A Short Survey on Business Models of

Decentralized Finance (DeFi) Protocols

Teng Andrea Xu1 and Jiahua Xu2

1
École Polytechnique Fédérale de Lausanne, Lausanne 1005, Switzerland,
arXiv:2202.07742v1 [econ.GN] 15 Feb 2022

[email protected]
2
University College London, UK, [email protected]

Abstract. Decentralized Finance (DeFi) services are moving traditional

financial operations to the Internet of Value (IOV) by exploiting smart
contracts, distributed ledgers, and clever heterogeneous transactions among
different protocols. The exponential increase of the Total Value Locked
(TVL) in DeFi foreshadows a bright future for automated money trans-
fers in a plethora of services. In this short survey paper, we describe the
business model for different DeFi domains - namely, Protocols for Loan-
able Funds (PLFs), Decentralized Exchanges (DEXs), and Yield Aggre-
gators. We claim that the current state of the literature is still unclear
how to value thousands of different competitors (tokens) in DeFi. With
this work, we abstract the general business model for different DeFi do-
mains and compare them. Finally, we provide open research challenges
that will involve heterogeneous domains such as economics, finance, and
computer science.

Keywords: Decentralized Finance · Value Investing · Blockchain

1 Introduction

Decentralized Finance (DeFi) aims to provide financial services on a blockchain-

based infrastructure. A plethora of cryptocurrencies form the DeFi ecosystem.
These tokens are able to replicate classical financial services, such as lend-
ing/borrowing, without any central institution, by exploiting smart contracts
and the immutable ledger. In the literature [1,2], DeFi’s key features are gener-
ally recognized to be open to anyone, transparent, non-custodial, and composable,
i.e. financial services can be arbitrarily composed to make new financial prod-
ucts. The Total Value Locked (TVL)3 has seen exponential growth with the
so-called “DeFi Summer”. The TVL grew from $600m by the end of March 2020
to $11bn by the end of September 2020 [3]. At the time of writing4 , DeFi TVL
is around $98bn. Similarly, from March 2020 to the time of writing, the top-100
DeFi tokens’ market cap grew by almost hundred times - from $1,8bn to $154bn
3
The sum of all assets deposited in DeFi protocols, and therefore locked in a smart
contract.
4
2021-10-31, https://defipulse.com/
2 T. Xu et al.

[4]. We show, in Figure 1, the top Ethereum DeFi tokens’ daily revenue over
time, smoothed by a rolling window of 30 days, for each DeFi financial service,
i.e. Uniswap (UNI) for DEXs, Yearn Finance (YFI) for Yield Aggregators, and
AAVE for PLFs. These recent DeFi milestones foreshadow a bright future for
both DeFi’s users and investors.

(a) (b) (c)

Fig. 1. The figure shows the top DeFi’s protocols by revenue for each financial service.
On the left, Figure 1a shows UNI daily revenue across time, smoothed by a rolling
window of 30 days. Uniswap is currently the biggest DeFi protocol by daily revenue.
Its treasury has more than $9bn in assets [5]. Similarly, Figure 4 and Figure 1c show,
respectivelyt, YFI and AAVE daily revenue, smoothed by a rolling window of 30 days.
We can clearly see how these protocols have built a steady and reliable cash flow in
less than one year. UNI and AAVE data are retrieved from CryptoFees API [6], while
YFI data is retrieved from YFIstats [7]

To date, there is a lack of literature that offers any clear abstraction on how
DeFi protocols generate their revenue stream, a key component for the sustain-
ability of a project. We claim that it is important for both investors and users
to understand how DeFi’s tokens profit. From an investor’s perspective, a clear
business model with a steady and constant revenue stream are key features be-
fore investing in the underlying project. As an end user, DeFi’s users look for
reliable protocols; hence, a protocol with an efficient and observable business
model is likely to be a “secure” protocol. Therefore, the central contribution of
this short survey paper is to describe and offer a clean comparison of different
DeFi services business models. In this work, we will look into the main DeFi
financial services namely: Protocols for Loanable Funds (PLFs), Decentralized
Exchanges (DEXs), and Yield Aggregators. We won’t analyze the variety of pro-
tocols in technical detail, but rather we will direct the reader to other resources.
The focus of this work is mainly on the protocols’ business model. The paper
is structured as follows. First, we describe the general PLFs business model in
Section 2. Subsequently, we explain the dominant cash flows within DEXs and
Yield Aggregators in Section 3 and 4. In Section 6, we present a first generalized
business model in DeFi. Finally, we provide a literature review and conclude the
work with Section 6 and 7.
 Title Suppressed Due to Excessive Length 3

