A few months ago, I wrote this post on the subsectors I expect to lead the charge into the next crypto summer. In the coming months, I’m planning to cover notable projects in each of these subsectors on a more granular level.
This week we begin with a clear fan favorite; the always exciting world of data storage ;)
*As always, this is not investment advice, insert disclaimer here, do you own research etc.
Public Cloud: Arguably the Best Business on Earth
The market for public cloud services is showing itself to be virtually unbounded - accelerating another 17% YoY to US$474b in 2022. The tech giants have realized the future is digital and are sprinting head-long into the void, promising to satisfy the unending thirst for digital infrastructure with pledges of 50 - 100 data centers annually to keep pace with demand.
The players cornering this capex intensive, yet heavily lucrative market are the familiar names: Amazon Web Services, Microsoft Azure, Google Cloud, Alibaba, etc. These top four players take up >70% market share today and, given returns to scale, that figure is likely to continue climbing.
The competition is fierce for good reason…
Market leader AWS (33% share globally) reported US$62b in revenues in 2021, growing 38% YoY with US$19b in profits. An absolute juggernaut despite public cloud adoption still in relative infancy. As more economic activity turns digital, IoT devices proliferate, and machine learning becomes more integral to decision making, public cloud infrastructure is the latest essential utility. But, a utility growing 38% YoY with ~30%+ margins…
Are there any businesses better positioned for our 21st century transformation into the digital economy? Massive data monopolies, throwing off billions in cashflows which are reinvested into a capital intensive, yet lucrative industry yielding 30% margins, crafting unassailable moats, and riding an exponential growth curve?
It’s hard to imagine any companies successfully taking these giants on head-to-head.
However, disruption often comes from unexpected directions. Marriott and Hilton were both slugging it out in the trenches, and now AirBnB is worth more than either in 14 short years.
Can something similar happen in a large market like data storage?
Integrated Incumbents vs. Protocols
While public cloud providers offer an array of services, data storage is a meaningful chunk. The market reached US$70b in 2021 and is set to accelerate by ~24% through 2029 reaching a staggering US$375b by the end of the decade: a colossal segment for integrated cloud providers.
Filecoin hopes to become the AirBnB equivalent in data storage - challenging the incumbents with a fundamentally different model. By using distributed incentives, Filecoin crafts a marketplace between a fragmented base of users demanding data storage and providers with “excess” storage capacity.
For the more technical, this is essentially a peer-to-peer version of Amazon S3.
Source: Filecoin documentation
As opposed to a company, Filecoin coordinates various parties through a decentralized protocol. Early network participants are rewarded with “equity-like” tokens to bootstrap a useful service (in this case file storage) which can incentivize a collection of smaller players to band together - taking advantage of excess capacity or incentivizing the building of capacity in a distributed manner - which ultimately provides a useful service at a (hopefully!) lower price point.
Far from being alone, this physical proof of work model is beginning to proliferate to other layers of the infrastructure and application stack:
Storage. Compute. Networking. Mapping. Data provisioning. All of these large markets may prove more efficiently managed via protocol incentives than companies over time. Instead of one centralized intermediary, a host of smaller actors with distributed incentives may prove superior. Decentralized marketplaces - with distributed capital aggregation and limited monopoly profits on the backend - undercutting incumbents.
An interesting parallel is Wikipedia. Out of the gate it was spotty, inefficient, and middle school teachers across the globe scolded aspiring 21st-century Herodotus’ for using such “unreliable” sources in their retelling of the battle of Thermopylae. Fast-forward a few years and the bottoms-up service is the undisputable champion in breath, depth, and quality of information. Even with the resources of a tech giant, there are always more talented individuals outside of your organization than within.
At least that is the thesis.
Filecoin: Priced to Perfection
As much as I would love to hype the “future of data storage”, the Filecoin investment thesis is challenging. Despite a ~95% decline since it’s April 2021 highs, the network still appears quite rich relative to traction.
Don’t get me wrong:
Great team? YUP.
