The Next Generation of Automated Settlement

We have elaborated the circuitry of DeFi, and must now realize its potential to replace the oil lamps, hand cranks and steam engines of the present financial landscapes.

AccessTimeIconDec 21, 2022 at 3:51 p.m. UTC
Updated Sep 28, 2023 at 2:24 p.m. UTC
AccessTimeIconDec 21, 2022 at 3:51 p.m. UTCUpdated Sep 28, 2023 at 2:24 p.m. UTC
AccessTimeIconDec 21, 2022 at 3:51 p.m. UTCUpdated Sep 28, 2023 at 2:24 p.m. UTC

“Quantity has a quality all its own.”—Thomas A. Callaghan Jr.

Although the ethical and legal dimensions of the FTX meltdown will be subject to ongoing debate, this much is clear in the wake of the biggest value destruction event in the history of our industry. CeFi, or centralized finance, is fatally flawed and no longer has a defensible value proposition to stand on. Exchanges have already seen the retail sector slip out of reach due to simple unit economics. Amid the bull rush, Super Bowl commercials and namesake stadiums pushed CAC/LTV ratios [lifetime value divided by the customer acquisition cost] beyond the realm of reality. And then when the market tanked in May, the not-all-that-large-to-begin-with pool of retail interest dried up abruptly, to the point that taking it on the chin became the default expectation for the family Thanksgiving dinner.

And now, after Sam Bankman-Fried and Caroline Ellison, there will be no good reason for large, well-run banks to use CeFi exchanges either. Whether the former heads of FTX and Alameda Research were guilty of criminal greed or just incredulous levels of stupidity is irrelevant. CeFi will head off into the sunset, as it deservedly should. “Disintermediation as a service” is prone to corruption and waste. It’s a bad premise on which to build a business, let alone a category.

Mike Maizels is the former head of research at Abra. This op-ed is part of CoinDesk's Crypto 2023 series.

As such, the next logical step for the industry is to turn its attention to where cryptocurrency potentials collide with the needs and frictions of legacy financial institutions. Decentralized finance (DeFi) may represent the financial “pipes” of the future, but it is the old world of traditional finance (TradFi) that still holds all the water. To date, adoption has been hindered by regulatory uncertainty, market volatility and the risks inherent in diving in to provable digital scarcity as an early entrant. But lying behind these immediate roadblocks, the way forward becomes somehow even less clear. Perhaps the most fundamental issue of all is sheer size. Consider the just market for foreign currency. JPMorgan settles forex trades with the combined value of the entire cryptocurrency sector every day.

Blockchain technology has much value to bring to this market, particularly through the novel capabilities of stablecoins. Their potential for instant settlement and close-to-zero marginal costs – which could unlock dramatically lower foreign exchange (FX) rates – is tantalizingly real, but the infrastructure needed to support adoption at scale is critically missing.

Scale matters because almost all present-day crypto settlement products were built under retail-sized premises. Namely, low marginal cost settlement has been made possible by the remarkable maturation of “automated market makers,” also known as AMMs. AMMs perform the work of legacy trading firms and banks, offering liquidity in the form of standings offers to buy and sell assets at specified prices – all without human intervention.

The ability to do so without human input lies in an elegant use of simple mathematical principles to maintain balance in liquidity pools. Two assets, X and Y, are held in a pool, with the product of their cumulative values linked to some constant. For the purposes of a simple explanation, assume the assets are equivalent to each other in price. 100 X units and 100 Y units makes a pool constant of 10,000. Now imagine a trader wants to acquire 10 units of Y. That would leave 90 in the pool, a deficit of 11.1 X units (10,000/90 = 111.1). This becomes the price (plus a small fee) paid by the trader for her new 10 units of Y. Note that 10 units of Y for 11.1 units of X represents an 11.1% premium over the abstract, 1:1 price parity under which the pool was established.

Importantly, this slippage grows exponentially as the trade size approaches the limits of the pool depth. Imagine if the trader wanted to acquire 99 units of Y. The price paid would be 9,900 units of X, a premium of just under 1 million percent. In practice, these unit economics limit the institutional utility of even the largest DeFi pools – Uniswap currently supports a daily trading volume less than a tenth of a percent of JPMorgan’s foreign exchange business.

Owing to such limitations, large institutions wishing to transact at scale have had no real choice other than to use custodial CeFi exchanges like FTX. Custodial here is key to the potential for abuse. Unlike traditional FX desks, which operate as deal-brokers but would rarely take custody of liquid assets, cryptocurrency exchanges take possession of the tokens in order to execute orders directly. This model, akin to permitting a real estate agent to move into your house and take possession of your deed as a precondition of sale, prepares the ground for disaster. Not only can exchanges front-run trades, profiting at the expense of clients, but the lack of deep liquidity makes it necessary to advertise large limit orders in a public manner. These limit orders may be anonymized to an extent, but even public awareness of the existence of large orders skews market intelligence to the detriment of all who seek to participate.

