Research Summary: Filecoin

An all-rounded storage solution for the future

Find us on Twitter or at palmdigital.io.

By Mahomed Ibrahim

This is a summary of our in-depth investment research report which can be found here.

Storage is a primitive function of computing. Peer-to-peer structures have always been popular solutions to scale storage as they were a way to leverage fallow capacity on the PCs of peers. But early attempts by protocols such as Napster, Gnutella and BitTorrent faced security problems around verifying files.

In 2014, inspired by Bitcoin, Juan Benet proposed creating unique identities for files using cryptographic hashing of their contents and overcome this problem. And in 2017, he launched IPFS based on this concept. It was revolutionary. Not only does the hashed identity make the files immutable, trustless, and permissionless, it also enables decentralisation. This eliminates the risks that arise from single points of failure and DoS attacks and at the same time, brings files closer to need, solving latency challenges. This method of addressing files by their content is also a game changer as the internet today is addresses files by location and as a result, a large part is missing, broken, or rotting.

IPFS, however, did not solve the problem of persistence – there was no guarantee that peers would store files. Juan therefore built Filecoin as an incentive layer on IPFS. Filecoin is a decentralised marketplace for storage, retrieval, and compute built on a blockchain with its own FIL token used for transactions.

The protocol raised $257m in a 2017 ICO  launched Mainnet in 2020. And it has successfully set up a storage marketplace that has 4,800 storage providers (SPs) providing 13 EiB of capacity and 1,300 clients – a quarter of whom have datasets > 1,000TiB. It is issuing its FIL tokens to bootstrap its network by subsidising storage prices.

As a solution for both web2 and web3, it has a large addressable market. And it has won significant clients including OpenSea, Metamask, LivePeer, and Nasa Hubble Telescope.

One of its main challenges, however, is scaling. Its blockchain is computationally heavy with long storage and retrieval times. This makes it more suitable to cold storage. Fortunately, this is where 80% of world’s data lies. And, importantly, ConsensysLabs flagship project – Interplanetary Consensus (IPC) – aims to solve for this. It will allow users to launch and customise their own subnets according to their individual needs (scalability vs security).

Other challenges include the fact that, as an infrastructure project, Filecoin does not have a front end and is very complex to use and understand. And barriers to entry for SPs are high. 1 PiB of storage costs requires c.$100k of capex and another $50-$100k of collateral at current token prices.

Importantly, Filecoin has just launched a Virtual Machine (FVM) which allows developers to customise what can be done on the blockchain aside from storage and retrieval. And FVM solves most of the challenges above. It will enable the development of dapps that make Filecoin user friendly and create lending markets for SPs collateral.

Combining a Virtual Machine with Filecoin’s storage markets will enable things not currently possible in blockchain, e.g. reputation scoring, data referencing for decision-making and tracking the lifecycle of data. And it will make it easier to monetise datasets. There are lots of interesting projects already building on FVM.

And the combination of FVM, IPC and compute has the potential to be a game changer in blockchain as it will empower web3 dapps with the programmability, computation and scalability needs to compete with web2.

As far as team goes, Filecoin is run by a solid founder and team with a track record of successful execution. The team is transparent and should have sufficient runway to keep building.

It also has a strong community with a large social media following and 1,400 active developers. They have run 73 hackathons to date and attend dozens of conferences each year. They also run accelerator programs and issue numerous grants. Governance, however, is technical to participate in.

Filecoin faces a fierce competitive landscape from both centralised and decentralised parties. In the decentralised space, each competitor targets a slightly different set of user requirements. Storj, which aims to compete with AWS S3, is much easier to use and extremely fast, but this has come at the cost of centralisation. Swarm, an Ethereum Foundation-supported project, which incorporates privacy, makes retrieval easy and is Ethereum compatible is one to watch. And Arweave, for permanent storage, looks like a winner.

As far as competitive advantages go, Filecoin is building a network effect of SPs, developers, and users. And it has some stickiness in the forms of the duration of deals locking in SPs, its ability to offer low prices, vesting terms for VCs and the sunk capex of SPs.

However, because of its complexity and the capital requirements, SPs are mostly professionals. This means that they have more overheads than if they were using fallow capacity. And because storage has economies of scale, it will be difficult to compete against the centralised cloud providers without subsidised pricing. And for subsidies to be sustainable, the network must grow.

In terms of tokenomics, Filecoin is minting a lot of tokens, but this is a necessity to bootstrap growth. And its baseline minting structure is a good way to cap minting by linking it to network growth.

At present, SPs, who need the tokens for collateral, are the only buyers. And while these tokens are vested, the network must sustain growth to ensure SPs renew deals and keep this supply off the market.

Positively, indications are that the token will be deflationary from 2027 as burning and vesting offsets issuance.

As a consumable asset, Filecoin is difficult to value. However, the current price is implying a 40% p.a. growth in revenue over the next 10 years. This is not unreasonable, given the low starting point, growing data and opportunities for other sources of revenue post the FVM launch.

Here is a Google Sheets workbook we have built that allows you to model token supply and solve backwards for implied protocol growth using both a Monetarist Theory as well as DCF valuation.

The biggest risk Filecoin faces is one of circularity. If the price collapses, SPs won’t be rewarded enough to add storage capacity. This will drive baseline minting to zero and in turn result in less token issuance. This will disincentivise SPs. And at same time, Protocol Labs token vesting won’t be able to fund their own overheads.

The good news, however, is that storage utilisation has been increasing even through this bear market – a sign of growing demand.