Modeling Generalized Mining from a Fund’s Perspective: A Livepeer Case Study
Disclaimer: Vision Hill is indirectly invested in Livepeer. The content provided herein should not be considered investment advice, and is not a recommendation of, or an offer to sell or solicitation of an offer to buy, any particular asset, security, strategy, or investment product. Vision Hill has not received and will not be receiving any compensation as a result of this report.
Executive Summary: Livepeer Network
The Livepeer project is an open source project that is focused on decentralizing live video transcoding and broadcasting over the internet that is built atop the Ethereum blockchain. Livepeer launched to Ethereum mainnet on April 30, 2018.
Project Background
The Livepeer project aims to deliver a live video streaming network protocol that is fully decentralized, highly scalable, economically incentivized, and results in a solution that is prospectively cheaper to an app developer or broadcaster than using traditional centralized live video solutions. The general value proposition of Livepeer is to launch a video transcoding network that is a common or public good as an alternative to a centralized provider.
Broadcasters or streamers send videos into the Livepeer network, and the network takes care of encoding it into all the different formats necessary to reach the majority of the devices on the planet, and distributing it to as many users that want to consume it. It should be noted that all the users who run the Livepeer software are connecting to one another in a peer-to-peer network. This is similar to how Limewire or BitTorrent worked, except with a focus on live video instead of media files. At present, the cost profile of live video broadcast on the internet makes it viable for only a certain segment of the market, whereas it is impractical for the rest of the market. Users can either pay these relatively steep prices or use platforms that control their audience, advertising, and monetization options. For instance, to broadcast to Twitch, an online site that allows users to watch or broadcast live streaming or pre-recorded video of broadcaster’s video game gameplay, users need to first acquire and be familiar with broadcasting software. While live-streaming, the Twitch dashboard lets broadcasters know if the streaming software video settings are optimal. Possible ratings are “Excellent”, “Acceptable” and “Incompatible”, and Twitch may refuse to accept a stream if it is rated “Incompatible”. A decentralized platform could create more scalable, pay-as-you-go, censorship resistant, and flexible solutions for broadcasters and streamers.
On December 3, 2018, Livepeer introduced The Streamflow Proposal that discussed plans to upgrade Livepeer’s network to be more reliable, cost effective, scalable and performant. Streamflow introduces some new and changing concepts to Livepeer. Notably: off-chain job negotiation (rather than on-chain job assignment), payments via probabilistic micropayments, relaxation on the limit of number of active transcoders, the introduction of a new active network participation role called Orchestrator (discussed later herein), among other proposals.
Technology
There are four components to the Livepeer project:
The Livepeer Media Server
The Livepeer Media Server is an open source media server that is capable of taking input streams of live video and audio and transcoding them into alternative encodings, transrating them into different bitrates, and transmuxing them into different delivery formats. This is necessary so that a video, recorded in one format and bitrate, can now reach every screen on every device on every platform, each of which may have their own requirements about format and bandwidth restrictions.
There is no shortage of media server solutions on the market, however most are proprietary, developed and sold by a single company, are not open source and come with high costs or licensing fees.
The Livepeer Media Server’s development will be financed by the protocol itself given it is a core component of the protocol and the network. This will allow it to be fully open source and transparent, and as a result will benefit from community in the form of peer review, open discussion around features requests and bugs, and potentially community driven contribution.
The media server can be used as a standalone component in any media streaming application that anyone wishes to build using it, however it is truly powerful as a tool that is running within the Livepeer Network.
The Livepeer Network
The Livepeer Network comprises a set of nodes that run the Livepeer Media Server and speak the Livepeer Protocol. The network operates primarily on two types of requests: publishing a stream, and requesting a stream. A publish request is a request to transcode a live video with certain parameters, and to serve the resulting video. A request is an attempt to access an existing live stream that is on the network.
When a node publishes a stream into the network, they can assume that the stream will be transcoded into whatever formats they request (and pay for), and that the network will scale to distribute the stream to as many nodes that are requesting it.
The Livepeer Token
Before getting into the details of the Livepeer token (LPT), it is worth mentioning that the service provided by the network — video broadcasting and transcoding — is priced and paid in Ether. In fact, in the future it could be priced and paid in any token, including the Dai Stablecoin, or any token that a user on the network is willing to accept. The LPT token, on the other hand, is used by the keepers of the network — the nodes who do work to make the network run securely and efficiently.