Massive market? YUP.
Shipping product aggressively? YUP.
Encouraging ecosystem partner growth? YUP
All look decent.
The red flags for me are the current valuation relative to proven demand as well as the token overhang for new investors.
For any marketplace business with a large addressable market (think horizontal ecommerce), its silly to gauge the opportunity using standard multiples early on. Profitability is a decade away and revenues can only be scaled up once the platform has reached escape velocity, cementing itself through superior liquidity as THE marketplace in the space. The cyber bazaar where users and sellers congregate.
Only at scale, does the intermediary increase take rates and begin showing respectable valuation multiples. This familiar story is demonstrated by Amazon, Alibaba, MercadoLibre, Flipkart, and SEA: 1) Gross Merchandise Value (GMV) multiples → 2) Revenue Multiples → 3) EBITDA multiples - the shift occurring as the network effects coalesce and the intermediary moves from market share wars into its value extractive phase.
I think about crypto infrastructure networks similarly (but with less ability to monopolize profits on the back-end given the opensource nature of the code). At an early stage of industry development, we must tease out the “GMV” equivalent figure - a measure of the network effects and increasing market liquidity. This doesn’t happen overnight.
To be fair, Filecoin has done a solid job of building out the supply side of the network. Using crypto incentives, the network has accumulated over ~16 EiB in network capacity from over 4000 different storage provides: equivalent to ~65,000 copies of Wikipedia1.
Not shabby. On the demand side, however, we have reason to pause.
While this chart flaunts an attractive trajectory, scale matters. 115 PiB of active deals represents just 0.7% utilization of the network capacity - a mere 467 copies of our favorite online encyclopedia. Crypto incentives are useful for bootstrapping the supply-side, but at some point, demand needs to arrive to make the network sustainable.
Could we be in the early innings of a potential inflection? Yes, it’s possible. Is unproven product-market-fit and unit economics something I plan to pay ~US$17b for?
It is not.
While the outstanding market cap of Filecoin (assuming Aug 14 trading price of $8.66) is only ~US$2.2b, just 13% of tokens have been released.
Filecoin went live in October of 2020 indicating the network is approaching ~2 years on the vesting schedule. As shown above, we are entering an inflection in the vesting curve where Fred Wilson (USV), Naval Ravikant, The Winklevoss twins, Sequoia, and other early backers will begin sprinkling a consistent stream of private tokens on unsuspecting retail plebeians. This presents significant sell pressure in the years ahead which the network growth will need to overcome.
Verdict: Because of the limited tokens outstanding, the fully diluted valuation of US$17b is likely a better valuation metric to use here. And based on traction, and despite a ~95% collapse from the April 2021 highs, I find it difficult to get excited about $FIL at these levels.
Catalysts to Watch…
However, this is a strong team, with a growing ecosystem, attacking a colossal market opportunity. A few things which could spark an inflection and help make today’s valuation make more attractive:
Early signals from large enterprise customers trusting the protocol enough to cede share from giants like AWS (unlikely near-term but would be big)
Increasing penetration of core customers like DeFi apps and NFT platforms keen for decentralized storage solutions (Early traction, but relatively small user base today)
The dramatic expansion of the web3 user and developer-base through the arrival of “main-stream” applications like gaming and web3 social built on decentralized infrastructure (potentially sizable catalyst on a ~3 - 5 year time horizon)
Large token “burns” to reduce the outstanding supply significantly (unlikely)
Absent these catalysts, I would simply wait for a material reduction in price from current levels for the risks outstanding to justify the upside. If we are in the early innings of an extended winter, bargain prices may materialize. However, if we do see some sort of Fed pivot heading into year end, it’s unlikely to enter a valuation zone attractive for long term accumulation. Crypto markets tend to get excited in a hurry.
Unfortunately, Filecoin can be both the future of storage and overpriced today at the same time.
Messari State of Filecoin Report Q2 2022. https://messari.io/report/state-of-filecoin-q2-2022