Liquid staking services are now the most important infrastructure for the blockchain industry beyond DeFi.
However, liquid staking has not found a place in the Cosmos ecosystem because the multi-chain ecosystem is inevitably fragmented.
Stride, which will provide interchain staking services to the Cosmos ecosystem, aims to solve this fragmentation problem with new interchain features.
It's been almost four years since the idea of liquid staking derivative and using them for DeFi and other services was born. At the time, liquid staking services were considered one of the many DeFi services, but as time has passed and more PoS blockchains have launched their own liquid staking derivatives, liquid staking derivatives have become not just a DeFi service, but a core infrastructure service required for PoS blockchains to thrive. As if to prove it, the liquid staking sector has become the largest sector in the blockchain industry as of this writing (as of May 17, 2023, the liquid staking sector has a TVL of $17.39B, about $0.5B ahead of the decentralized exchange sector with a TVL of $16.87B).
(TVL ranking by sector as of May 17, 2023 | source: DefiLlama)
In fact, this result is not surprising, because if you are going to stake tokens on a PoS chain, it is basically much better to use a liquid staking derivatives to receive staking rewards and use them for other services rather than just staking directly on the network and locking up your assets. In addition, in the case of Ethereum, where you need to hold 32 ETHs to stake, you can stake less than 32 ETHs through the liquid staking derivatives like Lido, and there is another advantage that you can get staking rewards without running your own node. With these advantages in mind, it is quite obvious why a lot of capital is flocking to the liquid staking sector.
Of course, with a lot of capital being flowed into this sector, there are also a lot of competitors in the space. At this point, it's safe to say that the Ethereum liquid staking market is already saturated. Starting with Lido Finance, a leading provider of Ethereum liquid staking derivatives, there are now several other players in the sector, including Coinbase's cbETH, Rocket Pool's rETH, and stablecoin project Frax Finance. The Ethereum liquid staking space is literally a red ocean. So, my question is are there any other ecosystems left with opportunities besides Ethereum?
In my opinion, the Cosmos ecosystem is the ecosystem where liquid staking derivatives can be highly successful, because it is basically the ecosystem of app-specific chains (aka app-chains), so even though it looks like one ecosystem in the big picture, it has a structure of various layer 1 governance tokens within the ecosystem. The combined market capitalization of the entire Cosmos ecosystem is around $60B, which is second only to the Ethereum ecosystem in terms of market capitalization (please note that this calculation includes the outlier called BNB Chain). Despite this, the current Cosmos liquid staking sector is not that large, why?
Before we dive into the Cosmos liquid staking market, let's take a look at how liquid staking actually works. First of all, the liquid staking service itself is composed of very simple processes: When a user deposits the tokens they want to liquid stake, they receive shadow tokens which have equavalent(almost equavalent let’s say) value to original tokens, and the tokens they deposit are distributed to validators based on the criteria set by the liquid staking services. However, this is only from the user's point of view, and from the point of view of building a liquid stakting service, it can be said that it has its own complex structure because all functions such as receiving, distributing, and delegating users' tokens to validators, setting fees for staking rewards, and issuing shadow tokens must be implemented through smart contracts (there might be security problems such as hacking because it is implemented with smart contracts).
It's a bit complicated, but it's not just liquid staking services that use smart contracts anyway. In fact, almost every application implemented on a blockchain uses smart contracts, so it's not a big deal. However, in the Cosmos ecosystem, where appchains exist independently, a new problem arises. Of course, if each of the many appchains in the Cosmos ecosystem implemented its own liquid staking protocol on their chain, there would be no structural differences compared to other blockchain ecosystems. However, since the Cosmos ecosystem has a unique structure of one chain for one application, if each of them implemented its own liquid staking derivative, they would have to add other applications for liquid staking derivatives, which would eventually dilute the advantages of being an application specific chain. In order for the liquid staking service to be meaningful, users must be able to utilize the tokens (shadow tokens) they receive in exchange for depositing original tokens into the smart contract, and if each Cosmos app-chain has different tokens, there will be great inefficiencies in terms of compatibility and utilization of the tokens themselves.