2 Protocols for Loanable Funds (PLF)

PLFs let users borrow/lend digital assets in a decentralized fashion. Automated
smart contracts behave as middle-men. They lock assets deposited by the lender
and allow borrowers to get liquidity in exchange for collateral. These types of
smart contracts are also called Lending Pools [8]. These Lending Pools typically
lock a pair of tokens, a loanable token, and a collateral token. By providing liq-
uidity, lenders gain interest rates depending on the supply & demand. Because
there is no guarantee of paying back, Borrowers must over-collateralize their
position. On top of that, when returning the amount borrowed, the borrowers
must pay an interest rate that is split pro-rata among the lenders and the gov-
ernance token. Moreover, when borrowers get liquidated, they will have to pay
an additional fee. We show in Figure 2 a typical generalized PLF use case.

Fig. 2. The figure above abstracts and generalizes the lending protocol framework by
showing the main actors and interactions. From left to right, Lenders can deposit their
crypto-assets, Ethereum in this case, to gain additional profits. They receive a PLFs
wrapped token or IOU as proof of their deposit. In the center, the smart contract acts
as a middle-man. It takes care of the deposited assets, loans, and liquidations - if any.
On the right, a borrower must deposit collateral before getting the loan. Finally, at the
end of the loan, the borrower will have to return the borrowed amount plus an interest
rate, part of this interest rate will split pro-rate among all lenders, and the rest will
generate revenue for the PLF itself.

Business Model - PLFs cash flow depends on the interest rate model, the
current underlying demand & supply, and the total amount borrowed. The in-
terest rate model can be either a linear model, a non-linear model, or a kinked
model [10]. Demand & supply increases the interest rate when demand is high,
and supply is low; vice versa, interest decreases. Given the rate of interest, the
PLF gets a percentage of it. For example, Compound takes 10% of the interest
4 T. Xu et al.

rate [11]. Besides traditional over-collateralized loans, PLFs offer flash loans5
that can bring more revenues to the protocol. Flash loans interest rate is usually
fixed, e.g. AAVE [12], or even without fees, e.g. dYdX [13].

3 Decentralized Exchanges

DEXs take distance from the classical order-book exchanges where traders match
market bids and/or asks. Again, smart contracts are the middle-men and, in this
case, are called Liquidity Pools. Investors or, in this scenario, Liquidity Providers
(LPs) can deposit a pair of equal worth tokens, say ETH/DAI as shown in Figure
3, into these Liquidity Pools. In exchange, they will receive LP tokens as proof
for their deposit and earn a percentage of the fee accrued to the Buyer when
swapping. A price is assigned for each token given the protocol’s price function -
usually determined by the constant rate formula [14]. The Buyer that is willing to
exchange - or “swap” - DAI for some ETH will deposit DAI in the Liquidity Pool,
plus some interest, and receive ETH. The whole mechanism is called Automatic
Market Making (AMM). Further reading on the topic can be found in [15] [16].
We show in Figure 3 a typical generalized DEX use case.