The Livepeer token will be a native protocol token for staking, used by participants who want to perform work on the network, that serves the following purposes:
· It serves as a bonding mechanism in a delegated proof-of-stake system, in which stake is delegated to transcoders (or validators), or, long-term, orchestrators, who participate in the protocol to transcode video and validate work. The token, and potential slashing that occurs due to violation of the protocol, is necessary in order to secure the network against a number of attacks. More below.
· It routes work through the network in proportion to the amount of bonded and delegated tokens, essentially serving as a coordination mechanism.
· It is a unit of account that is specific to the Livepeer ecosystem, which forms the basis of a SectorCoin concept, applicable to additional functionality to be introduced in the future. Services such as DVR, closed captioning, ad insertion/monetization, and analytics can all plug into the Livepeer ecosystem and potentially make sure of the security provided by bonding LPT.
The ledger of token balances and double spend prevention will be provided and secured by the underlying blockchain consensus mechanism. This will be an inflationary token in which new tokens are generated by nodes who run the Livepeer Media Server and perform transcoding, in proportion to the amount of work that they contributed to the network. Nodes who don’t perform work correctly or attempt to cheat will have their token stakes slashed. They will have the effect of routing more future work, and fees, towards the nodes who have successfully performed work in the past.
Assuming competition amongst nodes to win broadcasting business and the associated fees in Ether, any increase in proportional amounts of tokens, and hence future work, that can be earned for performing transcoding will equate to a drop in the cost that must be charged to the broadcaster. This creates a positive economic effect for broadcasters looking to transcode live video through the Livepeer Network.
The Livepeer Protocol
The Livepeer Protocol specification will lay out the various roles in the Livepeer Network, the various transactions types supported, the cryptoeconomic security measures in place to prevent collusion, and the token distribution mechanics. The Livepeer Whitepaper provides all the technical details of the protocol and its incentives. As a high-level summary, the Livepeer Protocol will operate as follows:
Roles
Node — anyone who runs the full Livepeer software and participates in the protocol. All subsequent roles are also nodes.
Broadcaster — the node who is publishing a stream. They pay the network in Ether for transcoding services.
Transcoder — a node who is currently transcoding a stream into a newly formatted stream. They get paid for this service by the broadcaster (Ether fee) and increase their proportion of newly generated LPT tokens. Transcoder is a special role that is delegated towards by other nodes. They not only transcode video, but also participate in the security protocol of verifying work and distributing new tokens. The economics of transcoding is explored in greater detail later herein.
Consumer — a node who is consuming an output stream to watch the video or to serve it over a gateway to additional users outside of the protocol.
Relay node — a node who is merely passing on video without the intent of watching or serving it itself.
Orchestrator — anyone who elects to run several concurrent instances of a node for disaster recovery, redundancy, load balancing and geo-routing.
Flow of Value
As described in the roles above, the value in the network will flow from broadcaster to the transcoding nodes in Ether as payment for their services. The price that a transcoding node charges for a given transcoding configuration, will be advertised via the transcoding node, and the broadcaster can set the price they are willing to pay. Transcoding nodes and delegators also increase the portion of future work which will be routed towards them, and hence fees, by provably doing honest work.
Newly Minted Tokens
Tokens will be generated by the protocol in order to reallocate the future work towards nodes who have bonded and successfully performed work in the past, and away from those who have cheated or not performed work. Tokens will be distributed to transcoding nodes in proportion to the amount of work that they performed in transcoding streams, as well as to nodes who delegated their stake towards transcoding nodes who behaved correctly in proportion to their stake.
Consensus
There are two levels of consensus in the Livepeer Protocol. The first, the state of the current ledger of the Livepeer token balances, and validation of all transactions which transfer tokens from one account to another, is provided via the consensus algorithm in the underlying blockchain that the Livepeer token is issued on (Ethereum). As a protocol implemented on top of a base layer, Ethereum (but, in principle, agnostic to its base layer), balances and transactions will benefit from the same level of security as the blockchain’s native token itself, secured by the hash power of the mining operations or underlying proof-of-stake algorithm.
The second level of consensus is unique to the Livepeer Protocol, and will govern how the newly generated Livepeer token is distributed in proportion to the amount of work that nodes in the network performed in order to transcode and verify live streams. This will operate on a delegated proof-of-stake consensus algorithm. Transcoders will be elected to play a special role in the network consisting of two responsibilities:
1) Run highly efficient and reliable hardware and networking setups to fulfill the demand for transcoding on the network.
2) Including proof of proper transcoding transactions on-chain, which the protocol will use as input to determine the allocation of newly minted LPT tokens.
