Thoughts on the incentive model of trading platforms: which is more worthy of incentive, liquidity or trading behavior?
The "easy business victories" brought by genius ideas will become increasingly rare in Web3.Author: Mint Ventures
Introduction
This content comes from an online discussion within the Mint Ventures team, during which I raised two questions. Our researcher Lawrence and investment manager Scarlett shared their thoughts, combined with my reflections and extensions on the topic, forming the main content of this clip.
The two questions I raised are:
Should the token incentives of trading platforms encourage liquidity or trading volume? Why?
Assuming the platform has a trading incentive mechanism, what are the pros and cons of wash trading for the protocol? Why?
*It should be noted that the term "trading platform" in the above two questions is generic, including spot Dexs (Uniswap, Curve), decentralized derivatives projects (Gmx, Gains), NFT trading platforms like Blur, and even centralized exchanges (although most of them currently do not have direct liquidity or trading mining mechanisms).
The following article reflects the author's and the team's interim views as of publication, which may contain factual errors and biases, and is for discussion purposes only.
Origin: Why Do Trading Platforms Issue Governance Tokens and Provide Incentives?
In my view, there are three main reasons why trading platform projects (including spot, NFT, and derivatives) issue governance tokens:
Token fundraising. This is straightforward and will not be elaborated on.
Decentralized governance. The emergence and distribution of governance tokens are prerequisites for community governance.
Growth and economic system adjustment mechanisms.
However, from the perspective of Web3 business practices, the first two points are not mandatory options because:
Projects can engage in equity financing.
Centralized governance remains the mainstream for enterprise and project operations, and even most projects that have begun to decentralize governance still have governance content and voting concentrated in the hands of the team and related institutions.
What may be truly important is the third point: once a project has a token, its resources for growth incentives and the ability to adjust its economic system are greatly enhanced compared to projects without tokens.
For trading platform projects, the core purposes of token incentives (whether to incentivize liquidity or trading behavior) generally boil down to two:
New projects achieve cold starts by attracting the initial bilateral users through token incentives (Uniswap also implemented token incentives during the early stages of the Dex wars).
Expand the bilateral market and expect to accelerate the cross-side network effects through token subsidies, creating barriers against competitors.
The term "bilateral" here refers to the liquidity providers on one side and the traders on the other for most trading platforms. Since the platform's two sides are interdependent and mutually beneficial: if no one provides liquidity, trading cannot occur; and without trading, there are no fee returns, so no one will want to provide liquidity. Conversely, the richer the liquidity (under the same product mechanism), the lower the trading friction, the more willing traders are to trade here, leading to more fees, and the more fees, the more liquidity returns, resulting in abundant liquidity.
This situation, where "the growth of one side's users provides greater value to the other side's users," is a type of network effect, specifically "cross-side network effect." Network effects are one of the most important moats for business projects.
This is especially true for new projects.
If we say that when Uniswap was first created, it could rely solely on the natural demand of early users due to a lack of competition (the value of DeFi had not yet reached consensus), then in today's crowded trading platform space, where product iterations are rapid, it has become exceedingly difficult for a bilateral/multi-sided trading platform to successfully cold start without providing incentives (or expectations of airdrops).
Thus, purpose 1 (cold start) is a prerequisite for purpose 2 (cross-side network effect), and the ultimate goal of 2 is to form a competitive advantage (monopoly position), based on which to achieve and expand profitability (protocol revenue > token subsidy expenses).
Given the existence of cross-side network effects, whether a trading platform incentivizes liquidity providers or traders, it seems that both can indirectly incentivize users on the other side. Does this mean:
Incentivizing LP = Incentivizing Trader?
Practice: Liquidity Incentives vs. Trading Incentives
Early Practices of CEXs
In fact, whether incentivizing liquidity or trading, this has existed long before DeFi Summer. For example, most centralized exchanges have incentive programs for cooperative market makers, aimed at ensuring they provide good trading depth on their platforms, allowing other traders to enjoy low-slippage trading experiences. This is similar to liquidity incentives, but these incentives are mostly provided through fee reductions and refunds, largely unrelated to the platform's own governance tokens.
