58 total views
Author: Tushar Jain and Spencer Applebaum, respectively Multicoin Capital co-founder and analyst
This article was co-authored by Tushar Jain, co-founder and managing partner of Multicoin Capital, and Spencer Applebaum, analyst of Multicoin Capital. Multicoin Capital is a theory-driven investment company that invests in cryptocurrency, tokens and blockchain companies.
In the past ten years, there has been a business trend that unleashes value by increasing the utilization rate of capital-intensive assets: Airbnb has increased the utilization rate of vacant houses, which has caused a shock in the hotel industry; Uber has increased the utilization rate of idle vehicles , Subvert the taxi industry. According to this logic, the promotion of decentralized storage can also be: they will increase the utilization rate of idle computers, thereby subverting the cloud storage industry .
Cloud storage is a market worth challenging. In the past ten years, we have witnessed a technological paradigm shift from internal servers to cloud computing, and new technology giants have also been born, such as Amazon Cloud (AWS), Alibaba Cloud, Google Cloud (GCP), etc. According to estimates by market research companies, the global cloud storage market will reach USD 137.3 billion by 2025, with a compound annual growth rate of 22.3% from 2020 to 2025. This market is huge.
Smelling the taste of opportunity, a large number of open source networks poured in at the end of 2016 and early 2017, wanting to get a share of the pie. The first projects include Filecoin, Storj, Sia, SAFE, etc. In 2018, Arweave released the mainnet, which brought an idea unprecedented in the market: permanent storage.
China has set off a wave of Filecoin mining. The price of Fil is currently $28, which means that the fully diluted network is valued at $56 billion. Before and after the mainnet went live, the market was excited and the hype was unprecedentedly high. Similarly, Arweave has recently raised funds from well-known investors, including Andreessen Horowitz, Union Square Ventures, and Coinbase Ventures. In late April 2020, Arweave (2.0) was successfully listed, entering the market before Filecoin and maintaining continuous growth.
When investors assessed the upcoming storage war, some major problems began to emerge. Between Filecoin and Arweave, which network will developers choose? Can the decentralized storage network compete with Web2 giants such as Amazon and Alibaba? How to compare these two decentralized storage methods?
This article discusses these issues and gives our evaluation framework for the decentralized storage market, and does not constitute any investment advice. In this context, the first step we need to understand is the differences and trade-offs between these two storage projects.
Based on contract vs. permanent storage
Both Filecoin and Arweave enable data to be stored in a decentralized, trustless, and censorship-resistant manner. Both projects have adopted blockchain-based solutions. And both networks can be used to store data for a long time, either as a reference or as a web hosting. At first glance, they are very similar. Therefore, instead of evaluating Filecoin and Arweave’s respective blockchain networks, it is better to choose a better entry point-how they intend to provide storage services to end users.
Filecoin follows the practice of AWS and other cloud service providers: contract-based storage . We’d better generalize contract-based storage as a pay-as-you-go model . The node network paid by the customer will store X bytes of data in the Y time period and guarantee the recoverability of Z. This is also the method adopted by protocols such as Storj, Sia and SAFE.
On the other hand, Arweave introduced a new model to the market: permanent storage . This was impossible before the advent of encryption. With permanent storage, users only need to pay a one-time prepayment and the network can store data permanently . Perpetual storage can serve both the addressable cloud storage market and create a brand new market (we will introduce it later). The Arweave protocol achieves this by using cryptoeconomic game theory and creating an endowment fund to compensate miners and ensure the availability, reliability and durability of data.
The main “function” emphasized by Filecoin and other contract-based decentralized storage protocols (Sia, Storj) is lower cost. They claim to be able to provide lower costs because they use free hard disk space. Filecoin also has a high degree of censorship resistance.
Given the early development momentum of Filecoin, we have reason to expect that it will occupy a certain share of the Web3 storage market. But in fact, it cannot continue to provide services at lower prices than Amazon. Filecoin will effectively subsidize the cost of storage service purchasers by issuing FIL in the early stage, but (1) the issuance subsidy will inevitably affect the currency price, it cannot last forever, and (2) Amazon, Alibaba, Tencent and Google Obviously it has more capital than Filecoin. If the price war really starts, the giants will have more ammunition .
Amazon’s entire business model is verticalizing the supply chain and cutting costs significantly. Storage customers will benefit from these economies of scale (as shown in the figure below), but at the same time bear platform risks.
Source: Thomas Vachon
An example of Amazon’s economies of scale is that they can negotiate bulk purchase agreements to obtain cheaper electricity. According to a study called “The Cost of Supporting Computing Power” conducted by Ponemon Institute LLC, the average annual cost/kW of a data center with more than 50,000 square feet is US$5,467, while the annual cost of a data center with 500 – 5,000 square feet It is $26,495.
Returning to Filecoin, it is quite difficult for an individual with free hardware space to participate in the supply side of the network. Therefore, the network (at startup) is likely to be dominated by miners who have invested a lot of money in powerful miners. We learned from Filecoin’s testnet Space Race that at least 230 PiB storage spaces will be used to support its network launch. From this point of view, the top ten miners drive 145 PiB. Amazon runs the world’s largest data center, and the data center cost of these medium-scale mining facilities is unlikely to be lower than Amazon.
According to this logic, the main “characteristic” of Filecoin is not low cost, but ideology and strong resistance to censorship .
As mentioned above, the Arweave protocol provides permanent storage services. It is not achieved by establishing a contract between the user and the storage provider, but by encouraging the miners to copy as much of the stored data as encrypted economic incentives . Perpetual data storage is a brand new market that Amazon, Google and other companies have not yet developed.