Essentially, transcoders will play an analogous role to block producers. They will earn a portion of the block reward for playing this role, and will be incentivized to act honestly, lest they lose the valuable position of holding this role.
Transcoder Selection
Nodes will optionally bond their existing LPT in order to elect a fixed number of transcoders, who will be expected to be running efficient hardware set ups in order to perform the role of transcoding and verification of work performed on the network. Protocol determined network parameters will determine the number of transcoders, the number of verifications that must be performed on each stream, and the routing procedure to randomly route jobs to transcoders in a distribution relative to delegated stake. In return for staking tokens to participate in the election process, nodes will earn broadcasting fees.
The protocol will also determine the appropriate time periods required for staking, for withdrawing stake, and for casting and changing delegation amounts for transcoders. These tunable parameters will move to ensure that the network operates in a game theoretically secure manner that also keeps up with the demand for transcoding verifications across the network.
Verification Procedure
Verification of transcoding work will be accomplished using scalable extensions to the Truebit protocol (discussed in Livepeer’s whitepaper). Transcoders will be responsible for claiming the work that they did on-chain, and they’ll be forced to pass verification of random segments of work in order to earn their fees and portion of the block reward.
Incentives
The goal of the above high-level description is to provide the following incentives:
1) Transcoding nodes want to transcode so that they can receive fees and increase their share of future work.
2) Transcoding nodes don’t want to cheat at transcoding, because they will be caught and get their stake slashed by a higher amount than they can expect to gain by cheating.
3) Nodes want to bond their tokens in order to earn fees and future work in proportion to their stake.
4) Nodes want to vote for properly acting transcoders because it ensures the efficient performance of the protocol, therefore increasing the usefulness of the network.
Initial Distribution
10,000,000 LPT tokens were generated in the Livepeer genesis block. New tokens are minted each round according to the protocol defined inflation schedule (relative to the outstanding float) and issued to transcoders pro rata to their delegated stake for performing transcoding services.
As noted previously, the main use of LPT is transcoder selection and network governance. Bonding means depositing some token into a protocol smart contract that is locked up for a period of time before it can be withdrawn. This serves as a security deposit that can be penalized or taken away if the user does not act honestly. The work that users who stake have the right to do comes in two forms:
· Users who stake can opt to run transcoding nodes on the network. These nodes contribute significant hardware (CPU & GPU) and bandwidth, and can charge broadcasters for the service of transcoding their live videos into all the formats and bitrates requested, and then serving them to the network.
· Users who do not wish to run their own transcoding on the network can instead delegate their stake towards other candidate transcoders (or orchestrators). These users are providing the service of vetting and QA’ing all the transcode candidates to choose honest, high performing, cost effective ones.
Users who do work on the network will earn inflationary LPT that the protocol mints each round. They will also earn fees that are charged to broadcasters in Ether. They will also generate more LPT which will route a higher proportion of work towards them in the future, should they choose to stake it.
Newly generated tokens in Livepeer are distributed to bonded nodes in relative proportion to the amount of tokens they have bonded (minus fees), as long as they’ve delegated towards transcoding nodes that behave according to the protocol. Bonds can be slashed (reduced by a certain percentage) if the nodes that they’ve delegated towards do not behave and violate one of the slashing conditions. Nodes who have bonded and delegated towards a transcoder also receive a portion of the fees that the transcoder generates through transcoding jobs on the network. In essence, nodes who perform work earn the fees that broadcasters pay for that work.
In summary, participants choose to bond their stake for the following reasons:
· Participate in delegating towards effective transcoders who will provide great service to the network, ensuring its value to broadcasters
· Build reputation and future-work allocation in form of allocated token in proportion to stake
· Earn fees generated from transcoders
· They may wish to be a transcoder
Current (Estimated) Distribution
11,021,918 LPT tokens were circulating as of November 30, 2018, implying 1,021,918 LPT tokens have been created since the genesis block. Based on internal analysis, we estimate the current distribution of LPT tokens to be as follows (with explanations provided):
An explanation of how we estimated the current distribution for each category is provided below. For purposes of simplicity, we assumed no LPT tokens that were put to work were unbonded to date.
Founding Team
We begin with the 1,240,000 owned by the founding team. We then modeled out the 36-month vesting schedule starting from April 30, 2018 (when the network launched) through the end of March 2021. We estimated 34,444 LPT tokens (2.8% of the founding team’s initial supply) are vested each month. Through the end of November 2018, the cumulative vesting amount was deemed to be equal to ~22.2% of the founding team’s initial supply, or approximately 275,556 LPT tokens. We then assumed the entirety of these vested tokens were bonded as we believe it rational that the network founders would choose to support their own network. As will be detailed later herein, we estimated the LPT inflation rewards collected by the founders to be 118,272 LPT tokens as of the date of this analysis, resulting in a total LPT token ownership of 1,358,272.
