Case of Decentralized Exchanges – DeFi✭ 

The Case of Decentralized Echanges

TABLE OF CONTENTS

Executive Summary ………………………………………………………………………………………………………………………………..3
Table of Contents……………………………………………………………………………………………………………………………………3
1. What is Decentralized Finance?…………………………………………………………………………………………………………….4
1.1 DeFi from a Technical Perspective…………………………………………………………………………………………………….5
1.2 The Promise of DeFi ……………………………………………………………………………………………………………………..6
1.3 DeFi Applications …………………………………………………………………………………………………………………………7
2. Decentralized Exchanges…………………………………………………………………………………………………………………….9
2.1 Market Size, Major Players, and Market Share ………………………………………………………………………………….10
2.2 Main Types of DEXs ……………………………………………………………………………………………………………………11
2.3 Automated Market Makers – the Current Market Leader …………………………………………………………………….12
2.3.1 How do AMMs Work ………………………………………………………………………………………………………………12
2.3.2 Benefits of AMMs …………………………………………………………………………………………………………………15
2.3.3 Risks and Drawbacks of Using AMMs ………………………………………………………………………………………..16
2.4 Beyond Automated Market Makers …………………………………………………………………………………………………17
3. Conclusion and Outlook…………………………………………………………………………………………………………………….19
References………………………………………………………………………………………………………………………………………….21

DEFI EXPLAINED – THE CASE OF DECENTRALIZED EXCHANGES

EXECUTIVE SUMMARY
Promising to enhance access and efficiency within financial services, Decentralized Finance (DeFi) has gained significant
attention – and some traction in the marketplace – in recent years. In essence, DeFi uses secure distributed ledger/blockchain
technology to facilitate peer-to-peer financial transactions, including borrowing, lending and trading.
DeFi seeks to disintermediate and decentralize the traditional financial services industry by automating complex financial
processes. The functional roles of trusted third parties such as brokerage firms, banks, and other centralized financial
institutions, are replaced by smart contracts – simple, self-executing programs that automatically carry out those functions.
Although from a big picture perspective DeFi is still a niche phenomenon whose long-term impact on the financial services
industry is yet to be ascertained, that potential for disintermediation means it is important that established financial
institutions understand how it might reshape the operational landscape and how they might themselves embrace the concept
of decentralization.
Certainly, a major market correction is currently underway in the digital assets space. That said, the ongoing travails of DeFi
challengers present an opportunity for traditional financial institutions to become more active in the space, stepping in to offer
robust solutions that draw upon on their core strengths of security and trust.
To this end, this paper aims to support practitioners in two ways: first, we offer an overview of DeFi, its promises and key
foundational principles to provide a broad framework to support firms as they assess the DeFi phenomenon. Second, we deep
dive into one of the most promising areas of DeFi – Decentralized Exchanges (DEX) and the perceived benefits of DEXs over
the more liquid centralized exchanges on which the majority of digital assets are currently traded.

1 . WHAT IS DECENTRALIZED FINANCE?
At its core, DeFi aims to provide financial products and services
based on blockchain technology. The term is used rather broadly
to describe the decentralized applications (DApps) providing the
necessary business logic for transactions as well as the underlying
blockchain networks and digital assets. The combination of
decentralized, smart-contract-based business logic solutions
with a blockchain-based settlement layer facilitates the creation
of financial services in a decentralized way.
In contrast to what their name implies, smart contracts are
neither smart, nor are they contracts in a legal sense. Rather,
smart contracts are executable code stored on layer 1 blockchain
protocols like Ethereum. These small software applications are
used to automatically execute program logic or rules written in
the code. If the conditions of the smart contract are fulfilled, the
code will self-execute its set of instructions. With no institution
needed to execute the business logic, this enables participants
to execute transactions in a ‘trustless’ environment.
In the DeFi world, intermediaries are replaced through a set
of automated smart contracts. Functional roles of trusted third
parties such as brokerage firms, banks, and other centralized
financial institutions, are replaced by smart contracts which fulfil
these functions automatically. In this sense, DeFi (similar to other
DLT based use cases) seeks to disintermediate and decentralize
the traditional financial services industry.

There is already a broad range of financial services or products
available in the DeFi space including trading, lending, investing,
deposits, and payments services. Furthermore, decentralized
applications are highly modular. This means, that very often
they can be combined and are interoperable to create new
applications.
DeFi’s rise in popularity can be partly explained by real and
perceived structural issues with the current financial services
industry. DeFi arose out of a desire to free financial services from
the control of centralized institutions and governments thereby
providing financial inclusion for more people. Proponents of
DeFi argue that traditional financial services are dominated by
large institutions and often characterized by tightly controlled
access, leading to organically grown inefficiencies, high and
opaque fees as well as financial exclusion. In addition, they point
to the high level of regulation fostering an environment that is
generally hostile to disruptive technologies or innovative business
models. While some industry commentators have cast doubt on
the sustainability of fully decentralized financial services, others
believe that DeFi has real potential as a disruptor of traditional
financial services markets.

1.1 DeFi from a technical perspective

the underlying infrastructure. DeFi applications as well as
digital assets reside on Layer-1-protocols (e.g., Ethereum). This
Layer-1-protocol is of crucial importance, as it represents the
execution and settlement layer for any transactions.

A basic understanding of the different layers of technology
used for DeFi applications will establish a mental map that is
helpful in analyzing and evaluating specific DeFi implementations
(Figure 1).

In addition to these core layers, three additional layers can
play a role. First, at the bottom of the stack, the interoperability
layer allows different settlement layers to directly communicate
with each other. It can be used to allow DeFi applications to
incorporate different Layer-1-protocols into their functionality. At
the top of the stack, an application layer normally provides user
interfaces. Finally, an aggregation layer allows to aggregate the
functionality of multiple DeFi applications.

