The Rise of Staking: From Theory to Building Large Infrastructure

• Staking is the process of holding funds to support the operations of a blockchain network through a reward-driven process. Staking has been one of the unique differentiating features offered in crypto-industry.

• Mimicking passive-income strategies in traditional finance, staking allows users to earn rewards directly on-chain for holding a given currency.

• Staking is available on many chains in many forms, but the precise mechanism depends on the staking project’s token design and consensus mechanism.

• However, staking yields cannot be considered in a vacuum - instead, they must be benchmarked against other chains based on potential risks and obligations like:

  • Different chains pay out rewards with various restrictions and payout timings.

  • Participants must address the technical risk of failure as well as their exposure risk to the underlying coin when deciding to stake long-term on a chain.

• Participating in different networks requires different resources and sometimes additional labor to collect rewards.

• With staking, existing players in the crypto-ecosystem and blockchain projects have aligned incentives continued symbiotic growth. As more infrastructure players support staking, the entire ecosystem will be able to mature.

• With Ethereum’s impending switch to Proof of Stake (PoS), staking is expected to take a most substantial portion of the market’s attention.

Staking promises rewards and other features (e.g., governance, transparency) that mirror incentive mechanisms in traditional finance. Like conventional financial products, staking relates to concepts such as currency risk, liquidity parameters, and interest rates.

In this report, staking is explained and defined. Afterward, some of the most compelling projects and mechanisms in the ecosystem are introduced. Furthermore, the importance of staking from the perspective of coin-holders and market participants is analyzed, before discussing in-depth other considerations and risks related to staking. Finally, market players are defined, along with the evolution of staking providers. Has staking's ecosystem matured enough for widespread adoption?

1. Definition(s) of staking

1.1 History and definitions

Binance Academy defines staking as:

“Staking is the process of holding funds in a cryptocurrency wallet to support the operations of a blockchain network. Essentially, it consists of locking cryptocurrencies to receive rewards.”

In its article, Binance Academy points out that in most cases, the process relies on users participating in blockchain activities through a personal crypto wallet (such as Trust Wallet or Coinbase Wallet).

Theoretically, projects utilize staking rewards as an incentive mechanism to encourage users to participate in the consensus finding of the chain actively. In practice, rational users are expected to be profit-seeking and therefore, must assess the potential rewards of participating in the ecosystem of each coin. However, staking rewards are typically denominated in a non-fiat pegged cryptocurrency, carrying potential volatility risks (which some of them will be discussed in the next subsections).

Historically, staking was first introduced as a concept in Peercoin (PPC), developed by Sunny King and Scott Nadal, who are now developing V Systems. Despite Peercoin beginning as a hybrid Proof-of-Work/Proof of Stake chain, it slowly transitioned to phase out any PoW rewards, thus solely relying on its PoS power.

There are several implementations for Proof of Stake algorithms, including several alternatives such as DPoS. At its core, the Proof of Stake algorithm is defined by Binance Academy as follows:

“The Proof Of Stake algorithm uses a pseudo-random election process to select a node to be the validator of the next block, based on a combination of factors that could include the staking age, randomization, and the node’s wealth.”

Fundamentally, “minting” a block in a Proof of Stake algorithm can be compared to “mining” a block in Proof of Work systems. To become eligible for minting blocks and consequently reaping block rewards, validators must stake cryptoassets on-chain. Depending on the chain details, these staked cryptoassets can be cut (in jargon “slashed”) to penalize malicious or unwanted behavior. Moreover, having locked (illiquid) coins in the network creates vested interests, as attackers are potentially dissuaded from performing any actions that could harm the value of the coin while they still hold this locked deposit denominated in the same cryptoasset.

Another variant of Proof of Stake is Delegated Proof of Stake (DPoS), the first version of which was displayed in BitShares, the first project created by Dan Larimer (who is also the creator of Steem and EOS). In addition to BitShares, other early projects supporting DPoS include Steem, Lisk, and Ark.

Binance Academy defines DPoS such as:

“A DPoS-based blockchain counts with a voting system where stakeholders outsource their work to a third-party. In other words, they are able to vote for a few delegates that will secure the network on their behalf.”

These delegates - also referred to as “witnesses” - are responsible for achieving consensus during both the generation and the validation of new blocks. In DPoS algorithms, voting power is (usually) proportional to the number of coins each user holds.

From an analytical perspective, stake-able coins can be segmented into five core families:

• Proof of Stake (PoS) like Algorand.

• Delegated Proof of Stake (DPoS) with assets like ICON and EOS.

• Distribution model with assets like Stellar.

• Dual-coin systems with assets like NEO/GAS.

