"The DeFi Déjà Vu"——An analysis of the current DeFi dilemma
星球君的朋友们
2019-08-26 06:53
本文约7549字,阅读全文需要约30分钟
What happened to the once highly popular "DeFi" ecology?

Editor's Note: This article comes fromForesighter Foreseer (ID: yujianjia2020)Editor's Note: This article comes from

Foresighter Foreseer (ID: yujianjia2020)

, author Jiang Xin JX, reprinted with authorization by Odaily.

In August 2018, Dharma Labs co-founder and chief operating officer Brendan Forster first proposed the concept of "DeFi" with the belief that decentralized finance will become the mainstream in the future. In his article "Announcing De.Fi, A Community for Decentralized Finance Platforms", he believes that "DeFi" projects need to meet the following four characteristics: build on a decentralized public chain, financial applications, open source code, and have a complete developer platform.

The prediction market platform Veil announced its closure in July 2019. As a star project invested by well-known venture capital institutions such as Paradigm and Sequoia Capital, it ended dismal after only 6 months of operation.

What happened to the once highly popular "DeFi" ecology?

1. "DeFi" Ecology - Utopia for Idealists

If the locked-up amount is regarded as a measure of the overall size of the "DeFi" market, before April 2019, the total market value of the overall ecology was less than 300 million US dollars. In the following two months, the total locked-up amount increased Nearly 6 times, the highest reached nearly 1.72 billion US dollars, and is currently stable at around 1.1 billion US dollars.

Figure 1: Distribution of locked-up amount of Defi projects Data source: Dapptotal

If "DeFi" itself is a single currency, then the market capitalization should be comparable to TRON and Cardano, and the market capitalization ranks around 10 on Coinmarkcap (Figure 2).

Figure 2: Coinmarketcap market capitalization ranking Data from: Coinmarketcap

However, when the locked amount of "DeFi" is counted according to the amount of ETH (as shown in Figure 3), it can be found that the overall locked ETH has increased from 2.33 million ETH in April 2019 to 2.77 million ETH on June 25. The increase rate was only 18.8%. At the beginning, the lock-up amount of the "DeFi" ecology skyrocketed only because the price of ETH doubled during the same period.

image description


  • Figure 3: Defi Ecological ETH lock-up quantity Data source: DappTotal

  • The total amount is still very low, only 600,000 ETH, according to the current market value (ETH price of 200 US dollars), it is only 120 million US dollars.


image description

Figure 4: Defi ecological project locks ETH Data source: DappTotal

Figure 5: Lock-up of Defi projects in the past 30 days Data source: DappTotal

Therefore, it is not difficult to conclude that the "DeFi" ecology is still: a carnival of a small group of people, and the loneliness of the majority.

secondary title

2. "DeFi" dilemma

Another example: In August 2008, exactly 11 years ago, a website called AirBed&Breakfast was launched, providing Internet services for searching for overnight stays in Sofa Hakka.

The interface of the newly launched website is very simple, the service provided is "staying with strangers", and it claims to make people "forget about hotels". The company is now called Airbnb. Although there were various problems at the beginning (such as security issues, the number of listings, etc.), the problem it was trying to solve was simple: give travelers cheaper and more convenient accommodation than staying in a hotel, which it started from the beginning. successfully resolved.

image description

Figure 6: The earliest interface of the Airbnb website

  • Then, to conduct an in-depth exploration around the core functions of "DeFi", first of all, put aside the debate on the use of virtual currency, we assume that the current virtual currency is only used for transactions, and "DeFi" products provide two types of tools to serve this business:

  • trading products

loan product

1. Loan products

According to the data of Loanscan.io, a statistical platform for "DeFi" lending projects (Figure 8), the best interest rate for using "DeFi" product financing (borrowing stablecoins bound to fiat currency, which can be regarded as long) is USDC, and the minimum annual The interest rate is 5.96%, while the average annualized loan interest rate of USDC is about 11.74%. The average annualized loan interest rate of DAI, the most commonly used stablecoin in "DeFi", is 20%.

image description

Figure 7 Defi Project Stablecoin Borrowing Rate Data Source: Loanscan

Binance’s USDC loan interest rate is 0.02% per day (VIP0, the most primary customer), which corresponds to an annualized rate of 7.3%, which is slightly higher than the lowest Defi interest rate of 5.96% and lower than the average interest rate. So for Defi users, the interest rate compensation for sacrificing the convenience of borrowing money on large exchanges is not obvious, and even many financing interest rates exceed those of centralized products.