Protocol, asset, and settlement layers form the core of the DeFi
technology stack. The protocol layer consists of DeFi applications
that offer some sort of financial service functionality such as
trading or lending. The asset layer defines which digital assets
can be processed by a DeFi protocol. It is important to keep in
mind that normally a specific DeFi protocol is offering its services
for only a few specific digital assets such as one fungible token
or a pair of fungible tokens. Finally, the settlement layer forms

AGGREGATION LAYER

AGGREGATOR 1

AGGREGATOR 2

AGGREGATOR 2

APPLICATION LAYER

PROTOCOL LAYER

ASSET LAYER

EXCHANGE

LENDING

NATIVE PROTOCOL ASSET
(E. G. ETH)

ASSET
MANAGEMENT

DERIVATIVES

FUNGIBLE TOKEN S
(E.G. ERC-20)

NON -FUNGIBLE TOKEN S
(E.G . ERC-721)

(ETHEREUM) BLOCKCHAI N

SETTLEMENT LAYER

(BITCOIN) BLOCKCHAIN
(…) BLOCKCHAIN

INTEROPER ABILITY LAYER

DIRECT COMMUNICATION BETWEEN DIFFERENT SETTLEMENT LAYERS

Figure 1: The DeFi Stack (based on IOSCO 2022 and Schär 2021)

…

1.2 The Promise of DeFi
Decentralized applications are the core building block of DeFi.
They make DeFi services fundamentally different to traditional
financial services: being built on public, permissionless blockchain
networks, DApps operate in a trustless environment. This means
that responsibility for the security of operations and assets need
not fall to a central intermediary, but is instead provided by the
underlying blockchain network. Furthermore, with assets stored
and transactions settled on a public blockchain infrastructure,
users are able to hold the cryptographic keys to their assets
directly without using a custodian.
It is important to note that the primary goal of DeFi remains to
facilitate existing financial services such as trading or lending.
While DeFi for the most part does not change the functionality
of the services, it does modify how the services are provided,
delivered and used. From a user’s perspective, this promises
various benefits, including:
•

Real-Time Execution and Settlement: Execution and
settlement is more tightly integrated in DeFi applications.
After executing a trade, the underlying blockchain is usually
automatically updated.

•

Transparency: Transactions are transparent for most DeFi
projects, and this ensures that users can check and audit
all activities.

•

Accessibility: Regardless of people’s location, they can
access DeFi services if they have an internet connection, a
crypto wallet, and a smartphone.

•

Disintermediation: Assets are normally directly held by
users in non-custodial wallets or in smart-contract-based
escrow accounts. This eliminates the need for expensive
intermediaries.

•

Programmability: DApps offer the opportunity to integrate
business logic into a system. This way, rules can be
automatically executed with no need for intermediaries or
human intervention.

•

Low Fees / High Interest Rates: Since two individuals
can transact directly, with no need for intermediaries,
transactions normally become less expensive.

However, it is important to also consider the risks involved.
Although DeFi solutions succeed in reducing some risks, such
as counterparty risk associated with trading partners and
intermediaries, they also add new risks such as software bugs or
more general smart contract risk (e.g., token contracts, domino
effects, dependency on off-chain data, etc.). Furthermore, from an
institutional perspective, regulatory risks have to be considered.
Important questions – such as concrete implementations for
AML or KYC policies – are still under discussion.

1.3 DeFi applications
DeFi solutions exist for major functions such trading, lending,
investing, deposits, and payments (see Figure 2) with more
services being added on an ongoing basis. Figure 2 gives
an overview of financial services verticals, with examples of
solutions across traditional finance (TradFi), Centralized Finance
(CeFi)1 and Decentralized Finance (DeFi).

When comparing TradFi with CeFi and DeFi, it is important to
distinguish infrastructures from assets. DeFi (and CeFi) solutions
currently focus on the processing of digital assets such as
cryptocurrencies, whereas TradFi handles traditional assets such
as bonds or equities. However, it would also be conceivable that
DeFi solutions process digitalized versions of traditional assets
(e.g., tokenized bonds).

Traditional Finance
(TradFi)

Centralized Finance
(CeFi)

Decentralized Finance
(DeFi)

T R A DI NG

Exchanges / Brokers
(e. g. Xetra)

Crypto Exchanges
(e. g. Binance)

Decentralized Exchange
(e. g.
)

L END ING

Secured and unsecured
(e. g. term loans)

Lending Platforms
(e. g. BlockFi)

Lending Protocols
(e. g.
)

INVESTING

Investment Funds
(e. g. ETFs)

Crypto Funds
(e. g. Grayscale)

Decentralized Asset Mgmt.
(e. g.
)

DEPOSITS

Savings Account
(e. g. Commercial Banks)

Staking Pool
(e. g. Coinbase)

dStaking Services
(e. g.
)

PAYMENTS

Payment Platforms
(e. g. SEPA, T2)

Centralized Stablecoins
(e. g. USDC)

DeFi Stablecoins
(e. g.
)

Figure 2: Main financial services categories in Traditional Finance, Centralized Finance and Decentralized Finance

1.

Centralized Finance refers to financial services for digital assets that are largely comparable to traditional financial services offerings from an organizational and procedural perspective.
However, CeFi also uses blockchain for the settlement of net positions.

As noted, DApps are currently available for users in financial
services such as trading, lending, investing, deposit, and
payment solutions.
•

Trading: In DeFi, decentralized exchanges (DEXs) perform
the function of centralized exchanges by using smart
contracts. DEXs enable users to exchange digital assets
without having to use or trust intermediaries or use
custodians. Major DEX protocols include Uniswap and
Sushiswap.