• Masternode with assets like Dash, TomoChain, and ZCoin.

1.2 The diversity of staking cryptoassets

In this subsection, examples for each of these five subgroups are described to illustrate the variety of stake-able cryptoassets.

1.2.1 Algorand: a pure Proof-of-Stake Network

Algorand employs a pure proof-of-stake (PPoS) consensus protocol whereby the users’ influence on the production of a new block is proportional to their stake (denominated in ALGO). Featuring one-block-finality, the chain prioritizes randomization in the sets of consensus participants and allows users to directly earn staking rewards without the need of any intermediaries. The project provides an initial 1.75B in tokens for a staking rewards pool, after which transaction fees will be utilized to replenish the staking rewards pool.

1.2.2 EOS: a DPoS giant

EOS is a cryptocurrency that was created to support large scale applications, without any transaction fee. EOS relies on its unique resource-sharing model (i.e., CPU, RAM, and NET/network resources1) with a lending/borrowing peer-to-peer lending platform named EOSREX. EOS features a 5% annual network inflation, with 1% allocated to block producers who stake and receive the delegation, and another 4% allocated for future utilization for the growth of the overall ecosystem.

1.2.3 Stellar: is inflation an easy money sniper?

Since its network inception, Stellar (XLM) has offered what it terms inflation, a steady 1% inflation rate to the supply on the whole network, which allows its network to maintain low transaction fees, as stakers collect the inflation rewards weekly at 12 AM GMT every Tuesday.

Stellar’s inflation rewards were initially designed to be easy to obtain, but not too high of an interest rate to dilute the total value of supply of the network. The initial intention was to simultaneously attract new platform users, as well as encouraging an active usage of the coin by introducing opportunity costs for merely hoarding coins.

However, Stellar has recently proposed an upgrade that, if ratified, may eliminate its inflation, citing a failure of fostering development and rewarding of positive actors within the Stellar ecosystem. Instead, Stellar has rolled out some unique airdrops, such as the one to Keybase users and Github users, targeting developers and decentralized technology users.

1.2.4 NEO and GAS: two-headed monsters

NEO’s rise to fame after its rebrand from Antshares largely stemmed from its introduction of the two-coin model, a model in which a gas coin, creatively named GAS, was issued to NEO holders as a reward.

The issuance of GAS, while dilutive at the start, is capped with a fixed maximum supply of 100 million coins. As a result, the issuance of the second coin slowly decays overtime for the supply to converge to a finite number.

This token economic model was unique for two reasons - firstly, it created a new system in which block rewards would not dilute the coin that provides the ability to earn the block rewards. Secondly, the model allowed for a separation between a pure utility token and a token whose value is derived from the ability to produce or earn the utility token.

Owing to its popularity, this model was replicated in other projects such as Ontology’s ONG, Vechain’s VET, and Theta Network’s TFUEL.

1.2.5 Dash: privacy coin with a masternode operating system

Dash, a privacy coin that was formerly known as Darkcoin, initially started as Bitcoin code-fork, but became probably most notable for popularizing the concept of a “masternode”.

Every masternode requires “bundles” of 1,000 DASH to operate and participate in the consensus mechanism. This model set the standard for tokens to be created as an “access token” to partake in the consensus, and commensurately earn “fruits for labor” provided to the network.

Currently, there are nearly 4,000 masternodes (MN) on the Dash network. Relying upon its masternode network, Dash has also unveiled an “Instant Send” function to safely accept transactions right away upon receipt, thus no longer requiring a multi-block time to consider transactions final. The project also implemented ChainLocks based on “Long Living Masternode Quorums” to mitigate the risks of 51% and double-spending attacks.

Generally, this masternode set-up allows for a fundamentally rule-based, yet open system network. Owing to its significant entrance barrier of having to own 1000 Dash, large token holders are aligned with all network participants. Hence, bad behavior from MNs supposedly would lead to a reduction in the asset value, i.e., the price of Dash.

Following Dash’s innovative model, many other coins rely on masternode models today with notable projects like Zcoin (XZC), Horizen (ZEN), PIVX, and Syscoin (SYS).

1.2.6 Ethereum: the largest elephant in the room

Perhaps the largest elephant in the staking room is Ethereum, with its pending switch to Proof of Stake being long-coming. With Ethereum Classic (ETC) firmly rooted in Proof-of-Work, Ethereum has many research camps developing various staking-powered mechanisms. To date, the Ethereum Foundation and related Ethereum stakeholders have funded almost $10 million worth of grants for Ethereum 2.0 solutions. Most of these solutions have been designed to make Ethereum more scalable while packing in more features such as privacy. With Ethereum being the second-largest market cap coin and one of the longest-running blockchain products, it could drive many projects to follow and switch to Proof-of-Stake.