Looking at the interest rate of "DeFi" for borrowing currency, ETH can provide a minimum annualized interest rate of 1.01%, and the average loan interest rate is only 3.4%, which is more than half of Binance's annual interest rate of 7.3% (and OKex , Huobi’s lending rate is higher). So is it more cost-effective to borrow ETH in "DeFi" products?

Let's take Dydx, which provides the lowest loan interest rate, as an example. The initial pledge rate is required to be no less than 1.25 times, and the lowest pledge rate is no less than 1.15 times. Less than 115 US dollars, and no less than 125 US dollars at the initial stage. Users need to have excess assets for collateral behind the loan.

image description

Figure 8 Defi project digital asset lending rate Data source: Loanscan

When the risk rate (Margin Level) of Binance loan is ≤ 1.1, the system will execute the forced liquidation process. The risk rate is the ratio of the market value of the total assets to the sum of the loan and the interest. Therefore, the user needs to borrow 100 US dollars of ETH, Only need assets of 90.9 US dollars. It can be seen that although "DeFi" products have more advantages in the interest rate of borrowing ETH, due to the need for over-collateralization, relatively speaking, the utilization efficiency of funds is less efficient.

Decentralized exchanges represented by 0x protocol and IDEX were the main direction of decentralized financial products before the concept of "DeFi".

According to DappTotal statistics, the total 24h trading volume of the top 10 exchanges is 27,000 ETH (August 15, 2019), which is roughly equivalent to the 60th place in the ranking of all ETH trading pairs. In other words, the combined trading volume of all currencies on the Top 10 decentralized exchanges is equivalent to the volume of one trading pair of ETH on a small exchange.

image description

Figure 9 Dex project statistics Data source: DAppTotal

The order model has obvious liquidity defects, so no specific analysis will be made here. Automatic market makers provide readily available liquidity by quoting traders based on algorithms based on predefined formulas. Alex Evans of the Placeholder Fund compared it to MMORPG games. Users only need to trade with asset pools. But for traders on Uniswap, exponential transaction slippage makes large transactions very uneconomical.

As shown in Figure 11, the price began to rise significantly after the transaction amount exceeded $10,000, the price slippage increased sharply after $100,000, and the price has risen exponentially after $1,000,000.

image description

Relatively speaking, it is optimistic that Uniswap has well met the needs of transactions with a single amount below $10,000. According to the statistics of Defi Pulse, Uniswap’s current total locked position is 43.9K ETH. Price) under the premise that the order of 10,000 US dollars can be sold 790 times. (Note: The actual number of transactions should be much smaller than this number, because the transaction slippage will increase after the reserve pool becomes smaller.) If the amount of each transaction is further reduced, take $10 as an example (due to bearing transaction and gas fees, a single No matter how low the amount is, it doesn't make much sense), then 790,000 transactions can be accepted. This number of transactions can already meet the application of some small-amount exchange scenarios.

The problem is that if a certain small-amount exchange scenario is opened up in the future, for more high-frequency demands, the total capital load of 790,000 transactions is not considered ample. For the current speculative trading market, it is obvious that the transaction carrying capacity of only a few hundred orders exceeding 10,000 US dollars is indeed insignificant.

secondary title

3. “DeFi” or Dilemma

In the course of the development of cryptocurrencies, countless cases have proved the fact that the original intentions of the creators are often based on idealistic spirit and self-interpretation of the future, but the reality always uses familiar contexts to give new ideas to the new ones without thinking aggressively. Things give ultimate definition.

Satoshi Nakamoto described Bitcoin as a "point-to-point electronic cash", hoping to subvert the existing currency system, so the early believers were full of liberals and Austrian diehards; The wave of the current wave of embracing the public has to give up the expectation of "monetization", and the research on the hedging and speculative value of "digital gold" has become the mainstream positioning of Bitcoin.