•

Lending: DeFi lending protocols offer loan services. These
solutions normally come in one of two varieties. With poolbased lending protocols interested individuals provide
liquidity or funds to a pool that others can borrow from.
Users putting their assets into the pool can earn interestlike income in return. With peer-to-peer based lending,
individuals borrow directly from a particular lender. In this
case, decentralized lending protocols enable borrowers to
take out loans with minimum barriers. Major DeFi lending
protocols include Aave, Maker and Compound.

•

Investing: DeFi DApps can also be used to execute
automated trading strategies. For example, TokenSets is
a DApp-based platform for portfolio management. Users
provide the boundary conditions and investment objectives
and then TokenSets trades, balances, and implements
strategies to achieve the users’ goals automatically. This
enables users to gain exposure to a basket of digital assets,
without the need to buy individual assets.

•

Deposits (Staking): While there are no deposits in the
traditional sense in the DeFi ecosystem, a very similar
mechanism – ‘staking’ – exists. Staking refers to the
process of locking up digital assets for a fee, and as such
is comparable to depositing money at a bank. Just as a
bank deposit represents a short-term loan that is given to
the bank, staked cryptocurrencies can be seen as a short
term loan to a protocol. For the time that assets are staked,

they earn a small income, but cannot be sold or otherwise
used by their owners. Rather, digital assets locked up in this
way are used for adjacent processes, such as supporting
the transaction validation mechanism of the underlying
blockchain network.
•

Payments and Stablecoins: The extreme volatility of
cryptocurrencies such as Bitcoin and Ether inhibits their
use for payment purposes. This problem is addressed
by stablecoins – digital assets that are pegged to fiat
currencies or some other stable asset. They aim to be a
means of payment with a similar volatility as fiat currencies.
As stablecoins are digital assets, they can be seamlessly
integrated into other DeFi applications. Users can quickly
conduct on-chain transactions with these coins without
the need for using traditional financial infrastructures.
Stablecoins come in two varieties. Asset backed stablecoins
are blockchain-based tokens that have their value pegged
to a reserve asset such as another digital currency, fiat
money or a commodity. In contrast, algorithmic stablecoins
try keep a stable value by managing the supply of the coin
based on demand from users. However, the recent collapse
of Terra (Luna) has shown that algorithmic stablecoins may
not always be able to retain their stable value making them
unsuitable as a means of payment.2 Also, the adequacy of
reserves of asset backed stablecoin projects have recently
been publicly challenged.3 Some of the most popular
stablecoins include Binance USD (BUSD), USD Coin (USDC),
and Dai (DAI). In additon to these private sector-driven
initatives, central banks around the world are looking into
the opportunities of Central Bank Digital Currencies (CBDC).

While the above examples represent the main solutions currently
observed in the DeFi space, there are a host of other DeFi financial
services that are being discussed and experimented with. These
include insurance services, derivatives, and prediction markets
(e.g., Augur).

2.

https://www.bloomberg.com/graphics/2022-crypto-luna-terra-stablecoin-explainer/

3.

https://www.wsj.com/articles/biggest-stablecoin-issuer-tether-switches-accounting-firm-to-bdo-italia-11660795062

2 . DECENTRALIZED EXCHANGES
One of the fastest growing areas in DeFi is Decentralized
Exchanges (DEX). With the rise of cryptocurrencies and other
digital assets after the financial crisis of 2008, people soon
searched for service offerings that enabled them to trade these
new assets.
With traditional financial services providers not offering services
for owners of digital assets, a new class of financial intermediaries
emerged, initially in the shape of ‘centralized exchanges’ (CEXs)
that replicate within the digital asset space the services provided
by traditional exchanges. In order to use the services of a CEX,
a users needs to sign up with the CEX and then, before being
able to buy or sell digital assets, has to fund their account using
either cryptocurrencies (e.g., Bitcoin) or some traditional form of
payment (e.g., bank transfer).
However, this also means giving up control over any assets that
are held by the CEX in a user’s account.4 Users need to trust
the exchange with their money, making them vulnerable to some

4.

degree of counterparty risk. This was especially problematic
given most CEXs were newly established entities with untested
operations and minimal to zero oversight from financial markets
authorities. As a results, hacks, scams, and other illegal activities
have been common in the early years of the CeFi space, often
leading to a loss of capital for users of these services.
Decentralized exchanges (DEX) emerged to solve this problem.
The primary goal of a DEX is full disintermediation via the
elimination of middlemen and allowing every user to deal directly
with other users on a P2P basis. By migrating trading functionality
directly onto the blockchain via smart contracts, a DEX serves
as a trustless platform for the trading of digital assets. Much
like traditional exchanges and CEX, these platforms coordinate
supply and demand from many users. However, assets either
remain in the custody of the user or – for a limited time – in an
escrow account of the fully automated smart contract.

Practically, the ownership of digital assets is represented by owning the private key to the address where the assets are stored. Sharing the private keys with third parties, enables these
parties to fully control the asset. “Not your keys, not your asset” is a familiar quote in the crypto community.

2.1 Market size, major players, and market share
While CEXs might dominate the digital asset trading space for
a long time, we are witnessing an increase in the popularity of
DEXs. Available reports indicate that by late 2021, the largest
DEXs have already started challenging some of the oldest CEXs
in terms of their trading volume.

Generally, the size of a DEX can be evaluated using two metrics:
trading volume and market capitalization. From a trading volume
perspective, Uniswap (V3) is the largest DEX, with an average
trading volume of close to $1.25 billion per day as of August
2022. Other large DEX include dYdX and Pancake swap (V2).
Figure 3 gives an overview of 24h trading volumes.
In comparative terms the 24h trading volume as on 15 August
2022 for Uniswap (V3) was $1.25 billion whereas that of Binance
(the largest CEX) on the same date was $17.52 billion.