Pre-Sale Investors
We begin with the 1,900,000 owned by the pre-sale investors. We then modeled out the 18-month vesting schedule starting from April 30, 2018 (when the network launched) through the end of September 2019. We estimated 105,556 LPT tokens (5.6% of the pre-sale investor’s initial supply) are vested each month. Through the end of November 2018, the cumulative vesting amount was deemed to be equal to ~44.4% of the pre-sale investors’ initial supply, or approximately 844,444 LPT tokens. We then assumed 60.0% of these vested tokens (844,444 * 60% = 506,667) were bonded per guidance from the Livepeer team. As will be detailed later herein, we estimated the LPT inflation rewards collected by these pre-sale investors to be 217,467 LPT tokens as of the date of this analysis, resulting in a total LPT token ownership of 2,117,467.
Early Advisor Grant
We begin with the 20,000 LPT token supply allocated to early advisors as a grant. We assumed 100.0% of these tokens were bonded as we believe it rational that early network advisors would choose to support the network they are advising. As will be detailed later herein, we estimated the LPT inflation rewards collected by these early advisors to be 8,584 LPT tokens as of the date of this analysis, resulting in a total LPT token ownership of 28,584.
Thus far, we have estimated 802,222 LPT tokens to be staked by the aforementioned network participants, as shown below:
Per Livepeer’s block explorers, we can view the number of total tokens bonded as of the date of this analysis (November 30, 2018) by active transcoders, which is 2,271,072 as of the date of this analysis, or approximately 20.6% of the total LPT token supply. Thus, the difference (1,468,850 LPT tokens) between this bonded figure and the 802,222 bonded figure estimated above represents the LPT tokens bonded by the remaining active participants of the network: the transcoders, the merkleminers and the public/delegators.
MerkleMiners, Public Delegators & Transcoders
We begin with the 6,340,000 allocated to the public via the merklemine. Per the completion of the merklemine, we estimated merkleminers collected approximately 16.3% of this allocation (1,035,101) while the public collected the remainder (5,304,894). While not a perfect science, for an upper bound (max possible output) we assumed merkleminers collected [0.4 / (0.4 + 2.05) = 16.3%] of the LPT tokens per account mined. While we could have estimated a lower bound [0.337 / (0.4 + 2.05) = 13.75%], as the amount that was allocated to “callers”/miners became 0.337 per LPT token on September 26, 2018 and the merklemine completed on October 3, 2018, we elected to stay conservative and assume the upper bound in this analysis for simplicity purposes to limit arbitrary assumptions. We assumed the public collected the remaining ~83.7%. Per discussions with the Livepeer team, it is our understanding they estimated between 800,000 to 1,100,000 LPT tokens were collected by merkleminers. Thus, while not perfect, we believe our 1,035,101 estimate is reasonable.
By way of iteration, we assumed 74.5% of the LPT token differential was bonded by the merkleminers, while the remainder while the remainder of the differential was plugged into the public. Per discussions with the Livepeer team, it is our understanding that the percentage of LPT collected by the public as bonded is likely closer to 0%-1% as of the date of this analysis and that many would consider those effectively “burned” (more just dormant and will be unlocked if the value is high enough to justify it) as they were delivered in ~2.3 LPT denominations to millions of Ethereum addresses.
For purposes of this analysis, we ran an iteration that assumed the merkleminers could not bond more LPT tokens than originally received (1,035,101), with the remainder plugged into the public. The result implies the public bonding rate is 6.7% of the LPT tokens received. While we realize we may be slightly overstating the amount bonded by the public (relative to the Livepeer team’s estimated 0%-1%), we nonetheless wanted to remain consistent in calibrating the amount of LPT tokens the merkleminers could bond with the amount of LPT tokens the merkleminers received as of the date of this analysis.