Coincidentally, Ethereum, whose market value is second only to Bitcoin, was originally conceived to build a decentralized "world computer platform". In 2017, when the market value exploded wildly due to crowdfunding, the Ethereum community might not have thought that it had become the most profitable "crowdfunding machine" in the world. Co-founder Joseph Lubin recently proposed a more pragmatic vision for ethereum: a "decentralized global settlement layer." From "World Computer" to "Global Settlement Layer", we can see that the vision that was once too idealistic is gradually settling down.

Recalling why we expected "DeFi" in the first place, there are no more than the following reasons:

  • Censorship resistance

  • transparency

  • Censorship resistance

  • transparency

Programmable Finance (Composability)

Self-hosted (no centralized counterparty risk)

Regardless of whether the features must be completed by the blockchain or not, we will not discuss which are "pseudo-requirements", we assume that they are all necessary factors for the future financial environment, and rely on a highly decentralized blockchain to complete.<>So, at what cost?

capital utilization efficiency

The average pledge rate of "DeFi" projects on Loanscan exceeds 3 times (319%), and the highest Compound v1 requires 4.74 times the asset pledge, which means that to lend 100 dollars of funds, you need a mortgage value close to 500 dollars assets. On Binance, users need to borrow $100 of ETH, but only need $90.9 in assets. The ratio of the utilization rate of the two funds can be said to be very different.

image description

Figure 11 "DeFi" project pledge rate Data source: Loanscan

But in any case, the practice of using personal credit to increase capital leverage needs to expose personal information, which is contrary to the original intention of "DeFi" to resist censorship.<>safety

Paradox 2: Transparency

safety

- "DeFi" products especially emphasize their transparency, and their transactions can be checked on the chain, and there is no risk of centralized institutions doing evil. But on the other side of the coin, take the order form of 0x as an example. The open order Maker’s pending orders are summarized by the relay party. At this time, TMiner sees the order flow summarized by the relay party (a manifestation of the so-called transparency). Then you can pay a higher Gas Fee to complete the transaction in advance, the so-called Front-Running.

For the Front-Running problem, there are currently different solutions, such as the improved solution based on the Bancor protocol proposed by Nate Hindman, and the zero-knowledge proof-based solution of Starkdex, which StarkWare and 0x are cooperating with. But generally speaking, they are still in the research or testing stage, and the actual effect is still unverified.

But single-point risks also arise from this. The asset platform Synthetix encountered an oracle attack in June 2019, and the attackers captured more than 37 million sETH in total. If multiple Oracle Feeds are created to extract data, it means that while the robustness of the system is improved, the complexity also increases.<>fluidity

Paradox 3: Degree of decentralization

fluidity

Taking an exchange as an example, due to reasons such as product operability, technical complexity, and gas fees, it is difficult for you to imagine professional traders making markets on decentralized exchanges to provide liquidity. At the current stage of market development, even Most of the liquidity of centralized exchanges also comes from market makers.

In the DeFi trading environment, Uniswap has cleverly designed an automatic market maker mechanism, that is, market makers only need to deposit coins to provide liquidity to obtain income. The income is accumulated transaction fees and temporary losses caused by price fluctuations (impermanent loss).

Figure 12 Yield analysis of Uniswap market makers (1) Data source: Uniswap.io

Mohamed Fouda calculated the profit and loss of market making under different trading volume and price changes in the article "Uniswap Market Making Profit and Loss Analysis". The results found that: 1) When the price changes significantly, especially when the trading volume is low at the same time, market making activities will cause market makers to suffer losses, and the return is not as good as holding assets. 2) The larger the transaction volume, the higher the transaction fee, and the higher the profit of the market-making business.

image description

Therefore, for users of decentralized "DeFi" products, "retail market makers" who need to make money by providing liquidity more conveniently, a more strategic strategy is to increase or decrease the number of tokens supplied according to price changes, and Choosing between calculating holding assets or providing market-making services, this threshold may even offset the convenience brought by the "automatic market-making mechanism" to retail investors.

Institutional investors will use more convenient and more profitable centralized markets for market making. This further reduces the trading volume on Uniswap, which makes the market-making income lower, forming a bad cycle of negative feedback.

secondary title

4. Where is the future of "DeFi"

  • As the largest P2P file sharing protocol, BitTorrent once accounted for 70% of the entire Internet traffic at its peak in 2006. The online update of the famous World of Warcraft uses BT technology to download and distribute. It is no exaggeration to say that BitTorrent is the decentralized application that aggregates the most user traffic.