In the next section, an overview of the most significant projects is mapped and put in perspective with the overall total industry market capitalization. Furthermore, some of the essential requirements for each chain are introduced.

1.3 Overview of the largest assets supporting staking

As of October 24th 2019, the largest 10 cryptoassets supporting (or planning on supporting) staking represent a cumulative market capitalization of $25.8 billion.

Table 1 - Overview of the largest cryptoassets for staking (data as of October 24th 2019)

Sources: Binance Research, StakingRewards, CoinMarketCap

According to StakingRewards, the highest yields include projects like Synthetix (61.9%) and Livepeer (102.7%), as of October 28th 2019.

Chart 1 - Staking yields (%) denominated in respective cryptocurrencies for largest marketcap cryptoassets

Sources: Binance Research, StakingRewards. Data as of October 28th 2019.

Synthetix Network displayed the highest yield amongst the fifteen largest cryptoassets which support staking. However, higher staking yields may not necessarily provide benefits from the perspective of users, as discussed in the next subsections.

Chart 2 - Comparison between volumes locked in DeFi vs. staked amount (in USD million)

Sources: Binance Research, DApp Total, StakingRewards. Data as of October 25th 2019.

Excluding Ethereum, the cumulative staking market capitalization, as of October 24th 2019, is worth around $11.2 billion with around $6.4 billion being staked.

Chart 3 - Cumulative staking participation rate across all blockchains offering staking

Sources: Binance Research, StakingRewards. Data as of October 25th 2019.

Furthermore, in comparison with the total industry market capitalization, it represents a staking dominance of around 5%.

Finally, there are large differences amongst staking ratios between blockchains. Staking ratio is defined as the ratio of the amount staked at a single point in time divided by the total circulating supply of the cryptoasset.

Chart 4 - Staking ratios across largest blockchains supporting staking2

Sources: Binance Research, StakingRewards. Data as of October 28th 2019.

As illustrated by chart 4, Synthetix Network showed the highest staking ratio across the largest fifteen blockchains by market capitalization. Regarding assets listed on Binance, Algorand, Tezos, and Cosmos displayed the high staking ratios (>70%) among all the largest blockchains supporting staking features. On the other hand, coins like TRON or Qtum exhibited a staking ratio of under 25%.

The next section discusses the “staking dynamics”, i.e., the key parameters to consider from the perspective of all stakeholders.

2. Staking dynamics: understanding the parameters that go into staking

In this section, we introduce the individual user’s staking dynamics: when deciding whether or not to stake a coin, users should be considering not only the rewards but also all the risks & restrictions and obligations associated with active participation in staking.

2.1 Staking rewards

Instead of block rewards being awarded to miners proportional to their hashpower in traditional Proof-of-Work blockchains, rewards are distributed to PoS participants, often proportionately to users (or probabilistically proportionately) who stake tokens on the network.

These rewards come in all shapes and sizes, with some systems having maximum rewards caps, requiring periodic claiming, lock-up periods (similar to Bitcoin coinbase reward time), as well as the frequency of compounding rewards. For example, Komodo caps rewards for wallets that have not made a “claiming transaction” within the last month, thus suggesting that the user is inactive and not participating in the ecosystem. Therefore, it is imperative to claim at least once a month. For projects like Stellar, which issue their inflation rewards every Tuesday at 12 AM GMT, rewards are automatically sent without requiring a new claim transaction, but the rewards compound only weekly.

Along with staking rewards, participants in staking may also receive other rights or access, depending on the structure and governance of a given cryptocurrency. While they may often be displayed or advertised as APY (annual percentage yield), beyond these rewards, a whole range of additional issues is bundled together, as not all blockchain rewards are created equally. Rewards do not come without any restrictions or concerns. Precisely, these costs of participating in staking (discussed in subsection 2.3) are impacted by obligations and requirements (paragraph 2.2).

2.2 Obligations and requirements

In this subsection, both the obligations and requirements for staking participants are discussed.

2.2.1 Required exposure to governance

Staking participants are reliant on the many parameters of staking, which are unique to each blockchain. Thus, they are positively incentivized to speak up in governance and vote for decisions regarding the chain and its ecosystem.

• Governance as a burden: However, the right to participate in the governance process may present a real burden on participants (for which they earn the staking rewards or can protect their interests as staking reward earners). For example, a staking participant may need to attend weekly governance calls (e.g., MakerDAO), participate in online discussions and forums, as well as communicate with other nodes to develop chain-wide standards and best practices.