Fig. 3. The figure above abstracts and generalizes DEXs protocols framework by show-
ing the main actors’ interactions and the protocol revenue stream. From left to right,
LPs deposit a pair of tokens, in this example DAI/ETH, in the Liquidity Pool. In
exchange, they receive LP tokens as proof of their deposit. In the middle, a smart
contract takes care of locked assets, new deposits, swaps, and fees. The Buyer, willing
to exchange his DAI for some ETH, will have to pay a fee. This fee will be partially
distributed pro-rata among all LPs, while the DEX treasury will collect a percentage
of it.
5
Flash Loans are a special type of loan where the borrower must return the borrowed
amount plus interest in the same transaction without the need for collateral. An
in-deep explanation, analysis, and exploitation can be found in [9].
 Title Suppressed Due to Excessive Length 5

Business Model - When Buyers swap, they pay a fee. This fee is split pro-rata
between the liquidity providers of the pool as a reward for their contribution to
the pool. A percentage of the interest rate is sent to the protocol’s treasury. This
share of the fee represents the primary income resource for most of AMMs, such
as Uniswap or Balancer that have variable swap fees [17] [18], or Bancor that
has a fixed interest rate [19]. To date, Uniswap has the biggest treasury in DeFi
with $9bn locked in its treasury [20].

4 Yield Aggregators
Yield Aggregators combine different strategies to maximize investors’ rate of re-
turn. Similar to PLFs and DEXs protocols, smart contracts have a central role.
Commonly, smart contracts are referred to as ”Vaults” in this domain. In this
scenario, shown in Figure 4, investors deposit their savings into a Vault. Differ-
ent Vaults run different strategies. These strategies can be straightforward, such
as finding the best lending protocol interest rate, or more complex, as borrow-
ing assets and leveraging some other position by exploiting different protocols
(Compound, Uniswap, Aave, to cite some), for example. There are a plethora of
solutions. To have a more in-deep technical insight refer to [21].

Fig. 4. This figure shows a typical yield aggregator use case. Starting from the top left,
the Investor chooses his Vault of preference to deposit his savings. The Vault, a smart
contract, will run its pre-set strategy; simple staking, lending, or providing liquidity are
just examples. More complex strategies combine borrowing and/or leveraging involving
multiple steps and protocols. The yETH vault is an example of a multiple-step strategy
[22]. Usually, Vaults apply a fixed performance fee on the strategy yield.

Business Model - Yield Aggregators cash flow is based on their Vaults per-
formance. That is, yield aggregators charge a commission fee on the strategy’s
 6 T. Xu et al.

profit. Hence, the investor yield will be equivalent to the Vault’s total profit
minus the protocol’s fee. Different tokens apply different interest rates: Yearn
Finance v2 applies 20% as performance fee and an additional 2% as management
fee [23], Pickle Finance and Idle have 20% and 10% performances fees respec-
tively [24] [25]. Harvest is the only yield aggregator that applies 30% of fees but
uses the whole reward to buy back FARM tokens (Harvest native tokens) from
DEXs and re-distribute them to FARM stakers [26].

5 DeFi Business Model

We now synthesize the business models reported and give a first general DeFi
business model framework. This framework involves different actors and actions
with their naming conventions.

Fig. 5. DeFi common mechanism and revenue strategy.

Table 1. An overview of DeFi’s naming taxonomy.

DeFi Protocol Smart Contract Investor User Financial Service

PLFs Lending Pool Lender Borrower Loan
DEXs Liquidity Pool Liquidity Provider Buyer/Trader Exchange
Yield Aggregators Vault Vault User - Asset Management

DeFi Protocol with multiple facets - Lending, AMM, or Yield Aggregator - provides open,
non-custodial, permissionless, and composable financial services in exchange
for a small fee. The fee is accrued to any asset movement, for example,
borrowing assets, swapping assets, or movements that yield profits.
 Title Suppressed Due to Excessive Length 7

Investor this actor is willing to hold the underlying protocol risk, such as protocol
misbehavior, impermanent loss, or rug-pulls, in exchange for a passive in-
come. The Investor mainly deposits his assets and provides liquidity to the
financial service. Usually, this figure has a passive role: He waits until the
return rate satisfies the risk held. The Investor is also known as Lender,
Liquidity Provider, or Vault User.
User Different from the Investor, the User usually exploits the protocol on the
fly, and he never waits for any long-term response. This actor moves asset -
borrowing or swapping -, and thus pays interest rates to the protocol. This
figure has an active role - he uses the protocol, and he expects an immediate
response. The User is also known as Borrower, Buyer, or Trader.
Financial Service a smart contract that lets the Investor and the User interact indirectly. This
actor is the core of the whole protocol: Locks the Investor assets, satisfies the
User requests, and prevents protocol’s misuse. Furthermore, it can behave
as a User by leveraging other DeFi Protocols. Finally, it delivers yields and
earnings to the Investor and DeFi Protocol. The smart contract in DeFi
is usually known as Lending Pool, Liquidity Pool, and Vault. In the latter
form, the smart contracts move assets to yield interest, thus paying some
interest to the protocol.