Given not all merkleminers were transcoders, and not all transcoders were merkleminers, we had to separately estimate the allocation of LPT token supply that accrued to transcoders to date. By taking the latest round (1180 as of November 30, 2018), we observe the following transcoder economics:
Given we know 1,021,918 LPT tokens have been created since the genesis block (discussed earlier herein), we next applied the above-shown transcoder economics to the minted 1,021,918 LPT tokens to estimate the inflation rewards collected by the active network participants since the genesis block. For purposes of simplicity, we assumed that these transcoder economics (e.g., reward cut, fee share & GWEI charged) have stayed static since inception. Per guidance from Livepeer, it is our understanding there were 10 transcoders for a period of time before the number of transcoders increased to 15. We also understand this number is expected to change over time and that it was initially arbitrarily selected. For simplicity purposes, we assumed 15 transcoders since inception with static economics matching the latest round (1180 as of November 30, 2018). Per historical monitoring, we do not believe transcoder economics change materially round-over-round. Additionally, we understand our bonding amounts for the founding team and the pre-sale investors are overstated in this calculation as per their respective vesting schedules (we computed the inflation rewards accrued to these participants based on the total vested tokens to date rather than incrementally computing the inflation rewards based on their respective monthly vesting allowances). Again, this was done for simplicity purposes coupled with the fact we do not believe the final numbers to be materially different than the estimates computed in this analysis. We then multiplied these minted LPT tokens by the reward cut and by the total LPT staked/bonded for each transcoder, to derive the following estimates of inflation rewards collected by each active network participant via delegation:
A quick rundown of some sample calculations:
Transcoders
Transcoder No.1 charges a 5.00% reward cut, as shown previously. Multiplying this 5.00% reward cut by the total staked by this transcoder (~12.88%) yields a product of ~0.64%, representing the total portion of the inflation rewards (1,021,918) collected by Transcoder No.1. This computation (1,021,918 * ~0.64%) equals 6,580 LPT tokens. Another way to estimate this is to rearrange the variables: 6,580 LPT / 1,021,918 LPT = ~0.64% collected by Transcoder No.1. If we divide this amount by the 12.88% bonded/staked amount by Transcoder No.1, we should arrive at the reward cut Transcoder No.1 charges: (~0.64% / ~12.88% = ~5.00%). We repeat this exercise for the rest of the transcoders and determine that transcoders have collected an estimated 80,294 LPT tokens to date, which we place in our current distribution table:
Delegators
As Transcoder No.1 charges a 5.00% reward cut, the delegators earn the remainder of the rewards generated by Transcoder No. 1 (95.00%). Repeating the same math as describe above: multiplying this 95.00% reward cut by the total staked by this transcoder (~12.88%) yields a product of ~12.24%, representing the total portion of the inflation rewards (1,021,918) collected by the delegators. This computation (1,021,918 * ~12.24%) equals 125,017 LPT tokens (note we are slightly rounding hence the exact numbers may slightly differ). Another way to estimate this is to rearrange the variables: 125,017 LPT / 1,021,918 LPT = ~12.24% collected by the delegators. If we divide this amount by the 12.88% bonded/staked amount by Transcoder No.1, we should arrive at the reward cut earned by the delegators: (~12.24% / ~12.88% = ~95.00%). We repeat this exercise for the rest of the delegators and determine that delegators have collected an estimated 940,307 LPT tokens to date, as shown previously.
We next created a pro-rata allocation to determine the amount of LPT tokens collected by each network participant that resembles a delegator:
*Calculated by taking the aforementioned difference in bonded LPT tokens (1,468,850) between total LPT tokens bonded (2,271,072) and the amount we estimated were bonded (per vesting schedules) by the founding team, pre-sale investors and early advisors via grant (802,222), reducing that difference by the amount we computed for transcoders (80,294), and then multiplying the result by the 74.5% allocation assumption (via iteration, discussed previously), with the remainder plugged into the public (as shown below):
Inactive LPT Tokens
This represents a small portion of leftover tokens from the merklemine as well as inactive tokens held by small inactive transcoders per the Livepeer block explorers that are deemed de minimis for purposes of this analysis.
Now that we’ve walked through all of the components, we can revisit the current distribution table we showed previously:
The last two remaining components are the merkleminers and the public.
MerkleMiners
As noted previously, we estimated the merkleminers to have collected 1,035,101 LPT tokens as a result of actively participating in the merklemine. With an estimated 1,035,101 LPT tokens bonded (74.5% of the active transcoder bonded token differential, solved for via iteration, and 100.0% of the original amount of LPT tokens received by the merkleminers), we computed the LPT inflation rewards collected by merkleminers to be 444,278 as of the date of this analysis, resulting in a total LPT token ownership of 1,479,379.