• Collusion prevention: to prevent collusion or other malicious actions that may harm the health of the chain or its staking rewards, participants must be active in ensuring the positive nature of the entire chain ecosystem, which is not always an easy task.

• Entrance barriers: in DPoS networks, where many validators may be jockeying for position, some validators may need to advertise or otherwise spend to publicize their efforts for candidacy in a system. Large token holders can typically exercise a proportionate amount of power in the decision-making of the community, and thus present a significant level of control over the chain’s governance. If users are outside of this large holdings circle, they may face some risk of being exposed to the whims of other (larger) staking participants.

2.2.2 Potential fees for a node application

The process for being eligible to become a node runner can vary drastically and may include considerable costs. Depending on the chain, there may be an application form, a registration fee, as well as a potential bond (or staked deposit) that validators must post prior to being considered for a validation spot. The effort required to pass these hurdles is another cost that must be considered when reviewing staking projects. For example, it costs 9,999 TRX to apply to become a Tron Super Representative, and 2,000 ICX to apply to become an ICON P-Rep.

2.2.3 Operational costs

For most of the PoS systems, there are physical hardware costs (such as a Ledger, HSM, or another signing). Additionally, there are costs to run masternodes which relate to general resources such as RAM, CPU, or other “resource commodities”. These resources may be provided directly via physical machines and servers, or via cloud providers. For example, the ICON foundation recommends an AWS C5.9xlarge instance, which currently costs $1.53/hour. This prorates to $13,400 per year of operating a node. As resource exchanges like EOSREX and file storage and decentralized computing protocols like Sia, Storj, and Golem gain popularity and capacity, it may make it easier for users to acquire these resources in a decentralized manner.

2.2.4 Technical understanding

Obligations of staking participants are not exclusively economical but may extend to having a certain technical expertise.

• Operating nodes: first and foremost, participants in the network’s consensus must operate nodes. Running nodes typically require general resources of varying degrees. Even if one has these assets already or has them readily available, with the introduction of resources exchanges such as EOS Rex, and Trontrade.io, there are true opportunity costs that can be calculated for deploying these resources to stake.

• Uptime requirements: node operators are usually required to have near-constant uptime. In some cases, violation of these requirements may also carry penalties. Cosmos, for example, has “hard slashing” for anyone who double-signs or is offline for a significant period of time. This mechanism ensures that validators are indeed actively participating in the consensus in a consistent manner.

• Additional security measures: in some cases, running a node may require additional security measures that may cost more than simply just raw computing resources. For example, users may even have to run their own segregated set-ups, complete with VPNs, firewalls, private connections, and other technological considerations. Familiarity with cloud service providers may provide individuals another option to host or provide the necessary resources for participation in the network. These cloud service providers may also be helpful if uptime is a crucial metric to consider, as it may make it easier and cheaper to provide redundancy across many providers and geographies.

• User interfaces: the ability to work through command line or programmatic access to the chain may also be a decisive skill, as not all projects allow staking to occur through a friendly GUI interface. To lower this barrier, many infrastructure providers such as Tronscan’s voting interface for Tron Super Representatives, Loom Network’s Plasmachain Dashboard, Atomic Lab’s KMD reward claimer, and many more create one-stop tools help let the average user earn and claim rewards more easily.

• Staking support in wallets: wallets and other existing infrastructure providers may also add additional staking features to their existing offerings. For example, Trust Wallet recently launched its Staking portal, allowing users to now enjoy the power of WalletConnect authentication and full custody of their funds while staking.

2.3 Restrictions and risks

2.3.1 Locking period and liquidity risks

When analyzing prospective staking options, users must check if the chain requires an un-bonding period or other loss of liquidity to count a balance as staked. Specifically, questions such as “Can users move funds while they are staked?” or “how easily can they stop staking at any given time?” must be considered.

• Unbonding restrictions: for example, a chain may allow users to unbond, or unlock their coins from staking delegation, at any time, but would forfeit any unclaimed rewards. Other chains may not allow any instant withdrawals, even if one is willing to forfeit interest earned, requiring all funds processing time to “unfreeze” or unbond. Different unlocking times both present a different time value of funds (a long unbond period in which the user cannot accrue interest is a direct opportunity cost), and a lockup period that renders funds illiquid for a fixed period may lead to other missed investment opportunities during that period.

• The frequency of reward payouts is also something to consider. If rewards are frequent, then there is a greater potential to compound these interests by frequently “re-staking” any new additional coins earned from staking. Conversely, large, instantaneous payouts may lead to discontinuous token distribution charts (e.g., Stellar’s Tuesday GMT snapshots), leading to potential price action or adjusted token dynamics due to the weekly issuance of coins. Furthermore, if rewards are paid out after a significant delay, the discounted value of these coins is lower, thus reducing the attractiveness of the staking rewards.