TRADING VOLUME (24H) FOR TOP 5 DEXS
0€

500,000,000€

1,000,000,000€

1,500,000,000€

VOLUME
(24H)

Uniswap (V3)

dYdX

Kine Protocol

PancakeSwap (V2)

Figure 3: Trading Volume (24h) for top 5 DEXs5

5.

Data obtained from coinmarketcap.com on Aug. 15th, 2022

iZiSwap

2.2 Main Types of DEXs
DEXs come in a variety of forms. Over the years, several designs
have been suggested and implemented, all in an effort to improve
on previous attempts and further streamline the functionality of
the solution and its user experience. Generally, there are three
major types of DEX.
•

Automated Market Makers based DEXs

•

Order book based DEXs

•

Hybrid / Alternative Platforms

Automated Market Makers (AMMs) based DEXs: AMMs use
pools of digital assets (sourced from so-called ‘liquidity providers’
– see Filling the Pool in section 2.3.1) to enable trading services
for its users. Prices are quoted automatically by the underlying
smart contract, hence the name automated market maker. The
main purpose of creating an AMM is to ensure liquidity at all
times. Given the prevalence of the term AMM for AMM-based
DEXs, we use the terms interchangeably in the following.

6.

Order Book based DEXs: Order book DEXs compile the details of
all open buy and sell orders for a particular pair of digital assets.
A buy order implies that a trader desires to bid for an asset at a
given price. A sell order on the other hand is an indication that a
trader is willing to sell a specific asset at a particular price. Much
like traditional exchanges, order book DEXs match these orders.
Hybrid / Alternative Platforms: Though most DEXs can be
classified as either as AMM or order book, a growing number
of platforms are beginning to merge the concepts of both these
types to create new, hybrid DEXs. This is mostly done to enable
additional functionality such as allowing users to seamlessly
trade their assets across multiple blockchains etc. An interesting
example for such a hybrid approach is Raydium.
In addition, a noteworthy phenomenon has been the rise of DEX
aggregators, which allow users to search for prices and liquidity
across multiple DEXs.6 As the name implies, they aggregate
liquidity from different DEXs to provide the users with the best
execution price available within the shortest time, while lowering
the level of slippage on large orders and optimizing fees.

Strictly speaking DEX aggregators are not a type of DEX, therefore they are not covered in this report.

2.3 Automated Market Makers – the current
market leader
At present, the most common type of DEX is the Automated
Market Maker. The main feature of traditional exchanges is to
match supply and demand for a given asset and determine a
price at which parties are willing to exchange assets. Automated
Market Makers operate in a fundamentally different way.

The liquidity pool in its simplest form contains units of two digital
assets. This asset pair can either consist of two risky assets (e.g.,
Ether and Bitcoin) or a risky asset and a base asset such as a
stable coin (e.g., Ether and USDC). From a technical perspective,
the pool is an address on the Layer-1 settlement infrastructure
(e.g., Ethereum). The liquidity pool is controlled and managed
by a smart contract that automatically, based on the pool’s
composition, generates prices for which a trader can exchange
the assets.

An AMM does not match trades as in the case of a centralized
or decentralized order-book, but rather provides a liquidity pool
to trade against. The two major components of an AMM are its
liquidity pool and a smart contract that manages the liquidity pool
by determining trading prices as well as fees and rewards for the
service.

2.3.1 How do AMMs work
Trading via AAMs is broken into multiple phases (see Figure 4).

AUTOMATE D / CONSTANT FUNCT ION MARKET MAKER
Peer-to-protocol Transaction

PROTOCOL & SETT LEME NT LAYE R

Trader

Order (illust.)

3

•

Input Asset

•

Output Asset

•

Input Asset
Quantity

Liquidity
provision /
withdrawal

Trader
submits order

Trader’s wallet

Liquidity
Provider
1

4
LP’s receive
transaction fees
(Liquidity reward)
Liquidity Pool (Smart Contract)
Token A

Swap / Settlement 5

Token B

Liquidity Shares

Ethereum Blockchain
Figure 4: AMM – Process overview (numbers represent sequence of process steps)

Token C

2
Receives /
Returns
‘liquidity
shares’ / LP
Tokens

Filling the pool: Initially the digital assets in a liquidity pool are
provided by liquidity providers, including long-term holders of
digital assets who provide their assets to the pool in order to
earn income on otherwise non-earning assets. In exchange for
bringing in their assets, the liquidity provider receives pool shares
(also known as liquidity provider shares) in the form of special
‘pool tokens’. These tokens represent a claim of the liquidity
provider on the portfolio of assets in the pool and can be used to
withdraw the liquidity provided. In addition to entitling the holder
to a share in the pool assets, pool tokens also give the holder a
claim on any trading fees earned by the pool.7
Trading assets: Once funds are available in the liquidity pool,
traders can send their orders. An order consists of the input and
the output asset as well as the quantity of one of the assets
(e.g., 100 Ether against Bitcoin). On receiving the order, the AMM
automatically quotes a price at which the trade can be instantly
executed.

Determining the price: Prices are automatically determined
by the AMM for traders who wish to trade the digital assets in
the pool. The AMM generates its quotes according to a predetermined formula (so-called conservation function or bonding
curve). The conservation function prices the two pool assets
relative to each other, making the asset that is in high demand
automatically costlier, and vice versa.
One common function to determine prices is the constant
product function. This function keeps the product of the amount
of the two assets in the pool constant (i.e., amount of token X *
amount of token Y = constant). Mathematically, such a function
determines a curve containing all ratios at which an AMM is
willing to exchange the assets of the pool (see Figure 5).