DeFi’s services don’t come for free. As we have seen, DeFi multiplies how
investors can yield passive income. However, this additional interest is somehow
”taxed” by the DeFi protocol. On the other side, the users are willing to use
the platform in exchange for a small fee. Hence, the DeFi protocol has a small
income from both sides. In classical finance, this market and business model is
known as the ”Two-Sided Markets,” first formalized by Jean-Charles Rochet and
Jean Tirole in [27]. Undeniably, the service couldn’t exist without both parties.
On one side, the investor provides liquidity to the financial service that peer
users can use. On the other side, by paying fees, the user provides income to
both protocols and investors.

6 Literature Review

Our work is strongly related to the canonical empirical asset pricing problem.
Evaluating and pricing an asset is perhaps the most renowned problem in clas-
sical finance literature. There exist two cardinal analyses to value and price an
asset. On the one hand, technical analysis deduces future underlying value from
its history of trading, which can involve price change, trading volume, price mov-
ing average, and other historical characteristics [28]. [29] shows how 90% of chief
foreign exchange dealers based in London in November 1988 exploited technical
analysis for their trades. We observe here a duality between firm shares and
cryptocurrencies trading. On the other hand, fundamental analysis is based on a
firm’s book value (assets and liabilities), stream of dividends, current earnings,
future investment opportunities, etc. [30] studies the role of dividends policies
in a firm’s share evaluation. The paper shows how firm’s value can be explained
 8 T. Xu et al.

by its current earnings, the growth rate of earnings, the internal rate of return,
and the market rate of return. Remarkable findings in the series of work [31],
[32], and [33] show how firm book value and market equity size are important to
explain future company’s return. Our work focused more on fundamental analy-
sis. Value investing [34] dominated financial decisions for years. Do high-valuable
tokens perform better than low-valuable cryptocurrencies? Finally, we would like
to stress the parallelism among firms’ share ownership and DAO tokens. Both
of them give shareholders voting power on the underlying future moves. We will
address and expand on these open research challenges in the next Section 7.

7 Conclusion
In this short survey paper, we have synthesized DeFi’s main business models. We
claim that it is important for DeFi’s investors and users to understand which
protocol has a reliable cash flow. Moreover, the literature is missing a clear
overview of the protocols’ business model to date. The main contributions of
this short survey paper of our work are as follows. First, we have provided a
clear understanding and explanation of the most important DeFi’s services busi-
ness model. Furthermore, we have generalized and synthesized a novel general
framework business model adopted by DeFi’s protocol. On top of that, the sci-
entific community can address multiple open research challenges:

Value Investing In classical finance, a firm has a high value when it has a high book-to-market
ratio. In the literature, it is shown that, by smartly investing in value stocks,
the strategy could increase the mean return of the portfolio [34]. Conducting
a parallel study on crypto-assets is a straightforward application of Value
Investing.
Voting Power How similar DAO tokens providing the same financial service with the same
business model are priced differently. These tokens grant the holder vot-
ing power on the underlying protocol improvement/change. However, DAO
tokens don’t have an initial value, but they acquire it by exchanging and
trading. For example, how is Aave and Compound’s DAO tokens value per-
ceived?
Regulatory Issues recently, the crypto-currencies ecosystem has seen big turbulence. While,
China has completely banned all crypto transaction [35], India is working on
plans to enforce similar regulations [36]. Finally, the US is willing to regulate
the market [37]. We claim that the literature lacks a deep analysis that
evaluates DeFi’s business models suitability with the current regulations.