• Custodianship risk: custody is another significant consideration when discussing staking. For some users, custody is a must, as they cannot afford to trust other third-parties with funds. For ordinary users, custody may be a burden for fear of the risk of losing private keys, or other daily hassles that may pop up. Depending on the chain, some staking mechanisms require funds to be directly held in the wallets or nodes participating in the staking. Others allow for delegation, which may lower the technical barriers to participating in staking. However, the delegation also opens up the ecosystem for more rent-seeking services that provide this service for a fee, which we will discuss in a later section. Further, if funds must be stored in a smart contract to stake on-chain, there may be a technological risk, instead of pure counterparty risk in lending, that a user must consider when staking on-chain.

2.3.2 Opportunity costs

In fiat terms3, users must consider how the chain’s staking reward rate compares relative to rewards for holding other coins.

For example, DAI, a stablecoin built on ETH, should theoretically mimic the expected opportunity cost of not holding ETH (and its corresponding expected appreciation). If ETH is projected to grow by 15% by the end of the year, the equilibrium rate to borrow DAI should be the amount that people should expect to earn from using DAI to purchase a cryptocurrency.

Besides, some market participants may expect to receive better risk-return profiles through active investment and trading than through passive staking, mirroring the dilemma in traditional markets between active and passive investment strategies. For instance, someone planning on holding Komodo long-term may think he can increase his number of KMDs by actively trading KMD over the course of a year by more than the staking reward yield provided on the chain (currently 5.1%).

Furthermore, some macroeconomic models could potentially be considered in assessing whether staking does provide benefits from the perspective of users. The choice of a macroeconomic model can be justified by the fact that operations are similarly financed; instead of relying on external value added activities, currencies and staking coins are financed by inflating the total circulating supply. This unique trait renders a comparison with most established financial assets an odd choice. Hence, the comparison with currencies appear the most comprehensive to understand why high staking yields (funded by inflation increases) may dilute the value, through price reduction dynamics that are in analogy with macroeconomic theories.

Chart 5 - Relationships between interest, forward, inflation, and spot rates

Sources: Binance Research, CFA Institute, Kaplan Financials

According to the Interest Rate Parity Theory, if staking returns are paid in the native cryptocurrency through an increase in its supply, then the interest rate parity states that the expected future spot price of the cryptoassets must be equal to the interest rate differential multiplied by the spot price between two assets.

Let’s consider a model with two assets: the US dollar and a cryptoasset (e.g., Tezos/XTZ) whose staking rewards are paid through consecutive supply increases.

However, based on the International Fisher Effect, it is possible to calculate the expected spot price change such as:

Hence, if a cryptoasset has a staking yield higher than the risk-free interest rate (denominated in USD), this cryptoasset should theoretically experience a future price decrease (in USD) ceteris paribus. However, as illustrated by the high volatility of the cryptoasset class4, the expectations theory often does not hold.

As a result, someone expecting the price of a coin to increase by more than the expected spot price change defined above should start staking this specific cryptoasset.

As a result, it is worth considering that an extremely high staking yield which appears appealing at first may not necessarily be so interesting from the perspective of users, owing to the high inflation rate across the network. Another interpretation could be the fact that high staking yields might be indicative of hidden risks.

Models similar to NEO’s two-coin model are even harder to analyze from a macroeconomic perspective, as one coin is never diluted at all from the staking mechanism, and instead serves as the base currency from which the staking rewards are calculated. All in all, staking yields cannot be considered in a pure black-box fashion as reward payouts affect price dynamics and other endogenous factors that make a coin more or less attractive as an asset to hold.

2.3.3 Environmental and exogenous factors

Environmental and exogenous factors also exist and remain critical aspects that must be considered. They include elements such as:

• Security risks: to participate in staking, some chains require funds to be held in a wallet that is hosted on the same machine as a live node, which may pose security concerns for those who wish to store their funds in cold storage. Other chains may require frequent actions from a wallet, such as signing messages or writing transactions online to collect rewards and participate in consensus and are thus not very compatible with cold wallet set-ups. We will discuss security and technical costs and ramifications from these types of situations in a later section. Other chains, such as Fetch’s staking auction, require users to send coins to a unique staking contract and thus relinquish the direct custody of funds. However, only the wallet that sent the funds may be able to retrieve them, thus ensuring the funds are still secure. Similarly, chains like TomoChain requires users to vote for, through a smart-contract, for masternodes.

