Moats of the DeFi protocol: to what extent are they forked?

Original title: "moats of DeFi protocol: to what extent are they forked?"

Blue fox notes

Introduction: in the common chain ecosystem, the development of DeFi is the most remarkable at present. So how defensible is the DeFi protocol itself? Can they be bifurcated? How does the composability and interoperability of the DeFi protocol affect the defensibility of ethereum itself? Based on the analysis of the current DeFi's main projects, this paper puts forward the defensible ranking of the current DeFi protocol, and considers that USDT and Maker are the most defensible among them, one is the strongest liquidity in actual combat, the other is due to the liquidity and external availability of synthetic stability currency DAI. Defensible in the middle are Tornado and Synthetix, one because of the network effect of their privacy pool itself, and one because of the bridge threshold between real assets and synthetic assets. However, the defensibility of the lending market and the trading market is not so strong, mainly through the subsidy of funds and the mutual utilization of capital pools, can lead to a certain amount of liquidity, and can not achieve an absolute advantage, which also leads to the emergence of competitors in these two fields. The original title of this article was "talking about forked DeFi protocol" by Kyle Samani, translated by "JT" in the blue fox notes community.

With a bunch of new layer 1 blockchains coming out, I've been thinking about the network effects of ethereum and the defensibility of the DeFi protocol built on ethereum.

A few years ago, I wrote about the network effects of non-sovereign Layer 1 currencies such as bitcoin and ethereum. Since then, the ethereum-based DeFi ecosystem has flourished: by taking advantage of a set of DeFi protocols, users can extract hundreds of millions of dollars in debt, secured by a larger pool of capital.

Together, these agreements now drive hundreds of millions of dollars in economic activity, so it is possible to reason about their defensibility. One way is to quantify and compare network effects. However, because the underlying dynamics of each protocol are unique, it is difficult to quantify network effects and thus difficult to compare them one by one.

In this article, I will consider:
1) fork to pay the effort
2) the amount of money you have to fork out

Then I rank the relative intensity of these network effects. Finally, discuss the level of defensibility of the ethereum ecosystem.

Synthetic stability currency - Maker

About a year ago, I wrote about how Layer 2 assets such as MKR (an equity token for the MakerDAO system) could capture value without permission or openness. In particular, the existence of "unbifurcated state" is the key to its value capture. In the case of Maker, it is clear that the non-bifurcating state is the collateral (mainly ETH) that sustains DAI's lending.(blue fox note: is this the main part of Maker that is defensible? Can't other agreements use ETH as collateral? Is the widespread use of DAI in different external protocols the real source of network effects? Welcome to explore here)

However, it is now clear that this framework is not complete. To understand this, let's assume that the only source of the Maker network effect is collateral. Wealthy third parties could fork out all of the Maker contracts, create a clone Maker ecosystem and deposit tens of millions of dollars in collateral to channel liquidity. But, so what? Shanzhai Maker's system is useless if no one wants to buy or interact with shanzhai DAI.

The most powerful source of Maker's defensiveness is not MKR, or the collateral backing DAI's loan, but DAI's liquidity and availability. DAI must be fluid in order to be usable. If someone takes DAI debt (with ETH as collateral), DAI is useless without liquidity. However, availability is a superset of liquidity. DAI's availability comes from the apparent adoption by merchants, from other agreements (such as Augur), and from the use of Compound, lendf.me, as collateral. DAI has been plugged into a variety of third-party applications, services, and infrastructure, which makes it more useful and available.

DAI's combination of mobility and availability is a powerful moat. The well-funded Maker copycat could offer a higher DSR and pay third parties for the integration of shanzhai DAI, but it is not clear whether it will gain meaningful traction.

The stable currency of fiat money collateral - USDT

Although Tether is not a pure DeFi protocol, its market value is over $5 billion, which I include because it is an integral part of the cryptographic ecosystem.Blue fox note: as of press time, USDT has a market value of more than $6.3 billion and has been accelerating. Blue fox notes predicts it's only a matter of time before USDT moves into the top three.)

The defensibility of the USDT is clear: it is the most liquid asset in the cryptography ecosystem, comparable to BTC. It is used by all the major non-us exchanges, is used as collateral on many derivatives exchanges and accounts for a large proportion of the clearing of OTC transactions.

Despite fierce competition from USDC, PAX, TUSD, GUSD and DAI, USDT still accounts for more than 80% of the stable currency market (by market value). This is the ultimate proof that the USDT is defensible.