DeFi’s financial services have seen massive growth since the “DeFi Summer”.
While some tokens achieved a stable cash flow (shown in Figure 1), many others
have been subject to cyber-security breaches. In 2021 only, the whole DeFi world
has suffered almost $1bn loss due to hacks [38]. The most recent hack is dated
Dec. 1st, 2021. BadgerDAO, a service that allows Bitcoin to be used as collateral
over different DeFis protocols, suffered a $120m loss [39]. Therefore, albeit the
exponential growth in DeFi’s users and revenue stream, DeFi is yet a risky
 Title Suppressed Due to Excessive Length 9

infrastructure that has to mature over time. Whether DeFi will replace or co-
exist with the classical financial services and infrastructure is unclear yet, leaving
us with open research challenges to unveil.

References
1. Werner, S. M., Perez, D., Gudgeon, L., Klages-Mundt, A., Harz, D., & Knottenbelt,
W. J. (2021). Sok: Decentralized finance (defi). arXiv preprint arXiv:2101.08778.
2. Schär, F. (2021). Decentralized finance: On blockchain-and smart contract-based
financial markets. FRB of St. Louis Review.
3. Cousaert, S., Xu, J., & Matsui, T. (2021). SoK: Yield Aggregators in DeFi. arXiv
preprint arXiv:2105.13891.
4. CoinGecko. (2021). Coingecko, Top 100 DeFi Coins by Market Capitalization,
https://www.coingecko.com/en/categories/decentralized-finance-defi.
5. Mihal, D.: OpenOrgs (2021), https://openorgs.info/, accessed: 2021-12-03.
6. Mihal, D.: CryptoFees (2021), https://cryptofees.info/.
7. Finance, Y.: YearnFinance (2022) https://www.yfistats.com/financials/RevProjections.html,
accesses: 2022-02-07.
8. Qin, K., Zhou, L., Gamito, P., Jovanovic, P., & Gervais, A. (2021). An Empiri-
cal Study of DeFi Liquidations: Incentives, Risks, and Instabilities. arXiv preprint
arXiv:2106.06389.
9. Qin, K., Zhou, L., Livshits, B., & Gervais, A. (2021, March). Attacking the defi
ecosystem with flash loans for fun and profit. In International Conference on Finan-
cial Cryptography and Data Security (pp. 3-32). Springer, Berlin, Heidelberg.
10. Gudgeon, L., Werner, S., Perez, D., & Knottenbelt, W. J. (2020, October). Defi
protocols for loanable funds: Interest rates, liquidity and market efficiency. In Pro-
ceedings of the 2nd ACM Conference on Advances in Financial Technologies (pp.
92-112).
11. Young, M., Alexandre, A.: DeFi Deep Dive: Yield Farm Pioneer Compound
Finance (2021-03-08), https://beincrypto.com/defi-deep-dive-yield-farm-pioneer-
compound/, accessed: 2021-11-29.
12. AAVE: AAVE Flash Loans (2021), https://docs.aave.com/devtelopers/guides/flash-
loans, accessed: 2021-11-29.
13. EthereumDev: Making a flash loan with Solidity (Aave, Dy/Dx, Kollat-
eral (2020-05-10), https://ethereumdev.io/making-a-flash-loan-with-solidity-aave-
dy-dx-kollateral/, accessed: 2021-11-29.
14. Buterin, V.: Improving front running resistance of x*y=k market makers (2021-
03-18), https://ethresear.ch/t/improving-front-running-resistance-of-x-y-k-market-
makers/1281, accessed: 2021-11-30.
15. Xu, J., Vavryk, N., Paruch, K., & Cousaert, S. (2021). SoK: Decentralized Ex-
changes (DEX) with Automated Market Maker (AMM) protocols. arXiv preprint
arXiv:2103.12732.
16. Lin, L. X., Budish, E., Cong, L. W., He, Z., Bergquist, J. H., Panesir, M. S., ...
& Zhang, S. (2019). Deconstructing decentralized exchanges. Stanford Journal of
Blockchain Law & Policy, 2.
17. Uniswap: Uniswap v3 Protocol Whitepaper (2021),
https://uniswap.org/whitepaper-v3.pdf, accessed: 2021-11-30.
18. Balancer: Fees (2021), https://docs.balancer.fi/getting-started/faqs/fees, accessed:
2021-11-30.
10 T. Xu et al.