In order to be able to store files on Arweave, developers will create a transaction and pay a certain amount of AR tokens as a network fee (the current price is $0.005/MB) to store the data permanently. The low-end package of Amazon S3 charges 0.000023 USD/MB per month. This means that every Arweave user will have to break even after 18 years. Arweave has not said that he wants to compete with Amazon on cost. Obviously, Arweave users obviously paid a premium for permanent storage .
We believe that there are currently two market segments that will adopt Arweave soon. They are less price sensitive and have a greater demand for permanent storage space.
Blockchain for data availability
Blockchain is often used to permanently store historical records of transaction networks. Arweave allows Layer 1 and Layer 2 teams to permanently store a copy of their ledger. This is the key to auditability and redundancy. Some well-known teams such as Solana and SKALE are exploring the use of Arweave for this purpose.
Arweave recently partnered with the Internet Archive (a non-profit organization under the famous Wayback Machine project) to make the access channel of this popular archive permanently fault-tolerant. Although “link rot” is not widely known, the problem is already very serious. For example, in the ruling of the U.S. Supreme Court, more than 49% of links were broken. The Internet is rapidly consuming data, and the more data is discarded, the more links are broken.
Although we don’t know exactly how big the market for permanent storage space is, we know that once the mass blockchain reaches the network scale, it needs to store massive amounts of data. We are also well aware that these massive amounts of data created by humans are equivalent to human history, and some organizations are actively seeking to record them in a fault-tolerant way.
In addition, there are several other markets that we can reasonably expect that they will also value the durability of data storage and are willing to pay a premium for this:
- A group of journalists who hope that their reports can be retained forever to prove the truth.
- A group of lawyers engaged in personal property or trust business.
- NGOs or foundations that want to keep their career records permanently.
- Individuals who want to store personal memories for future generations to read.
Image source: ViewBlock
Related network valuation and token economics
Filecoin’s native token FIL, as a utility payment token, should be valued using Fisher’s formula MV=PQ . For a simple model, we can use Filecoin’s own valuation-the market size by 2021 is 75 billion U.S. dollars, and the circulation velocity of M1 USD is 3.9 (considering that FIL is affected by circulation velocity, this is very large). We use the speed of 3.9 cautiously, because there is a small speed sink, during which storage nodes must pledge FIL to obtain rewards denominated in FIL. At a speed of 3.9, if Filecoin can occupy the entire cloud storage market including Amazon and Alibaba, FIL’s market value of US$19 billion (US$75 billion / 3.9) will be reasonable, which is 66% lower than today’s price ( $28 per FIL or a fully diluted network valuation of $56 billion).
Filecoin’s potential network valuation seems to have exceeded the fundamentals after being fully diluted, especially when you think of it as a pre-release network.
On the other hand, Arweave’s native token AR, which is also a Layer1 token, was priced at US$2.69 during the writing of this article, and was valued at US$178 million when fully diluted. It is worth noting that AR is subject to a powerful token speed receiver (more important than FIL’s cyclic payment structure, in which storage nodes receive…FIL by hosting FIL). When users pay to store data on the Arweave network, they don’t just have to prepay the miners for the entire storage fee. On the contrary, more than 83% of the expenses went into the fund pool. When the block is needed to reach a consensus at some point in the future, a part of the fund pool will be paid to the miners, who will refer to the old blocks to dig new blocks. Diggers who mine new blocks that reference old blocks. The decline in the circulation rate of tokens reduces the circulating supply of AR, and we believe that this is a better system than the high circulation rate of FIL tokens.
In other words, Arweave, as a fully released and fully functional network, has a better token economy, and its current value is 0.31% of Filecoin . The numbers are difficult to understand. Let’s take a look at the chart. Filecoin is the red line in the figure below. Arweave is the blue line.
It is difficult to say whether the valuations of these two networks will grow or shrink. It is still unknown whether Filecoin will realize the vision of annexing Amazon Cloud and Alibaba Cloud, or whether Arweave will continue to provide users with paid file storage services 20 years later. .
What we know is that Filecoin has a thriving community, and many early investors participated in the 2017 $257 million ICO. As we all know, Filecoin has been in the development stage for the past three years and Juan Benet is a very good engineer. The Filecoin team has created the Interplanetary File System (IPFS) and CoinList . IPFS used in conjunction with Filecoin demonstrates the team’s ability to create large-scale digital infrastructure, while CoinList demonstrates their ability to create well-functioning user-oriented applications. These two points are critical to success.
On the other hand, Arweave is also very interesting because it serves a whole new market, the only market supported by blockchain technology. It can satisfy both the existing storage market and this new market. The network is running and developing. Recently, the interfaces of decentralized applications such as SushiSwap, Uniswap V2 and Yearn.Finance have also been ported to the network, and the data stored in Permaweb of Arweave is growing rapidly .
We are also aware that Filecoin will encounter circulation speed problems at certain times, and Arweave tokens are subject to the inherent design of the protocol, and the circulation speed will decrease.
In short, after careful consideration of the following four points: (1) the technical and economic design of Filecoin and Arweave, (2) the current dynamics of the cloud storage market, (3) Arweave has opened up a new market and has functionality beyond price Stand out on the Internet (and the price is much lower than Filecoin). (4) The valuation of AR is approximately 0.31% of FIL. We expect that the future competition between the two will be quite exciting.
Information disclosure: Multicoin has established, maintained, and implemented reasonably designed written policies and procedures to identify and effectively manage conflicts of interest related to its investment activities. Within 3 days (the “No Trading Period”) after the publication of this report, Multicoin will comply with the “No Trading Policy” for the assets listed in this report. At the time this article was published, Multicoin Capital held Arweave ($AR) tokens.