Several teams are working on the stable currency clearing house, including the DeFi protocol (e.g. StableCoinSwap, Shell); And central clearing houses like Stablehouse. If they succeed, and thus reduce the friction of stable currency transactions, then this has a negative impact on the USDT.

For example, if these agreements/companies provide strong guarantees that a large number of stable currency conversions can be done with minimal slip-points, derivatives exchanges may start to use other stable currencies as collateral. FTX now offers this service locally; However, if the existence of a liquidity-stabilized currency clearing house may accelerate this trend, it may not be good for the USDT.

Mortgage money market -- Compound/ lendf.me

The Compound nonforking state is its collateral. So the Compound defensibility can be understood as: as the pool of collateral increases in value, the borrower can lend more, which attracts more lenders, and so on.

How difficult is it to guide the flow between forking and Compound?

There are several ways to do this. The Compound team can:

& have spent , & have spent & have spentAssets that support Compound unsupported assets (such as USDT)
& have spent , & have spent & have spentIntroduce more favourable mortgage rates and settlement penalties
& have spent , & have spent & have spentKeeping its own assets in pools and lending them to users can offer more competitive or even lower interest rates
& have spent , & have spent & have spentSubsidize third party lenders to lower interest rates.

Today, Compound's collateral is less than $100 million. If the team were to compete with the Compound by subsidizing users, say, on the order of 100 basis points a year, the annualized opportunity cost of channeling liquidity would be less than $1 million. This magnitude of risk is investable.

But in addition to Compound internal liquidity (in the form of borrowing rates), it is subject to several unique forms of external liquidity that may provide additional defensibility.

First, there are third-party aggregators, such as Instadapp, Zerion, RAY, idle.finance, Aave, and so on. These systems route deposits to the Compound, which in turn lowers borrowing rates and attracts more borrowers. While natural capital flows are certainly good, it is not clear that they matter at the margin (since the Compound team can subsidise interest rates to guide liquidity). (Blue fox notes: here not only has the interest-rate subsidy problem, there is also the question of balance, for lenders and borrowers, what is the most appropriate interest rate, it may be hard to lead to dominance, is not only a Compound is difficult to have powerful defensive, and fake Compound project is also facing the same problem, from this Angle, the market can accommodate at least several fake Compound borrowing agreement project, and each one can alone)

Interestingly, the existence of aggregators can be counterproductive, as aggregators have an incentive to send user assets to the highest-yielding lending pools. Assuming that contracts, governance, and predictive mechanisms are similarly reliable, aggregators will not be loyal to Compound at the expense of users, so Compound counterfeiters can easily win over aggregators with subsidies. Moreover, a large enough aggregator can siphon liquidity from the Compound into its own pool or Compound copycat. Although this has not happened, I expect it to happen in the next few years.

In general, it is not clear whether third-party aggregators will be a significant source of defensibility.

Second, let's consider cToken. CToken is somewhat similar to DAI. If third-party applications integrate cToken
This makes cToken more usable outside of the Compound core protocol, for example, as collateral. This makes it difficult for the lender, the cToken holder, to transfer from the Compound to another Compound.

While there are certain analogies between Maker/DAI and Compound/cToken, it's not a perfect analogy: the only reason to create DAI is to sell it and buy something else (like more ETH). Therefore, copycat makers are of little use unless there is a market for copycat DAI.Blue fox note: the implication here is that even if it can be done well, it will not provide more functionality, due to its off-contract traction. However, things would be different if other collateral could be pledged, such as Kava against BTC, ATOM, BNB, etc., which opens up other possibilities. In a sense, this is similar to the relationship between Compound and lendf.me.)

For the Compound, however, this is not the case. Even if third-party applications don't use cToken, the Compound is still useful.Note: cToken does not depend on external scenes to survive. DAI needs external scene requirements to survive.

Empirically, this is what one would expect. The dForce community from China has split the Compound code base and launched the mortgage money market agreement lendf.me, and in just a few months they've introduced $20 million in collateral into their system.

They did this by:

* provide Compound unsupported products (e.g. USDT, imBTC, HBTC)
* integrate with local third parties for Chinese users

It follows that the dForce community does not have to subsidize lendf.me products to achieve this. Just after the bifurcation Compound lendf.me attack today, there is still a long way to go.