  • However, in 2018, it was acquired by Tron for only 140 million US dollars. On the 30th day after the acquisition, the founder Bram Cohen announced his resignation. For the dismal ending, Simon Morris, a senior employee who has worked in BitTorrent for ten years, summed up the experience of several decentralized applications worth mentioning here:

  • Although the BitTorrent ecosystem has a certain scale and gained enough influence in the industry, as a disruptive and innovative technology, BitTorrent itself seems to have never succeeded in relying on the technology itself.

  • While BitTorrent is a great idea, it's not a good business model. Decentralization has unleashed an unstoppable thug trying to upend traditional rules, but we must not mistake rule breakers for the future winners of the internet.

With the advent of BitTorrent, the whole concept of "file" seemed to disappear from people's view, and we found that many old media were reborn in a new and better form. The most typical example of this is the music streaming media provides. provider Spotify and video content service provider Netflix.

BitTorrent technology has indeed brought subversive innovation, but the "winners" who use this technological achievement have actually gotten rid of the appearance of decentralization. For companies like Spotify and Netflix, there is no need for decentralized technology at all, because it adds complexity, and as a result, it will make things worse and the user experience will be worse. But it is undeniable that their success is due to the so-called "paradigm shift", that is, a disruptive innovation that abstracts "files"

In the first stage, make a single function 100% available, forming a closed loop of business scenarios.

"Leverage lending transaction" can be said to be the largest application scenario of decentralized financial products at present, but Maker, which has the highest mortgage amount of "DeFi" assets, has achieved effective leverage, but lacks transaction functions. According to the data of StableCoinindex, DAI currently has a 24-hour trading volume of 14 million U.S. dollars, ranking fifth among stablecoins. Although the ranking is not low, it is far from the top 4. Compared with the top-ranked USDT, which has a 24-hour trading volume of 17 billion U.S. dollars The trading volume is only 0.08% of it.

image description

Figure 14 Ranking of Stablecoin Market Value and Trading Volume Data Source: cryptoslate

In terms of the number of trading pairs, according to Coinmarketcap data, there are about 77 trading pairs of DAI (excluding 0 trading volume), while there are more than 400 trading pairs of USDT according to incomplete statistics.

If the current work progress of Maker is 50%, then Maker needs to find trading scenarios for DAI in the future, and continuously increase the number of trading pairs and transaction amount of DAI in the exchange. Don't put shackles on yourself because of "DeFi". Cooperation with large centralized exchanges may be an optional path.

Looking at Uniswap again, the AMM model has initially run through, and the next step is to achieve the closed loop of the scene, what is needed is the exchange link of small and micro transactions. The above article has stated that large-amount transactions are not suitable for AMM, so what are the possible scenarios? In-game token changer, or micropayment (phone top-up/offline shopping...)?

The second stage is the birth of a paradigm shift.

All in all, the first thing that the designers of "DeFi" need to do is to get rid of their aloof identities as utopian fantasists, and start building real application scenarios from a loose community-based organization to a more efficient commercial operation.

epilogue

secondary title

epilogue

"Space View" by Partho Mondal

Placeholder VC researcher Mario Laul used the phrase "the DeFi Déjà Vu" to describe the feeling of trance caused by the withering of the "DeFi" ecology. Looking back at the "DeFi" boom, the scenes of subverting traditional finance seemed like a lifetime ago .

But perhaps, when "DeFi" falls into the valley of despair, it is also the best time for Buidls to build a new financial order Tower of Babel.

  • "P&L analysis of Uniswap Market Making",by Mohamed Fouda

  • “Trade-Offs: Decentralized Exchange”,by Justine Humenansky

  • “DeFi: What it Is and Isn’t”,by Justine Humenansky

  • “The future of decentralized finance” by Linda Xie

  • “Uniswap Deep Dive”by ZPX | Satoshi&Co Newsletters

  • “UniSwap Traction Analysis”  by Ganesh 


星球君的朋友们
作者文库