Public (Additional Delegators)
As noted previously, we estimated the public to have collected 5,304,894 LPT tokens from the merklemine. With an estimated 353,454 LPT tokens bonded (25.5% of the bonded token differential, solved for via iteration, and 6.7% of the 5,304,894 starting amount), we computed the LPT inflation rewards collected by public to be 151,707 as of the date of this analysis, resulting in a total LPT token ownership of 5,456,601.
Inactive Tokens
As noted previously, we estimated a very small amount of LPT tokens to be inactive in the ecosystem.
With the (estimated) current distribution now identified, we can move on to model the economic returns of active network participation (aka “generalized mining”) in the form of transcoding.
Modeling Generalized Mining from a Fund’s Perspective: A Livepeer Case Study
In this section, we model a hypothetical fund’s active network participation / generalized mining economics in the Livepeer network through the lens of a fund operating one transcoder. Generalized mining can generally be thought of as bootstrapping early-stage networks by provisioning various types of supply-side services to meet the demand side such that decentralized networks can launch themselves and organically create network effects with the support of non-Byzantine actors. The purpose of this exercise was to better assess the economic impact of operating a transcoder in the Livepeer network.
*At the time of this analysis, the LPT token was trading around $2.62 on Radar Relay. It should be noted that we do not believe Radar Relay’s LPT token value of $2.62 is a viable representation of fair value due to the very small trading volume. The foregoing price was nonetheless used in this analysis for calibration and illustrative purposes as well as for enabling price sensitivity for analytical purposes. Source: https://app.radarrelay.com/LPT/WETH.
Note (A)
We assume this hypothetical fund operates in the Livepeer network as a transcoder and has also previously operated as a merkleminer, and therefore estimated the participation specific to this hypothetical fund by taking the current estimated distributions of the merkleminer and transcoder supply (1,479,379 and 80,294, respectively, as of the date of this analysis) and dividing it by the total bonded tokens among the active transcoders (2,271,072). This produces a participation of approximately 68.6%, which we rounded to 2/3 participation for simplicity purposes (with delegated participation comprising the remaining 1/3).
Note (B)
Estimated per the projected inflation schedule of Livepeer. Using 0.0329% inflation as of July 4, 2018 (the first data point available in this query), we added 0.0003% inflation for each round thereafter per the Livepeer protocol to derive a projected inflation schedule. Using the following sources for supply and bonding, we derive actual inflation and the bonding participation rate to date, and simple extrapolate the historical data on a go-forward basis.
Supply (Source)
https://supermax.cool/livepeer/mainnet/ethcall/0x58b6A8A3302369DAEc383334672404Ee733aB239/18160ddd
Bonding (Source)
https://supermax.cool/livepeer/mainnet/ethcall/0x511Bc4556D823Ae99630aE8de28b9B80Df90eA2e/5c50c356
As of the date of this analysis (November 30, 2018), according to our inflation schedule, we estimated the current 20.6% bonding rate to grow into a 50.0% bonding rate by June 28, 2019, illustrated as follows:
Once the network reaches 50.0% bonding participation, it is our understanding the inflation rate will fluctuate as follows:
· If the bonding rate exceeds 50.0%, the inflation rate will contract by 0.0003% per round.
· If the bonding rate declines under 50.0%, the inflation rate will increase by
-0.0003% per round.
Per guidance from Livepeer, it is our understanding that inflation rewards are not always entirely claimed each round by transcoders. If there is a case where a certain number of inflation rewards are not claimed by a transcoder, they are never generated by the Livepeer network. Thus, the actual supply has historically been lower than the maximum possible supply due to an unclaimed portion of inflation rewards never being generated by the network.
All else equal, the higher the bonding rate, the lower the inflation reward becomes, and the less incentive there is to bond (generally), thus participants are expected to start unbonding (a withdrawal period that lasts 7 days currently) which in turn will decrease the participation rate. If this goes below 50.0%, the cycle resets (the inflation rate per round reverses to the upside and bonding incentives become reignited). It is expected an equilibrium will be reached over time.
It is worth noting that the aforementioned equilibrium prediction ignores Ether, DAI (or other) fees that currently do not flow through the Livepeer ecosystem. All else equal, as fee earnings tick up over time, participants will have incentives to collect the fees rather than only the inflation rewards, so they may not necessarily unbond as much as predicted. In equilibrium, participants will likely earn low inflation but collect high fees.
As of the date of this analysis (November 30, 2018), we projected the network’s inflation rate to reach approximately 0.13%-0.15% by the time the network hits 50.0% bonding participation, shown as follows:*
It should be noted that the volatility portrayed in the royal blue line is driven by the fact we are mixing calendar rates with rounds. There are 5,760 blocks per round (one block every 15 seconds), but due to current compute power supply sometimes the network accelerates and computes 2 rounds in a given calendar day (one block every 13–14 seconds), essentially doubling the inflation effect in the process.