• Interest rate uncertainty: with regards to interest rates for staking projects, “what you see may not always be what you get”. Interest rates are not always fixed for all chains - for some chains, the rewards distributed to all stakers is a fixed pool, and thus, the percentage of participating (staked) coins affects the true yield of staking on the network. For any network with fixed aggregate network rewards, the larger the proportion of participants, the lower the staking yield. This is no different from higher hashrates competing for fixed, finite block rewards on a PoW chain. For other networks, such as Komodo, the yield from staking is indeed a fixed percentage on the balance held in the wallet. Even if the stated interest rate is fixed, it may not truely be an exact APR (Annual Percentage Rate). Almost all interest rates depend on one crucial variable: block times. Most interest rates on blockchains are calculated on a per-block basis according to an estimated blocktime (or blockchain difficulty). However, if the chain happens to speed up or slow down because of a fault in its difficulty adjustment algorithm, the effective interest rate will also change in terms of APY (Annual Percentage Yield). Further, projects may introduce additional staking promotions (such as Algorand) that last for specific periods, thus further incentivizing new users, while also changing the rewards calculus. For instance, some chains may make it easy to compound yields automatically. In contrast, others require a specific action to re-stake earned interest, allowing for lower frictions en route to achieving potential maximum APYs.

• Uncertainty of chain functionality: if the chain undergoes any attack in which it may be stalled, this could also hamper potential returns that stakers could earn. Thus, the predictability and reliability of a chain must be considered when considering the reliability of the interest rate itself. This factor may also affect the price performance of the coin as well, as an unstable chain may lose some user confidence and thus the expected network value. Given the uncertainty in chains, we propose a metric of interest per block, which might help in allowing users and developers to provide an accurate estimate of how much rewards are owed to users. This metric is particularly useful in the construction of interest-bearing tokens, such as Compound’s cTokens and bZx’s iTokens, and other DeFi-based abstractions of interest rates.

• “No free lunch”: The most glaring risk that goes into staking is price uncertainty - if the price of the underlying asset that is staked goes down by a percentage higher than the staking yield, then chasing the yield is a fruitless effort. Due to the uncertainty of price (i.e., volatility risk), we can only treat yields as nominal interest rates (APR). As a result, it is essential to consider the token release schedules of each project and understand the amount of circulating supply and other factors that may drive token price dynamics while participating in staking. Users must also be aware that specific attacks that may cause the loss or freezing of staking rewards may impact the value of what can be collected by a staking participant. The technological risk that is presented by on-chain functions may be mitigated or at least covered using on-chain insurance providers - such as Nexus Mutual - and prediction markets5 that explicitly detail the exact situation. For DPoS chains, a delegation service may lose its status of being inside the validator set, and thus may earn fewer block rewards to pass on to its participants. As a result, a change in parameters, such as the number of validators, or minimum token delegation amounts may affect the competitiveness of acting as a validator and thus hurt the reliability of earnings for delegators. Therefore, this uncertainty is passed along to delegators, as the validators may incur more significant costs or may take additional measures to ensure their status as a valid participant in staking rewards.

• Vested interests: one of the core underlying assumptions of any PoS - the reliance on vested interests derived from locked assets - may lose its salience as lending and borrowing markets become more liquid and interest rates become too low to be a barrier to borrow or obtain more coins. If higher staking participation rates do not lead to greater degrees of decentralization and network participation, the assumption that staking is necessary to incentive positive network contributions may also break.

3. The maturation of the staking infrastructure

In the preceding sections, several tradeoffs were discussed with various degrees of complexity. Fortunately, users do not have to face alone all these tradeoffs as there are many staking providers that provide options and tradeoffs for users, depending on their individual needs. In this section, market players are described and categorized based on the various types of service(s) provided to existing and prospective users.

3.1 Staking pools

Staking pools are typically (on-chain) addresses or individual delegated candidates that accept votes or on-chain pledge support, without requiring relinquishing custody of one’s coins. Often, staking pools are coin-specific, much like mining pools, and are designed specifically for helping reduce the barriers to entry to participating in earning some piece of the consensus rewards. Hence, it allows small players to take advantage of any economies of scale that may exist on any chain. One of the best examples is Lumenaut Pool for Stellar, a free pool that allows users to set them as an “inflation address”, which then subsequently passes the rewards back to the delegator.

For many Ethereum-based or EVM-compatible staking mechanisms, smart contracts may be deployed to pool together funds and earn optimal rewards. Examples include Tomochain’s Tomo Pool or ThunderCore’ staking pool service.