Asset Y

Constant product function represents
all combinations of assets Y and X,
where Y1*X1 = Y2*X2 = Yi*Xi. All
rectangles under the curve have the
same area.

Y 1 X1

Y 2 X2

Y 3X3
Y 4X4
Y 5 X5

Ass et X
Figure 5: Constant product bonding curve

7.

With some DEX-applications, a liquidity provider might also receive so-called “protocol tokens” which grant the right to vote on decisions concerning the governance of the protocol (e.g., fee
structure, code updates etc.).

While constant product functions are one of the most frequently
used functions to determine prices for the assets in a pool,
they are by no means the only one. Other functions that can be
found implemented or discussed include Constant Mean Market
Makers, Constant Sum Market Makers, Hybrid Function Market
Makers and, Dynamic Automated Market Makers.

100/100 = 1) but only 0.90909 units of B. This price impact
of trades (so-called ‘slippage’, in this case equal to -0.0909)
depends on the size of trade relative to the size of the pool. The
bigger the trade relative to the total size of the pool the higher the
slippage. Figure 6 gives an exemplary illustration of the interplay
between conservation functions and slippage.

Example: If the liquidity pool AB consists of 100 units of digital
asset A and 100 units of digital asset B, the product A*B is
10,000. If someone wanted to sell 10 units of A (against B) he
would supply 10 A to the pool and receive an amount of 9.0909
units of B. This keeps the product constant at 10,000 (110 A *
90.909 B = 10,000).

Arbitrage Opportunity: As can also be seen from the example
above, the price quoted by the AMM reflects the ratio of the
assets in the pool. Every trade changes this ratio. If the slippage
is large enough, this could result in a price deviation from a fair
ratio of exchange between the two assets. If the assets in the
pool are also traded on other markets, this in turn could lead
to potential arbitrage opportunities. An arbitrageur would take
advantage of this price difference and bring the price ratio back
to its fair value.

As can be seen from this example of a constant product function,
the price that the trader receives is not 1 unit of B for 1 unit of
A (as implied by the ratio of the quantities of the two assets,

Asset Y

In order to keep X*Y constant, the trade
is realized at (y+dY) / (x-dX). This move
along the constant product function
creates slippage for a trader.

Constant
pro duct
f unct i o n

P realized = Y R X R
dY
P c ur rent = Y c X c

dX

Asset X

Figure 6: Constant product function and slippage (simplified, based on Mohan 2022)

2.3.2 Benefits of AMMs

the liquidity pool. This eliminates counterparty risk which has
been an issue with lightly regulated centralized crypto exchanges.

AMMs offer several benefits to traders as well as liquidity
providers.
For traders, the benefits are instant execution and continuous
liquidity for otherwise illiquid assets. Another benefit lies in the
operational setup. In contrast to centralized exchanges, traders
do not have to store their digital assets with the exchange –
instead, trades are instantly and automatically executed against

POOL ASSET 1

For liquidity providers, the main benefit lies in generating
income for otherwise non-earning assets. From this perspective,
providing liquidity to a DEX can be compared to traditional
securities lending of long-term portfolio positions. This can be
quite profitable as the below example shows.

POOL ASSET 2

POOL
TOTAL

NOTE

ETH

$-Value

USDC

$-Value

$ Value

Pool Total

9

$9,000

9,000

$9,000

$18,000

Constant Product
(9*9k= 81k)

Liquidity
provision

+1

+ $1,000

+1,000

+ $1,000

+$2,000

Liquidity Provider owns
10% of Pool = $2,000

Pool Total

10

$10,000

10,000

$10,000

$20,000

Constant Product
(10*10k = 100k)

BUY 1 ETH
vs. USDC

(10-1)

Pool after
Trade

9

$9,000

10,000+x

Constant Product must
stay at 100k

11,111

Solving for X; Constant
Product = 100k, Liquidity
Providers owns 10% of
Pool = $2011,1

$11,111

$20,111

Table 1: Example for trade via an AAM (Assumptions: 1 USD Coin (USDC) = 1$, 1 ether (ETH) = $1,000; example based on BIS 2021 p34)

2.3.3 Risks and drawbacks of using AMMs
Alongside the benefits of an AMM, it is also important to take an
detailed look at the potential risks and drawbacks for traders as
well as liquidity providers.
For traders, the main disadvantage of using an AMM is the
potential for high slippage (see Figure 6). As AMM prices follow a
curve, there will always be slippage, and that slippage depends
on the conservation function of the AMM. The main factors
driving slippage are the size of the trade relative to the pool size,
as well as the current composition of the pool.
A second risk traders face is frontrunning. In order for a trade to
settle it needs to be recorded on the underlying blockchain (e.g.,
Ethereum). Generally, the Layer-1-blockchain of a DeFi protocol
stores transactions in a public transaction pool. Validators /
miners then group these transactions into blocks, which are
subsequently written onto the blockchain. Once written onto the
blockchain, a trade is executed and settled at the same time.
While this process seems pretty straight-forward, validators /
miners have large discretion with regards to the order in which
they process transactions from the transaction pool.
In a largely unregulated market, this opens up the risk of
frontrunning, which exists even if validators / miners automatically
group transactions according to predefined rules. By structuring
transactions smartly, malicious third-party traders might be able
to sandwich an order (buy and sell order before and after the
order of the AMM) in order to profit from a price impact.