Note (C)
Calculated by taking the hypothetical fund’s estimated $20.0 million AUM and multiplying it by the assumed ~2.0% (1.985%, solved via iteration to calibrate to market price) portfolio allocation to Livepeer. ($20.0 million x ~1.985% = $397,000). It is worth noting this economic participation relates to only one transcoder that we assume our hypothetical fund manager operates.
Note (D)
Per inflation schedule, also found by observing the Livepeer block explorers: https://explorer.livepeer.org/transcoders; https://supermax.cool/livepeer
Note (E)
Calculated by taking the hypothetical fund’s estimated 10.0% participation and multiplying it by both the current bonding participation across active transcoders (approximately 20.6%) and the estimated current supply (11,021,918).
11,021,918 x 10.0% participation x ~20.6% current network bonding participation = 227,400 (numbers may not perfectly tie due to rounding).
Note (F)
Calculated by dividing the figure in Note (A) by the estimated current supply.
151,600 / 11,021,918 = ~1.4%
Note (G)
Calculated by dividing the figure in Note (C) by the figure in Note (F).
$397,000 / ~1.4% = $28,863,453 (estimated Livepeer network value; numbers may not tie due to rounding)
Note (H)
Calculated by dividing the figure in Note (G) by the estimated current supply (11,021,918).
$28,863,453 / 11,021,918 = ~$2.62 (estimated LPT token value as of the date of this analysis (November 30, 2018)).
Also calibrated to listed price on Radar Relay as of November 30, 2018 (https://app.radarrelay.com/LPT/WETH). It should be noted that we do not believe Radar Relay’s LPT token value of $2.62 is a viable representation of fair value due to the very small trading volume. The foregoing price was nonetheless used in this analysis for calibration and illustrative purposes as well as for enabling price sensitivity.
Note (I)
Per our inflation schedule, this is the estimated supply when the network hits 50.0% bonding participation by June 28, 2019.
The final step needed before sensitizing the network’s economic returns is to estimate the inflation rewards (in LPT token-denominated terms) earned by the Fund on behalf of its transcoding work.
As of the date of this analysis (November 30, 2018), we estimated the network’s inflation rate to be 0.0673%. Assuming the inflation reward remained constant until the network hits 50.0% bonding participation, the LPT token-denominated yield earned by the Fund amounts to 15.18%.
It should be noted that the aforementioned estimate assumes a constant inflation rate going forward (e.g, that the inflation rate will remain at 0.00673%). Realistically, this is not expected to be the case. Thus, we experiment with a different methodology to derive an estimated inflation yield. Using time value of money (as we understand the inflation rate is not likely to remain constant), we estimate the LPT token-denominated yield earned by the Fund to be 50.56%.
For purposes of our analysis, we assumed the midpoint between the two inflation yields determined by different methodologies (15.18% and 50.56%). The average amounted to 32.87%. This means that our Fund is expected to earn a 32.87% inflation yield (in token-denominated terms) on behalf of its transcoding work between the date of this analysis (November 30, 2018) and the date in which the network is expected to hit 50.0% bonding participation (June 28, 2019).
We made the assumption that the subject hypothetical fund manager charges a 35.0% reward cut for the transcoding work it performs; as such, we compute the following supply when the network is expected to hit 50.0% bonding participation:
As a percentage of the total LPT token supply when the network is expected to hit 50.0% bonding participation (13,947,839), our hypothetical fund’s network ownership amounts to ~1.5%, implying that as a result of being an active network participant, our hypothetical fund manager not only avoided getting diluted by the inflation supply, but was also able to grow its stake within the Livepeer network by 10 basis points.
In consideration of the foregoing, we sensitized the economic returns of our hypothetical fund manager’s active ecosystem participation around the current price point of $2.62 by $0.25 increments, and measured the impact of the extra 10 basis points of network ownership our hypothetical fund manager earned. In particular we sensitized the LPT token price with a range of $1.00 to $10.00. By way of iteration, we identified our hypothetical fund manager’s breakeven MoC to occur at an LPT token price of $1.10, all else equal as of the date of this analysis.