3.2 Delegation services

For some chains, further work is necessary beyond just pledging funds on-chain. In these cases, service providers may operate a node and handle most (if not all) of the obligations that may exist for a given chain. Delegation services, such as Staked.us, Infinity Stones, Certus One, and Chorus One, participate in many chains. These services are particularly useful for delegated Proof-of-Stake systems, for which validators must attain a critical mass of coins to participate in the staking mechanism.

These delegation service providers also roll out other utilities and tools for blockchain users to better understand the networks they are interacting with. For example, Staked rolled out its Robo-Advisor for Yield (RAY), Infinity Stones offers reports on the coins it supports, Figment.networks offers block explorers for Cosmos & other Tendermint-based chains through its Hubble domain, and Chorus One offers biweekly roundups on all things staking-related.

In exchange for this access, education, and to cover the real fiat costs of operating nodes, many “Staking-as-a-Service” providers charge a fixed rate fee on the staking rewards generated from any amounts pledged by users. Thus, staking providers operate as a “revenue-sharing” business, taking a cut of the total staking rewards claimed by their users.

3.3 Dedicated custodian solutions

These delegation services also offer significant overlaps with custody solutions, as the fund holder can help staking on behalf of the user and also take a cut, similar to delegation services.

3.3.1 Custody’s first step

For example, with Coinbase Custody, users will be able to collect rewards for select custody coins in a revenue-split model, with Coinbase taking a share of the rewards. From the perspective of coin holders, the revenue-split model allows them to delegate custody while still benefiting from margin-positive returns (vs. not staking).

Holding one’s own funds may be both a blessing and a curse - for some users, retaining custody is of utmost concern (“not your keys, not your coins”). However, for others, maintaining private keys, wallets, and other necessary infrastructure may be a burden. In this case, an attractive alternative could include relying upon a custodian or other staking providers who provide some Service Level Agreements (SLAs). For instance, Staked.us offers an SLA on uptime for its nodes, ensuring that users can be reassured of any technical difficulties that may come up during staking operations that prevent the user from receiving proportionate rewards. Uniquely, staking providers are held to a level of accountability through the transparency of the blockchain, as staking providers competing for users’ delegations are forced to display their rates publicly on-chain for users to select right alongside their uptimes and past performances, thus providing clear transparency into their business model and reliability.

3.3.2 Ecosystem tools as staking providers

Platforms that are already actively engaged in the ecosystem and offer trust-based services are ideally positioned to further monetize their existing user bases by funneling users to their staking validator. Because these trusted network participants must have already established a reputation for providing something of use to the community, they are naturally well-positioned to steer users to delegate assets to a trusted validator further.

The ability to cross-sell existing customers is facilitated by the fact that users may already have accounts/balances with these platforms and can thus avoid going through burdensome onboarding processes. For example, TrustWallet has rolled out its staking platform, providing its existing token holders an opportunity to participate in network staking rewards directly through their interface.

3.3.3 Exchanges and liquidity providers

Exchanges are the newest players in the staking game, with Binance, Poloniex, Huobi, OKEX, and Coinbase offering staking rewards to their users.

However, in the case of Huobi, OKEX, and Coinbase, the products are separate from their exchange wallets themselves, and thus, users cannot both trade and stake at the same time in the same product.

Staking rewards on exchanges for user balances held directly on trading accounts is unique because it provides a dynamic similar to maker-taker fee spreads. While users set limit orders, the coin they are providing to the trade still remains in their wallet and accrues staking rewards. As a result, limit sell orders of staking-supported coins are incentivized, rather than market sell orders6. Uniquely, users who want to partake in staking rewards may also be incentivized to set market buy orders, rather than limit buy orders, to guarantee the order completion and the beginning of staking rewards accrual.

Project teams have new opportunities as exchanges partake in the process of staking, as these exchanges can help act as a last-mile conveyor of education regarding their technology and chain to the crypto-audience masses. Particularly on exchanges, the front lines of price war battles between bulls and bears, incentivizing longer-term holdings through staking rewards mechanisms, may prevent some “pile-on” investing mentality in which the momentum of trading may drive prices significantly down during a short period.

Though the term “soft staking” has been thrown around, the effect of exchanges adopting staking may give rise to a more robust staking ecosystem for all projects. Uniquely, exchange providers can offer more access and liquidity to token holders while simultaneously increasing the network staking participation rate. From the perspective of exchange platforms, the support for staking is a very compelling decision, as it allows exchanges to align interests with the longevity of the products they offer on their marketplaces.

4. Conclusion

Concluding, Binance Research proposes a working taxonomy of staking projects. This taxonomy is derived from Section 2 above and can be expanded on in the future. It gives a simple overview of the core differentiating elements of staking projects, including:

1. Protocol: the underlying protocol sets the fundamental framework, as it defines how staked tokens ought to be used. Relevant variations are PoS, DPoS, BFT, pBFT.

