is a Research Partner at a16z crypto, which reviewed a draft of this article for compliance prior to publication (see general a16z disclosures here). Additionally, both Bernstein and Kominers hold various crypto assets; advise crypto projects; and have used both centralized and decentralized crypto finance platforms.