The Maker is more defensive than the Compound. With a subsidy budget, anyone can Compound and channel the internal liquidity of the lending markets. But successfully branching Maker requires more than a subsidy budget: it also requires liquidity from DAI, and availability of DAI outside the agreement.

General synthetic asset agreement -- Synthetix

Synthetix is a specific type of exchange focused on trading synthetic assets. The defensibility of an exchange is usually its liquidity. However, Synthetix is not a traditional exchange because it does not offer a centralized limit order book (CLOB), as nearly all exchanges in the markets and cryptographic markets do.

One of Synthetix's most important features is that Taker does not cause any slip-points when trading the synthetic assets in the mortgage pool, but the liquidity is also limited, as it is subject to the amount of mortgage assets in the system. This means that liquidity, which is defensible, is primarily a function of the assets that are available as collateral.

Interestingly, the Synthetix exchange's growth is actually constrained by taker's need to tap into the Synthetix ecosystem by trading its real assets (such as ETH) into synthetic assets (such as sETH). Today, most users enter the Synthetix ecosystem through Uniswap, where the largest liquidity pool is sETH/ETH. So while the need for a liquidity bridge limits growth, it is in turn a moat: if someone tries to fork off Synthetix, a similar bridge needs to be channelled first.(note: in other words, there is a certain threshold for having a large liquidity pool of real/synthetic assets like the one on Uniswap.)

How do Synthetix's network effects compare to those of Maker and Compound?

First, let's consider the collateral in the agreement. Much like a knockoff Maker or a knockoff Compound, anyone who splits Synthetix can capitalize his or her own pool of assets or subsidize others to do so. As a result, mortgage assets are unlikely to be defensible.

Second, let's consider exotic assets: DAI, cToken, and synthetic assets. Unlike Maker's DAI, synthetic assets do not require liquidity outside the protocol... Instead, synthetic assets are more comparable to Compound cTokens: like cTokens, synthetic assets can be used as collateral for third-party applications without making the protocol work. Although this could be a defensible source, it is not.

The main form of defensibility is real assets < - & gt; Bridge of synthetic assets. While Synthetix is using Uniswap to do this today, teams that copy Synthetix can easily offer their own real assets through Uniswap, Kyber, or other free DeFi tools. - & gt; Bridge of synthetic assets.(note to blue fox: this is also essentially done through assets.)

Automatic market maker (AMM) -- Uniswap, StableCoinSwap, Shell, Bancor, FutureSwap, and Kyber

Compound is an automatic market maker, although it is used for borrowing/lending rather than trading. As a result, most transaction-centric AMM's defensiveness can be interpreted as equivalent to Compound, except for cToken.

Empirically, this is true. While not all of these amms are in direct competition due to different product priorities (e.g. StableCoinSwap and Shell focus on stable currency trading, while FutureSwap focuses on futures), the defensibility of each project is essentially dependent on the amount of liquidity in each protocol.

Although the larger liquidity pool in the Compound allowed for tighter lending rates, the larger liquidity pool provided taker with a lower slip point in the trading-centric AMM.

Kyber became the most liquid AMM in the past 12 months mainly because:

1) access to other AMM liquidity pools, such as Uniswap
2) the taker order flow will be routed due to third-party integration

Clearly, over time, all amms will tap each other's liquidity pools.

Paradoxically, once all amms in a particular vertical (such as stable currency conversions) enter each other's liquidity pool, all these amms become perfectly replaceable. None of AMM can compete in distribution.(blue fox note: the reason is simple, because any expansion in distribution will also make wedding clothes for others.)

Unmanaged central limit book trading platform - dYdx, IDEX, NUO, 0x

These agreements are as defensible as centralised exchanges, albeit with some drawbacks.

First, all of these agreements are subject to the constraints of their underlying blockchains, where they need to settle trades, including non-deterministic order execution, high delays and front-running by miners. All of these restrictions discourage liquidity providers, thus increasing the slippery slope.

Second, in general, all decentralised exchanges do not support cross margin or net position settlement. While I would like to eventually see this development in the DeFi ecosystem, it is clearly many years away. Meanwhile, the centralised exchanges FTX and Binance now offer cross-margin services and are rapidly expanding their product range to maximise traders' capital efficiency.

Mixer, Tornado. Cash

Tornado Cash is unique in the above DeFi protocol. While other companies focus on the loan/loan and trading areas, Tornado focuses on hybrid capital to maximize user privacy.