2. Token model: the token model (i.e., singular token vs multi-token) defines the use cases thereof.

3. The concentration of staked coins: is a relevant metric that is presumably dependent on multiple factors such as the ease of staking or the fact whether the supply has a cap.

4. Rights: staking tokens can grant governance rights and open passive revenue streams (i.e., block rewards, transaction fees) as described in Chapter 2.1.

5. Obligations: the unique set of obligations (i.e., lock-up periods, operation of a full node, technical expertise) is another differentiating factor that is described in Chapter 2.2.

6. Risks: lastly, risks (i.e., opportunity-costs in a highly volatile ecosystem, counterparty risk from custodial staking, etc.) such as the ones described in Chapter 2.3.

Besides this general overview, staking can be assessed from several different perspectives.

For users, staking provides a new assortment of investment opportunities and strategies to generate passive income along with their underlying bets on a blockchain. To serve these users, many staking providers will create business lines bridging the needs of users to participate in blockchain rewards. As the evident growth of the staking ecosystem continues, network effects are likely to draw in more and more players into the staking game.

For projects, considering the staking dynamics during mechanism design and governance may allow them to provide a more compelling offering to compete for staking-seeking users. Staking can also be described as a lending system where the chain is the lender of last resort, substituting technical risk (of contract failure or exploit) for counterparty risk. As such, projects setting staking rewards will also have to consider the interplay between their rates and the rates that the open market may command in lending markets.

The concept of being able to set a “native interest rate” on-chain promises the ability to provide significant rewards for users without significant physical or technical capital to participate in consensus. As the infrastructure matures for receiving staking-related rewards and participating in DeFi, blockchain users will have more options at their fingertips to choose the products that best fit their needs.

With Ethereum slated to switch to Proof-of-Stake in the not-too-distant future, the blockchain space may be forced to reckon with staking sooner, rather than later.

References

• Algorand. Staking Rewards (Accessed on October 25th 2019). https://algorand.foundation/stakingrewards/

• Binance. Staking Portal (Accessed on October 25th 2019). https://www.binance.com/en/staking

• Butterin, Vitalik. A Next-Generation Smart Contract and Decentralized Application Platform (2015). https://github.com/ethereum/wiki/wiki/White-Paper

• DeFi Pulse. Website (Accessed on October 25th 2019). https://defipulse.com/

• Duffield, Diaz. Dash: A Payments-Focused Cryptocurrency (2016). https://github.com/dashpay/docs/raw/master/binary/Dash%20Whitepaper%20-%20V2.pdf

• EOSAsia. What is CPU? What is RAM? And How Does the EOS blockchain Utilize These Resources? (2018). https://medium.com/@eosasia/what-is-cpu-what-is-ram-and-how-does-the-eos-blockchain-utilize-these-resources-a7a52e158652

• EthHub. Eth 2.0 Economics (Accessed on October 25th 2019). https://docs.ethhub.io/ethereum-roadmap/ethereum-2.0/eth-2.0-economics/

• Figment Network. (Mis)Understanding Yield and Inflation in Proof of Stake Networks https://figment.network/resources/misunderstanding-yield-and-inflation-in-proof-of-stake-networks

• Krohn, Steven. Everything You Need To Know About Staking Coins (2018). https://medium.com/@stevekrohn/everything-you-need-to-know-about-staking-coins-686ea95041c

• Lumenaut Pool. Website (Accessed on October 25th 2019). https://pool.lumenaut.net/ (Lumenaut Pool is a Stellar inflation pool with no fee)

• Mazieres, David. The Stellar Consensus Protocol: A Federated Model for Internet-level Consensus (2016). https://www.stellar.org/papers/stellar-consensus-protocol.pdf

• Staking Rewards. Website (Accessed on October 25th 2019). https://stakingrewards.com/

• Staked.us. Yields. Key Staking & Financial Metrics(Accessed on October 25th 2019). https://staked.us/yields/

1. See this article about resources in EOS. https://medium.com/@eosasia/what-is-cpu-what-is-ram-and-how-does-the-eos-blockchain-utilize-these-resources-a7a52e158652↩

2. Stellar, NEO, VeChain, Kucoin Shares, Komodo, and Nexo do not require any lock-up.↩

3. As costs are typically denominated in fiat terms from the perspective of users.↩

4. In traditional financial markets, this also does not hold in the short-term.↩

5. For instance, Augur. Augur was discussed in one of our past reports about design flaw attacks. https://research.binance.com/analysis/augur-design-flaws↩

6. Future analysis might be conducted to analyze the effects of staking from a trading perspective.↩