Thirdly, traders also face some uncertainty with regards to the
total cost of their trade. While direct fees of the DEX are known
and comparatively small, settling a trade requires recording the
trade on the underlying blockchain. This normally also incurs
fees. Most Layer-1 protocols used for DeFi have dynamic fee
structures based on the demand for the protocol’s services (e.g.,
Ethereum). This can lead to prohibitively high settlement fees.
The extent of this risk is protocol dependent.
Lastly, from a trading perspective AMMs have very limited
functionality when it comes to permissible order-types, as
normally only market orders are supported.
For liquidity providers, the main risk for providing liquidity to an
AMM lies in the possibility of a so-called ‘divergency loss’. As
described above, liquidity providers do not have a claim on the
exact amounts of digital assets they have initially provided to
the pool but rather claim on a share of the pool’s total assets
at a given time plus any fees earned by trades (see the above
example in the Table).
If the price of an asset pair on external markets moves away
from the price quoted by the AMM, this opens the opportunity
for arbitrageurs to trade against the liquidity pool and profit
from these price differences. This can lead to situations where
the value of the liquidity provider’s share in the pool is worth
less than the value of the original positions provided to the
pool. In other words, it would have been better for the liquidity
provider not to have provided the liquidity. This loss is sometimes
euphemistically referred to as ‘impermanent loss’, as it only is
realized when a liquidity provider chooses to exit the pool.

2.4 Beyond Automated Market Makers

However, over time new approaches and new Layer-1 protocols
have provided new impetus to projects working on blockchainbased order book DEXs. Several approaches can be identified.

Currently, the majority of DEXs are based on the AMM model.
This is noteworthy because – as discussed above – that model
has several drawbacks when compared to an order book model.
However, the major reason for the dominance of AMMs lies in
the difficulties associated with implementing an order book on a
blockchain infrastructure.

DEXs with off-chain orderbooks: In order to combine the
benefits of limit-order-book markets with the advantages of
DEXs, one solution is to move the order book off-chain.
The 0x protocol is one example for such a solution. In this setup, a party wishing to place an order sends a signed order to
a ‘relayer’ who enters the order into the order book. This way,
a party looking for liquidity can check their order against the
order book. On finding an appropriate matching order, the relayer
forwards the order to the DEX. The smart contracts of the DEX
then verify the validity of the order and execute the asset transfer
(Figure 7).

Maintaining a central limit order book on-chain means that every
quote needs to be recorded as a transaction on the underlying
blockchain settlement layer. This requires the underlying
infrastructure to be able to processes transactions in both a fast
and cheap way. Both these requirements were not met by early
Layer-1-protocols (e.g., Ethereum). This made a normal quoting
process – and hence maintaining an order book – prohibitively
expensive and practically infeasible.

APPLICATION LAYER

0X PROTOCOL (DEX) WITH OFF-CHAIN ORDER BOOK
Message format for broadcast orders

3.a

Trader A

2
Trader A creates and signs
order with private key
3.b
B) Alternative: broadcasts order over any arbitrary communication medium

A) Trader A
transmits
signed order to
Relayer

3.1
Relayer checks order validity and
posts it to order book

PROTOCOL & SETTLEMENT LAYER

Relayer

3.2
Trader B receive/view updated
version of order book
Order book

Trader B

5
Trader B submits makers’ signed order to
DEX
Trader B Account

Trader A Account
1

Token A

Trader A approves DEX contract

DEX

Trader B approves DEX contract 4

6
DEX contract verifies Trader A’s signature, that order has
not expired / been filled then transfers tokens at specified exchange rate

Figure 7: DEX with off-chain order book – Process overview (numbers represent sequence of process steps)

Token B

DEXs with on-chain order books: DEXs with on-chain order
books look to combine the advantages of limit order books
(e.g., order functionality, slippage) and a secure blockchain
infrastructure for the order book (e.g., resistance to censorship,
no need for central intermediaries). However, this requires every
order and every quote to be recorded on the blockchain.

In addition to 0x, DEXs with such a set-up include IDEX and
EtherDelta.
The off-chain order-book approach has two main advantages.
First, it leads to higher performance when compared to purely
blockchain-based order books, as the underlying Layer-1
protocol ceases to be a limiting factor. Furthermore, it normally
also leads to lower overall costs, as no fees for the use of layer 1
must be paid in the quotation process.

The requirement to record every action as a transaction on the
underlying blockchain-infrastructure made this type of DEX very
hard to implement on early Layer-1 protocols such as Ethereum.
As discussed above, transaction recording on these Layer-1
protocols was simply too expensive and too slow. However,
newer Layer-1-protocols (e.g., Solana) enable fast and low-cost
execution of transactions on the underlying blockchain, therefore
enabling decentralized order books as a pricing mechanism.

On the downside, the approach introduces a centralized element
in the whole structure. While the settlement infrastructure is
still based on blockchain, the order matching is done off-chain.
This represents a single point of failure, as the centralized order
book can be compromised. In other words, this solution requires
trust in the central relayer to act properly on behalf of its users.
However, in contrast to a CEX, the relayer does not take custody
of the asset nor executes orders.

DEXs using an on-chain order book approach include for example
Stellar and Serum.

STE LLAR PROTOCOL WIT H ON-CHAIN ORDER BOOK
1

1

‘Manage Buy Offer’
Operation (bid)

‘Manage Sell Offer’
Operation (ask)

Trader A

Trader B

PROTOCOL & SETTLEMENT LAYER

2
P ara mete r

Ty pe

De s cription

Selling

Asset

Asset the offer creator is selling.

Buying

Asset

Asset the offer creator is buying.

Amount

Integer

Amount of buying being bought. Set to 0 if
you want to delete an existing offer.

Price

{base asset,
quote asset}

For example, if you wanted to buy 30 XLM
and sell 5 BTC, the price would be {5,30}

Offer ID

Integer

The ID of the offer. 0 for new offer. Set to
existing offer ID to update or delete.