Today, Tornado Cash does not support privacy payments in the pool. Instead, it can only be used for anonymized funds. Tornado's defensible source is the size of its anonymous pool. Because money circulates relatively quickly in Tornado (the entire asset base flows once every 1-2 weeks), the network effect can be fleeting.

Moreover, beyond a certain point, the larger pool of anonymity is marginal. For example, as the anonymous set grows from 500 addresses to 1,000, it is not clear that the next marginal user will care. Who thinks 1/500 is not good enough and 1/1000 is good enough? As a result, Tornado Cash is not as defensible in its current form.

However, in a future version of the service, Tornado Cash aims to support asset transfers of privacy in the privacy pool (rather than just anonymized funds, which is currently available). In this model, capital is stickier because it doesn't leave the system as quickly. This will allow the anonymous pool to grow larger, making it more useful for large-scale capital.

The idea that large amounts of capital will only go into large privacy pools is unique to the other services mentioned above. For example, if the entire privacy pool has only 1000ETH, then the pool is not available to people who want to anonymize the 9000ETH, and it is actually not good for the original 1000ETH in the pool, because the owners of the original 1000ETH May not want 90% of it to be associated with another 9000ETH.

For users who want to anonymize 10000ETH, they may need a pool with 90000ETH. This model, while not yet available, is clearly more defensible than the current situation because it enables the richest to use the service, and the richest have the greatest incentive to hide their wealth.

Defensive positioning

Considering the hypothetical difficulty of forking these protocols, and our observable experience to date, I rank them defensible from strong to weak. Please note that the ranking is a guess because it cannot be quantified:

1. The stable currency of legal currency collateral -- USDT
2. Synthesize stabilization currency -- Maker
3. Mixers - Tornado. Cash
4. General synthetic asset agreement -- Synthetix
5. Mortgage money market -- Compound/ lendf.me
6.AMM -- Uniswap, StablecoinSwap, Shell, Banco, Futureswap
7. Unmanaged centralized order book trading platform -- dYdX, IDEX, NUO, 0x

The USDT was put in the first place because it faces the biggest competition, and is still 5-10 times bigger than its biggest rival, USDC. Although the USDT is controversial, it is clearly highly defensible. Coinbase, one of the best-funded companies in the sector, has been unable to substantially replace the USDT over the past 18 months. Although a stable currency liquidation could change that, it is still too early.

Based on the above, it should be clear why Maker came in second. If DAI is not liquid and is not available outside the core protocol, then the Maker protocol will not work. These two characteristics are not easily bifurcated and are difficult to replace with subsidies.

I put Tornado in third place, above loan agreements, because wealthy users, who provide most of the funding in these agreements, need other wealthy users to participate. And because wealth is not evenly distributed (there are relatively few wealthy users), I expect the market to support only one or two privacy markets, rather than the 10 or so that the lending and trading markets can support.

Although Synthetix is similar to Compound, I rank Synthetix above Compound because of the bridge between real and synthetic assets.

One of the common features of the second half of the agreement is increased competition. Empirically, this is clear: entrepreneurs and venture capitalists are betting that these markets are not very defensible. Moreover, as mentioned above, competitors can relatively easily channel the same liquidity as most markets today by subsidising liquidity.(note: in addition to the weak moat, the market size potential is also an important reason for them to enter, in essence, it is still early, there is no single big possibility)

Network effects at the ecological network level

Before concluding, consider the impact of the above mentioned on ethereum defensibility.

Simply put, ethereum is defensible, at least in relation to the DeFi protocol, which is substantially better than any single DeFi protocol. The main source of ethereum defensibility is not capital or liquidity, but the overall composability and interoperability of these protocols.

Amazingly, someone can use ETH as collateral and make DAI from it, lend it DAI in Compound or lendf.me, then borrow ZRX using DAI as collateral and sell ZRX for the ETH... This is just a single transaction over a single time period.

The final test of DeFi's ethereum-level network effects was the first bZx attack. The attacker's transaction was probably the most complex the ethereum network had ever handled, linking five complex protocols together. It would take years to rebuild this level of interoperability infrastructure in any other ecosystem (as ethereum has experienced), so I recommend that most of the new layer 1 teams focus on use cases outside of DeFi, at least until they reboot their respective ecosystems.

Original link:https://mp.weixin.qq.com

Welcome to join the official BlockBeats community:

Telegram Subscription Group: https://t.me/theblockbeats

Telegram Discussion Group: https://t.me/BlockBeats_App

Official Twitter Account: https://twitter.com/BlockBeatsAsia