As can be seen, our hypothetical fund manager is able to boost its return to the upside as a result of extracting the inflation rewards from the Livepeer network by actively participating within it. Additionally, as will be discussed in greater detail later herein, our hypothetical fund manager is also able to lower its cost basis relative to a passive (non-participative) investor. A graph illustrating the return our hypothetical fund manager earns as the LPT token price rises from $1.10 (the LPT token price in this sensitivity where our hypothetical fund manager has a 1.0x MoC — solved for via iteration) to $10.00 presented below.
As can be seen in this chart, our hypothetical fund manager is able to boost its investment returns (earning the green MoC return, aka the delta) by being an active participant in the Livepeer network relative to a passive holder (that earns the blue MoC returns). If we start with the $1.10 LPT token price, the implied Livepeer network value based on current supply (11,021,918) is $12,104,030. Using this same $1.10 LPT token price when the Livepeer network is expected to hit 50.0% bonding participation (13,947,839 token supply), we calculate a Livepeer network value of $15,317,212. As calculated previously, our hypothetical fund manager is projected to grow its stake within the Livepeer network by 10 basis points from ~1.4% to ~1.5% by the time the Livepeer network hits 50.0% bonding participation. Multiplying the network value based on current supply assuming a $1.10 LPT token price ($12,104,030) by the ~1.4% current ownership yields an observed investment value of $166,484 whereas multiplying the network value based on the future supply at a 50.0% bonding participation assuming a (flat) $1.10 LPT token price ($15,317,212) by the ~1.5% future projected ownership yields an investment value of $227,552. Thus, with a flat LPT token price of $1.10 (a MoC for our hypothetical fund manager equal to 1.00x in this sensitivity, as solved for via iteration), our hypothetical fund manager in this case is able to earn a 1.37x MoC by collecting the inflation rewards as a result of performing transcoding work on behalf of the network. The power trendline (in yellow dashes) clearly demonstrates the increasing acceleration of upside return per each incremental unit increase in LPT token price ($0.25).
This is not surprising, because in crypto networks, there are financial incentives whereby a global community is catalyzed to pull together and provision a decentralized digital good or service, self-organize in a specific way and then remunerate for either the contribution of that resource or pay for a specific resource on a network. In the case of Livepeer, transcoders are reformatting video content and collecting and distributing the inflation rewards and fees accrued in each block to all active network participants. Thus, it makes sense that transcoders should be rewarded for the work they are performing on behalf of the network, boosting their return potential in the process (as shown previously).
Finally, we analyzed the reduction in cost basis that our hypothetical fund manager earns as a result of collecting inflation rewards in the Livepeer network. If we recall from traditional finance the dividend discount model, which is a method of valuing a company’s stock price based on the theory that its stock is worth the sum of all of its future dividend payments, discounted back to their present value, we can assess the present value of each dollar of LPT token put to work in the Livepeer network.
The cap rate in this analysis is akin to the inflation yield an active participant earns in the Livepeer network. As the cap rate is a blended rate of return used to convert a point estimate of cash flow into value, the inverse of the cap rate is akin to the multiple earned on a given investment:
Thus, taking the inverse of each multiple on investment our hypothetical fund manager earns per dollar of LPT tokens put to work to earn inflation rewards, produces the estimated cap rate (i.e., the inflation yield, aka the discount rate) earned by our hypothetical fund manager on each dollar of LPT tokens at work. Thus, with a dollar worth of LPT tokens being the final value, we can back solve for the present value (aka, our hypothetical fund manager’s cost basis) of each dollar worth of LPT tokens being put to work to earn inflation rewards.
As can be seen, the cost basis of each dollar at work in the Livepeer network declines as the LPT price rises as a result of the inflation rewards active participants earn in the network. It should be noted that this analysis was based on one transcoder, and thus slightly limited in scope. Nonetheless, we believe our hypothetical fund manager’s active network participation / generalized mining investment strategy is evidently effective.
Grateful thanks to Doug Petkanics, Raffi Sapire & the rest of the Livepeer team, and especially Jake Brukhman for all their help & feedback with this analysis.
— — — — — — — — —
Vision Hill Advisors is a crypto asset and blockchain focused fund of funds. Through a proprietary fund manager selection process and institutional level due diligence, Vision Hill aims to lead investors into the future of digital assets. Vision Hill brings together a team with extensive experience in traditional financial markets, a deep passion and understanding of crypto and digital asset markets, and a history of risk and portfolio management.
Vision Hill is indirectly invested in Livepeer. The content provided herein should not be considered investment advice, and is not a recommendation of, or an offer to sell or solicitation of an offer to buy, any particular asset, security, strategy, or investment product. Vision Hill has not received and will not be receiving any compensation as a result of this report.