Builders are hard at work creating solutions for this blockage across a spectrum of time horizons. In the most immediate term, regulated cryptocurrency service providers are beginning to stand up over-the-counter (OTC) desks along the models of currency exchanges of yesteryear – working the phones on behalf of clients rather than taking custody of their assets in order to execute trades on their behalf. This will no doubt drive value; it is a matter of seconds to move stablecoins on a blockchain rather than days to settle fiat exchange through SWIFT.

However, the use of human brokers foregoes the signature advantage of DeFi: trustless, automated functioning. Uber would be a remarkably less-efficient company if human switchboard operators were required to match riders to drivers. Tech-driven systems such as on-chain order books suffer from the converse problem. Their trustless nature necessitates a publicly visible set of trading positions, a flaw akin to poker players betting with cards displayed to all of their opponents.

However, a bottom-up approach to new tools to connect AMMs and to expand the kinds of instruments that they support (e.g., options with an expiration date) has the potential to push back the horizons of the possible. These topics have received important theoretical treatment by Dr. Maurice Herlihy, An Wang professor of computer science at Brown University, who has demonstrated that AMMs may be considered like electrical circuits. Properly constructed, they can operate either in series (trade asset A for B on one AMM, and B for C on another) or in parallel (split a large trade between A for B between multiple AMMs), arrangements that may be considered as mathematically equivalent to one larger AMM. Significantly, these different configurations can redistribute the costs of participating between traders (slippages, as described above) and capital providers (the much rehearsed impermanent loss). Such costs can never be fully eliminated, but they can be subjected both to rational minimization and strategic hedging (e.g., a synthetic “impermanent gain” token that offers insurance for impermanent loss).

The analogy with the electrical circuit is extremely important. The modern discovery of electromagnetic phenomena occurred in the late 18th century, and it took decades for the rules governing their operation to be formulated mathematically. It took several additional decades for the discipline of “electrical engineering” to emerge, which sought to harness the principles of electricity to perform socially useful functions. The first product, the telegraph, enabled only the exchange of direct messages, akin to the first permissionless, peer-to-peer payments on the Bitcoin ledger. The modern electrical infrastructure required numerous other independent inventions to come online together – the dynamo for generation, the battery for storage, the resistor and capacitor for modulation. But soon, these and other devices permitted not just inventors, but entrepreneurs to create technological marvels that transformed the fabric of the world.

The cryptocurrency sector is fond of the analogy of “speed running the history of finance,” but we are also reprising the development of many other infrastructure-level technologies. From an electrical perspective, we are perhaps now in the 1860s. We have elaborated the circuitry of DeFi, and must now realize its potential to replace the oil lamps, hand cranks and steam engines of the present financial landscapes. Wiring AMMs together will require huge complementary efforts: rethinking security procedures and value capture through transaction ordering, a topic animating the widespread interest in grappling with maximum extractable value (MEV).

The history of electricity has one other valuable lesson to teach us. While the mythos emphasizes independent inventors with their lightbulb moments, the reality was that the path from idea to impact was blazed by new modes of collaboration between institutions of higher learning and the world outside their walls. Prior to the 19th century, colleges and universities saw themselves as space apart – zones of aristocratic privilege or monastic scholarship in which inquiry was prized for its own right and disparaged if it became tinged with social or commercial usefulness. That began to change in Germany and the U.S., with emergent departments of electrical engineering often leading the charge to carry out research designed for commercial or civil application. In all likelihood, the present crisis in graduate education, with soaring tuition costs and vanishing opportunities for “pure” academic positions, represents another epoch-defining shift. It will be to the great benefit of the cryptocurrency sector to find ways to capitalize.

With thanks to Robin Malik for his contributions.

Learn more about Consensus 2024, CoinDesk's longest-running and most influential event that brings together all sides of crypto, blockchain and Web3. Head to to register and buy your pass now.


Please note that our privacy policy, terms of use, cookies, and do not sell my personal information has been updated.

CoinDesk is an award-winning media outlet that covers the cryptocurrency industry. Its journalists abide by a strict set of editorial policies. In November 2023, CoinDesk was acquired by the Bullish group, owner of Bullish, a regulated, digital assets exchange. The Bullish group is majority-owned by; both companies have interests in a variety of blockchain and digital asset businesses and significant holdings of digital assets, including bitcoin. CoinDesk operates as an independent subsidiary with an editorial committee to protect journalistic independence. CoinDesk employees, including journalists, may receive options in the Bullish group as part of their compensation.

Mike Maizels

Mike Maizels is the former head of research at Abra.

Read more about