The earliest experiment with trading incentives/mining was also initiated by centralized exchanges, with Dragonex (龙网) being the first to implement a mechanism in 2018 that rewarded platform tokens (DT) based on trading volume, establishing a "trading mining" mechanism based on platform token dividends from trading fees.
This mechanism was later further "promoted" by Fcoin. After the launch of trading mining, both exchanges experienced rapid business growth. Particularly, Fcoin, as a latecomer in the "exchange wars," had a peak daily trading volume that surpassed the sum of all other leading CEXs.
However, neither Dragonex nor Fcoin's glory lasted long. The former collapsed due to theft and insolvency, while the latter went bankrupt due to a so-called "internal financial error" leading to massive losses. The reasons for the downfall of the two platforms differ slightly, but the commonality is financial insolvency; the former was due to external attacks, while the latter was due to internal issues.
But we must ask: if there had been no theft incidents and internal financial problems, could trading mining have helped them stabilize their market share in trading volume and create barriers?
Later Practices of DeFi Projects
The emergence of Uniswap pushed the AMM mechanism further into the market, significantly lowering the barriers for liquidity provision and providing natural convenience for liquidity mining incentives. Sushiswap quickly siphoned off a large amount of liquidity from Uniswap through its own token subsidies to LPs, leading Uniswap to launch its liquidity mining program in response.
In addition to Dexs, NFT trading platforms like Looksrare, X2Y2, and Blur have also engaged in extensive practices of liquidity incentives, shifting the subsidy battlefield from fungible tokens (FT) to the NFT space. It was from Blur that Opensea felt a significant threat and began to implement zero fees in response.
On the other hand, perhaps due to the harsh failure of Fcoin still echoing in memory, significant trading mining practices in the DeFi space did not occur until DyDx launched its token. Subsequent practitioners included the Dex project Cherry on Okchain and Dinosauregg on BSC, among others. More recent projects incorporating trading incentives include level, gridex, kwenta, etc. However, overall, projects primarily incentivizing trading remain few.
Liquidity Incentives vs. Trading Incentives
As previously mentioned, the purpose of token incentives for trading platforms is to more quickly establish a lead in cross-side network effects and form a monopoly advantage. One of the prerequisites for monopoly is that "users find it difficult or are unwilling to leave the existing platform," which in the context of internet product operations relates to the metric of retention rate.
Thus, whether trading platforms should incentivize liquidity or trading, an important consideration is "which behavior or behavioral object" is more likely to retain users long-term after the incentives diminish? Users and behaviors that are more likely to retain long-term can contribute higher "lifetime value" (LTV) to the platform, and they should be the primary targets of the platform's token incentives.
So, whose behavioral habits are more entrenched, liquidity providers or traders? Our current thinking leads us to believe it is liquidity providers, for several reasons:
LP users are more easily bound to the platform compared to traders.
For example, projects like Curve guide LP users to enhance their market-making returns by staking CRV tokens through the ve mechanism, which increases the migration costs for LPs and aligns their interests with the platform.
LPs have more concerns and willingness regarding product migration.
Traders have a transient financial interaction with the platform, with minimal risk as no funds are entrusted to the protocol; whereas LPs authorize their funds to the platform's smart contracts, which carries greater risk. Therefore, they tend to choose platforms they are familiar with, have a good safety record, and have a long history for providing liquidity. Trying new platforms involves higher psychological risks, and even if the APR is higher, they are often reluctant to migrate.
With the existence of aggregators, trading behavior is more rational, adhering to the principle of minimizing trading friction.
While liquidity provision also has aggregators (putting funds into liquidity pools or yield aggregators), the process of investing through aggregators introduces an additional layer of smart contract fund authorization risk, and aggregators often take a higher share of returns (in contrast, trading aggregators usually do not charge explicit fees). People tend to prefer to provide liquidity directly, especially larger funds that hardly need to consider the gas friction caused by yield reinvestment.
Incentives for trading behavior dissipate due to trading friction.
The frequency of trading is much higher than that of market-making, and the friction generated during trading, such as GAS (including MEV) and NFT trading royalties, is taken away by third parties outside the protocol and users, effectively dissipating part of the incentives.
Incentives for trading may "inflate trading demand."