Relayer checks order validity and
posts it to order book

3.b
Order book

3.a

Not marketable orders saved on
order book until either consumed
or cancelled

Matching (“crossing”) orders are filled/
executed and settled automatically

Trader A’s wallet

Trader B’s wallet

Stellar network
Figure 8: DEX with on-chain order book – Process overview (numbers represent sequence of process steps)

3 . CO NCLUSION AND OUTLOOK
DEXs in their various forms are still a niche phenomenon in the
emerging digital asset space. Therefore, it would be easy to
dismiss them as largely irrelevant. However, from a bigger picture
perspective, several trends support the growth potential of DeFi
in general and DEXs in particular.

Third, DeFi and DEXs should profit from the ongoing drive to
digitalize traditional securities and assets. Such fully digitalized
assets will require new processes and infrastructures, which
DeFi is already offering today, such as the integration of the
securities and the cash leg within one infrastructure.

First, DeFi has the potential to be a key component of the ongoing
trend of improving financial services processes. At its core,
DeFi promises to make financial services more efficient mainly
through automation and by cutting out middlemen – objectives
achieved by building on the specific properties of blockchain
technology. This could enable, for example, a tighter integration
and automation of execution and settlement processes.

Finally, the technology and solutions underlying DeFi are
developing rapidly. New Layer-1 blockchains such as the
emergent Solana ecosystem or the Binance Smart Chain, as
well as innovative Layer-2 solutions (e.g., Polygon) promise
better performance at a lower price while building on the key
advantages of a decentralized system. This in turn creates a
basis for newer and better DeFi solutions – the DEX space in
particular has benefitted from constant work on major protocols,
resulting in tremendous improvements to the available solutions.

Second, while the crypto asset space has emerged out of the
crypto community, in recent years we have seen increasing
interest in the space across broader target groups. Not only have
institutional investors become more interested over the past few
years but established financial service providers have started
initiatives and offerings for digital assets. We are starting to see a
similar development in the DeFi-sector. While most DeFi-solutions
are driven by the technology affine crypto community, there are
first examples of regulated financial institutions becoming active.
The engagement comes in a variety of forms including stable
coin offerings (e.g., by JP Morgan8) and active participation in
DeFi-protocols by established institutions (e.g., Societe Generale
in MakerDAO9) to name two examples. Given the development
of firm regulation of digital assets in various jurisdictions (e.g.,
Europe or Switzerland), there is an expectation that DeFi will
persist and mature in the coming years.

8.

https://www.jpmorgan.com/onyx/coin-system.htm

9.

https://www.thedefiant.io/makerdao-members-vote-on-issuing-30m-loan-to-societe-generale

While new disruptive solutions come with risks, they also offer
opportunities. With regards to DEXs, traditional financial service
providers can identify opportunities at various levels of the
technology stack. Naturally, their expertise means trading-oriented
financial institutions have an opportunity to become active players
in DEXs as liquidity providers, traders or arbitrageurs. However,
banks could also become more deeply involved by offering their
own DEX or DEX-based services and solutions, built on their
strong expertise in financial markets. Lastly, opportunities exist
to become active in providing infrastructure services for the
settlement layer of DeFi (e.g., staking).

How do I become active in the space?

•

We are currently seeing a major market correction in the digital
assets space. Such corrections should be expected with new
technologies – and as with previous financial crises, such
corrections typically lead to a shakeout of weaker companies
and solutions. They also present long-term participants with an
opportunity to learn from their own and others’ mistakes.

Implementation plan: Typically new DLT components will
have to be integrated with existing internal systems. In most
cases, four components will be required. Depending on the
positioning and the new service offering, these include:
–

a custody solution

–

an ‘intermediate layer’ (custody integration layer) that
serves as an interface between the standardized DLT
applications and legacy systems

–

specific trading software

–

an anti-money laundering solution

When considering engaging with DEXs, market participants
should consider the following aspects:
•

•

•

Customer segment: A commercial bank with a strong retail
service offering might be inclined to offer its customers
secure and easy-to-use access to DEXs while a provider of
traditional exchange services might focus more on creating
infrastructure solutions.
Regulatory requirements: The integration of existing
DEX solutions into your own offering can have a potentially
negative impact on your standing with regulators. With
supervisory authorities globally casting an eye on DeFi, it is
important to follow regulatory developments closely.
DEX business case: Due to the dynamic nature of DeFi,
such a business case should ideally consider numerous
branching scenarios.

•

Build or buy: Whether components are built or bought,
the know-how and capabilities for operating the DLT
components should be established internally to ensure both
independence and sustainability.

Capco and ABC Research support the individual implementation
path end-to-end with comprehensive technical and
methodological knowledge of capital markets as well as DLT
expertise, from digital asset strategy to the implementation of
new business models.

REFERENCES
•

21Shares Research Team. (2022). State of Crypto. Our Insights into Web3, The Future of The Internet. [Industry Report].
https://21shares.com/research

•

Aramonte, S., Huang, W., & Schrimpf, A. (2021). DeFi risks and the decentralization illusion. BIS Quarterly Review, December
2021, pp 21-36.

•

Aspris, A., Foley, S., Svec, J., & Wang, L. (2021). Decentralized exchanges: The “wild west” of cryptocurrency trading.
International Review of Financial Analysis, 77, 101845. https://doi.org/10.1016/j.irfa.2021.101845

•

Barbon, A., & Ranaldo, A. (2021). On The Quality Of Cryptocurrency Markets: Centralized Versus Decentralized Exchanges
(arXiv:2112.07386). arXiv. http://arxiv.org/abs/2112.07386

•

Capponi, A., & Jia, R. (2021). The Adoption of Blockchain-based Decentralized Exchanges (arXiv:2103.08842). arXiv. http://arxiv.
org/abs/2103.08842

•

ConsenSys. (2021). DeFi for Institutions (Insight Report) [Industry Report].