19. The Serenity Fund: Bancor 2.1 and its impermanent loss insurance (2021-03),
https://medium.com/coinmonks/company-watch-bancor-2-1-and-its-impermanent-
loss-insurance-71d87d092e9f, accessed: 2021-11-30.
20. Mihal, D.: DAOs are the new companies. What’s on their balance sheets? (2021),
https://openorgs.info/, accessed: 2021-11-30.
21. Cousaert, S., Xu, J., & Matsui, T. (2021). SoK: Yield Aggregators in DeFi. arXiv
preprint arXiv:2105.13891.
22. Cronje, A.: yETH vault explained (2020-09-04), https://medium.com/iearn/yeth-
vault-explained-c29d6b93a371, accessed: 2021-11-30.
23. Yearn: YIP-51: Set Vault v2 fee structure (2020-11),
https://gov.yearn.finance/t/yip-51-set-vault-v2-fee-structure/7752, accessed:
2021-11-30.
24. Pickle: Fees (2021), https://docs.pickle.finance/fees, accessed: 2021-11-30.
25. Idle: Fees (2021), https://docs.idle.finance/advanced/fees, accessed: 2021-11-30.
26. Harvest Finance: Profit Sharing and iFARM (2021), https://harvest-
finance.gitbook.io/harvest-finance/general-info/what-do-we-do/profit-sharing-
pool-ps, accessed: 2021-11-30.
27. Rochet, J. C., & Tirole, J. (2003). Platform competition in two-sided markets.
Journal of the european economic association, 1(4), 990-1029.
28. Edwards, R. D., Magee, J., & Bassetti, W. H. C. Technical Analysis of Stock
Trends, —. A Professional Journal Published by The International Federation of
Technical Analysts, 75.
29. Taylor, M. P., & Allen, H. (1992). The use of technical analysis in the foreign
exchange market. Journal of international Money and Finance, 11(3), 304-314.
30. Miller, M. H., & Modigliani, F. (1961). Dividend policy, growth, and the valuation
of shares. the Journal of Business, 34(4), 411-433.
31. Fama, E. F., & French, K. R. (1993). Common risk factors in the returns on stocks
and bonds. Journal of Financial Economics, 33(1), 3-56.
32. Fama, E. F., & French, K. R. (1992). The Cross-Section of Expected Stock Returns.
Journal of Finance, 427-465.
33. Fama, E. F., & French, K. R. (1995). Size and book-to-market factors in earnings
and returns. The journal of finance, 50(1), 131-155.
34. Piotroski, J. D. (2000). Value investing: The use of historical financial statement
information to separate winners from losers. Journal of Accounting Research, 1-41.
35. BBC News: China declares all crypto-currency transactions illegal (2021-09-24),
https://www.bbc.com/news/technology-58678907, accessed: 2021-12-03.
36. BBC News: Indian government set to ban cryptocurrencies (2021-11-24),
https://www.bbc.com/news/technology-59402310, accessed: 2021-12-03.
37. Haar, R, Time: U.S. Officials Send Mixed Messages on Crypto
Regulation. Here’s What It All Means for Investors (2021-11-02),
https://time.com/nextadvisor/investing/cryptocurrency/crypto-regulation-talks-
heat-up/, accessed: 2021-12-03.
38. Cointelegraph: Cointelegraph Consulting: Recounting 2021’s biggest DeFi hacking
incidents (2021-11-03), https://cointelegraph.com/news/cointelegraph-consulting-
recounting-2021-s-biggest-defi-hacking-incidents, accessed: 2021-12-03.
39. Thurman, A, CoinDesk: Badger DAO Protocol Suffers $120M Exploit (2021-12-
02), https://www.coindesk.com/business/2021/12/02/badger-dao-protocol-suffers-
10m-exploit/, accessed: 2021-12-03.