This leads to behaviors like "trading to obtain incentives," which may seem to contribute to trading fees and protocol revenue, but may not significantly help achieve the project's incentive goals, namely "gaining advantages through token subsidies to realize cross-side network effects," because these "inflated trading behaviors" will immediately disappear once the incentives stop, failing to help the protocol achieve "a larger market share in trading."
Overall, LP users have a closer and more enduring relationship with trading platforms compared to purely trading users. Incentives directed at the LP side may yield higher long-term total user value. Perhaps this is why most trading platforms still primarily focus on liquidity incentives, with relatively few explorations into trading incentives.
Outlook: How to Better Incentivize Trading?
Although in most cases, choosing between liquidity and trading incentives, incentivizing liquidity is often the more prudent choice, exploring incentives for trading behavior still holds significant value because the ultimate revenue of trading platforms comes from trading. Incentives for liquidity are also aimed at providing traders with a better trading experience to achieve higher trading volumes and fees. When designing trading incentives, paying attention to the following points may yield better results:
Do not directly incentivize trading behavior or trading volume, but incentivize those liquidity pools or trading pairs with higher trading volumes.
A typical case is the concept proposed by AC when designing the ve(3,3) project Solidly, where holders of ve governance tokens can only earn trading fees from the pools they vote for, rather than receiving fees from the entire protocol regardless of which pool they vote for.
This will guide ve token holders to vote more for those pools (trading pairs) that have high trading volumes and contribute more fees to the protocol, allowing these pools to receive more liquidity guidance (increasing the token emission incentives for that pool), thereby achieving better depth and further gaining more trading volume and fees, effectively "indirectly incentivizing trading."
Currently, in addition to Solidly's fork projects, some other projects have also begun to experiment with such incentive methods. For example, Pendle has adopted the approach of "distributing 80% of pool fees to those ve users who vote for the pool," meaning ve users will be more willing to vote for pools with high trading volumes, increasing incentives for LPs.
Control the value of incentives to avoid "inflated" trading volumes.
As long as the value of trading subsidies is lower than the friction costs of trading, people are less likely to engage in "trading for incentives" behaviors. Instead, they will rationally choose platforms with trading subsidies when they have actual trading needs, conducting those "trading behaviors they would have undertaken anyway." Over time, users may tend to trade more on incentivized platforms under similar conditions, and once the behavior solidifies into a habit, the purpose of the subsidies will be achieved.
Transform trading subsidies from "token distribution mechanisms" into "operational activities."
Token distribution mechanisms are long-term and require careful adjustments, while operational activities are short-term and flexible. Trading platforms can design short-term trading activities to activate key behaviors of new users (for example, defining a user's first trade on the platform as "activation") or to achieve specific short-term business goals through activities like trading competitions (for instance, to align with fundraising or combat competitors).
For example, the Gains clone project Vela on Arbitrum has adopted short cycles and continuously adjusted reward mechanisms and amounts in its trading incentive activities to attract users from platforms like Gains and Gmx.
The above design directions for trading incentives also align with the retention rate thinking mentioned earlier, focusing on real user behaviors that can lead to long-term retention for more refined incentive designs.
Conclusion: In the Low-Barrier Web3 Business World, Operational Efficiency is the Key to Long-Term Development
As Web3 business development progresses, one important observation I have made is that monopolies seem harder to establish here. This is determined by the inherent account systems, permissionless flow of funds and project creation, open-source transparent code, and composability of the Web3 world. Building moats in such an environment is challenging because users become harder to "foster and monopolize."
For any project to survive and thrive in this "low-barrier" environment, it must continuously strive for operational efficiency. Here, "operation" is a broad term encompassing product innovation, marketing activities, team management, and more, with accurate and efficient incentive design being one of the important tasks.
It is foreseeable that the distribution and incentive models of tokens will increasingly need to iterate with changes in competition and user demands, becoming a long-term endeavor. The notion of a "static economic model" designed from a god's-eye view may ultimately be a delusion for most projects.
Therefore, as investors, when selecting long-term investment targets, the project team's qualities of "long-term combat capability," "sustained business ambition," and "willingness to actively respond to market changes" should be key considerations.
Genius ideas leading to "easy business victories" will become increasingly rare in Web3.