•

Deshmukh, S., Warren, S., & Werbach, K. (Eds.). (2021). Decentralized Finance (DeFi) Policy-Maker Toolkit. In collaboration with
the Wharton Blockchain and Digital Asset Project [White Paper]. World Economic Forum.

•

FATF. (2021). Updated Guidance for a Risk-Based Approach to Virtual Assets and Virtual Asset Service Providers. FATF. www.
fatf-gafi.org/publications/fatfrecommendations/documents/Updated-Guidance-RBA-VA-VASP.html

•

Fidelity Digital Assets (2021). The Institutional Investor Digital Asset Survey. September 2021. Avaliable at: https://www.
fidelitydigitalassets.com/research-and-insights/2021-institutional-investor-digital-assets-study

•

Han, J., Huang, S., & Zhong, Z. (2021). Trust in DeFi: An Empirical Study of the Decentralized Exchange. SSRN Electronic Journal.
https://doi.org/10.2139/ssrn.3896461

•

IOSCO. (2022). Decentralized Finance Report (OR01/2022; p. 45pp). The Board of the International Organization of Securities
Commissions.

•

Kitzler, S., Victor, F., Saggese, P., & Haslhofer, B. (2021). Disentangling Decentralized Finance (DeFi) Compositions
(arXiv:2111.11933 [cs]; arXiv:2111.11933). arXiv. http://arxiv.org/abs/2111.11933

•

Lehar, A., & Parlour, C. A. (2022). Decentralized Exchanges. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.3905316

•

Meyer, E., Welpe, I. M., & Sandner, P. (2021). Decentralized Finance—A systematic literature review and research directions.
SSRN Electronic Journal. https://doi.org/10.2139/ssrn.4016497

D E FI EX PLAI N E D – T H E CAS E O F D EC E N T R A LIZ E D EXC H A N G ES / 21

RE FE R E NC E S
•

Mohan, V. (2022). Automated market makers and decentralized exchanges: A DeFi primer. Financial Innovation, 8(1), 20. https://
doi.org/10.1186/s40854-021-00314-5

•

Moncada, R., Ferro, E., Favenza, A., & Freni, P. (2021). Next Generation Blockchain-Based Financial Services. In B. Balis, D.
B. Heras, L. Antonelli, A. Bracciali, T. Gruber, J. Hyun-Wook, M. Kuhn, S. L. Scott, D. Unat, & R. Wyrzykowski (Eds.), EuroPar 2020: Parallel Processing Workshops (Vol. 12480, pp. 30–41). Springer International Publishing. https://link.springer.
com/10.1007/978-3-030-71593-9_3

•

Paruch, K. (2012). SoK: Decentralized Exchanges (DEX) with Automated Market Maker (AMM) protocols [Working Paper].

•

Quinlan, B., Chung, J., & Trehan, E. (2021). Cracking the Code. The Evolution of Digital Assets to the Mainstream. [Industry
Report]. Quinlan & Associates.

•

Schär, F. (2021). Decentralized Finance: On Blockchain- and Smart Contract-Based Financial Markets. Federal Reserve Bank of
St. Louis REVIEW, 103(2), 153–174. https://doi.org/10.20955/r.103.153-74

•

Wang, Y., Chen, Y., Wu, H., Zhou, L., Deng, S., & Wattenhofer, R. (2022). Cyclic Arbitrage in Decentralized Exchanges
(arXiv:2105.02784). arXiv. http://arxiv.org/abs/2105.02784

•

Werner, S. M., Perez, D., Gudgeon, L., Klages-Mundt, A., Harz, D., & Knottenbelt, W. J. (2021). SoK: Decentralized Finance (DeFi)
(arXiv:2101.08778). arXiv. http://arxiv.org/abs/2101.08778

•

Wharton Blockchain & Digital Asset Project. (2021). DeFi Beyond the Hype [Working Paper]. The Wharton School, The University
of Pennsylvania.

•

Zhou, L., Qin, K., & Gervais, A. (2021). A2MM: Mitigating Frontrunning, Transaction Reordering and Consensus Instability in
Decentralized Exchanges (arXiv:2106.07371). arXiv. http://arxiv.org/abs/2106.07371

AUTHORS
Vinzenz Treytl, ABC Research
Jan-Michael Steiner, Capco
Arindam Bhaumik, Capco
Gerald Hessenberger, Capco

CONTACTS
Alexander Eisl, ABC Research, alexander.eisl@abc-research.at
Christoph Ruth, Capco, christoph.ruth@capco.com

ABOUT CAPC O
Capco, a Wipro company, is a global technology and management consultancy dedicated to the
financial services industry. Our professionals combine innovative thinking with unrivalled industry
knowledge to offer our clients consulting expertise, complex technology and package integration,
transformation delivery, and managed services, to move their organizations forward.
Through our collaborative and efficient approach, we help our clients successfully innovate,
increase revenue, manage risk and regulatory change, reduce costs, and enhance controls. We
specialize primarily in banking, capital markets, wealth and investment management, finance,
risk & compliance, and insurance. We also have an energy consulting practice in the US. We serve
our clients from offices in leading financial centers across the Americas, Europe, and Asia Pacific.

ABOUT AB C R E SEAR C H
The Austrian Blockchain Center (ABC), located in Vienna, is a COMET competence center with the
mission to be Austria’s first scientific contact point for Blockchain and related technologies.
Blockchain is a technology for secure cooperation between different participants with a wide
range of use cases, not only as a digital currency, but also in industry, finance, energy, logistics,
and public administration.
ABC is an interdisciplinary and application-oriented research institution dedicated to all aspects
of blockchain research. Technological, economic, and legal topics are the focus. Projects with a
high practical relevance, which directly lead to innovations in the economy, are made possible by
the experts of the ABC and its scientific partners – Austrian and international universities,
universities of applied sciences and research institutions.