What if there was an app-chain dedicated to liquid staking only? In fact, such a structure would be the best fit for the characteristics of the Cosmos ecosystem. However, there's still a problem. Since the Cosmos ecosystem is the ecosystem where independent chains have independent sovereignty and communicate using a communication protocol called IBC(Inter-blockchain communication protocol), even if an app-chain exclusively for liquid staking were to emerge, that chain would need to be able to communicate with other chains in the Cosmos ecosystem in an organic way (the liquid staking chain would need to be able to get control over the staking and governance of other chains, as well as get the state values of other chains as reliably as possible) to be able to liquid stake various Layer 1 tokens in the Cosmos ecosystem. Up until now, this has not been technically possible, which is why the Cosmos ecosystem hasn't seen any significant liquid staking services. In a way, it can be said that the characteristics of the interchain ecosystem pursued by Cosmos have prevented liquid staking from being activated. In other words, in order for the concept of ideal liquid staking to be established, the communication protocol of the Cosmos ecosystem, IBC, needs to be more advanced.
Of course, the developers of the Cosmos ecosystem were clearly aware of these limitations, so they tried to make the communication between appchains more advanced, and the technologies that emerged from these efforts are Interchain Account (ICA), Interchain Queries (ICQ), and Replicated Security (RS).
ICA is a feature that allows accounts on one chain (e.g., Stride) to communicate with and control accounts on another chain (Cosmos Hub). As long as the chains that want to communicate have upgraded their ICA, they can create accounts on each other's chains.
To understand ICA, you need to understand two concepts: “Controller Chain” and “Host Chain”. A controller chain is a chain that "controls" an account on another chain, and a host chain is a chain with a controlled account. So in my example (Stride<>Cosmos Hub), Stride would be the controller chain and Cosmos Hub would be the host chain.
Here's how ICA works:
The controller chain (Stride) sends IBC packets to the host chain (Cosmos Hub).
The relayer receives the packet and sends it to the IBC module in the host chain (Cosmos Hub).
The ICA on the host chain receives information about what to do with the IBC packet (it can be about staking, governance, or opening a CDP, etc.).
The ICA on the host chain performs its duties by issuing transactions as passed to it by the relayers.
Before ICA was enabled, different Interchain Standards were required for IBC transactions coming from different chains, but after ICA is enabled, the IBC module does not process incoming transactions, but only forwards them to the interchain account, and the processing of incoming transactions is performed by the chain's own logic (the interchain account performs the transaction as received).
With the addition of features like ICA, “an appchain dedicated to liquid staking only” can be created, because the liquid staking appchain will have access to other independent appchains through ICA. Using Stride as an example, the process of staking ATOM on Stride might look like this:
A user deposits ATOM to Stride (and receives a proof-of-stake token, stATOM, in return).
Stride forwards the ATOM via IBC to the Delegation ICA (ICA) in the Cosmos Hub.
The ICA at the hub verifies the message and stakes it by firing a transaction that stakes the ATOM (which also determines which validator it delegates to).
In addition to the staking liquidation just described, ICAs are also used to reinvest rewards accumulated in tokens that are already staked: In fact, every 6 hours(every epoch), Stride processes the rewards accumulated in staked tokens, sending 90% of the rewards to the Delegated ICA to be re-staked, and 10% to the Fee ICA to pay fees. In this way, Stride organically reinvests staking rewards by directly controlling the events that happen on Cosmos Hub.
While these technologies may seem complicated, they don't actually require users to do much, but rather allow users to stake/reinvest their tokens on the Cosmos Hub without having to access the Cosmos Hub directly through the ICA.
If ICA is about giving a controller chain access to exercise certain actions on a host chain, then Interchain queries (ICQ) is an interchain solution that allows you to retrieve data from other blockchains without relying on a third party. A blockchain is essentially a database that is independent of the environment outside of the blockchain. This means that it's hard to trust data outside of the blockchain unless it's produced by the blockchain itself. However, in order for a blockchain to become a more useful database, it must eventually be able to accept data from environments outside of its own chain (including other blockchains), as well as validate that data. Third-party services called oracle solutions (Chainlink, Pyth Network, Band Protocol, etc.) have traditionally done what blockchains can't do on their own. But in the end, relying on third-party services to bring data from outside the chain is risky, and hackers have been known to attack oracle solutions and steal funds, so oracle solutions have not been good alternatives so far.
So here we have ICQ. ICQ is a technology that can be implemented because the Cosmos ecosystem has already been using a protocol called IBC to communicate between chains based on the Cosmos SDK. The way the Cosmos ecosystem chains exchange data through ICQ is as follows:
In the figure above, the chain sending the data query is called the "home chain" and the chain receiving the query is called the "satellite chain." The node in the home chain configures the query values in the IBC packet and forwards the packet to the satellite chain through a relayer.
The satellite chain then receives the packets through the IBC module and forwards them to the appropriate modules.
The ICQ module checks the query values and forwards them to the application modules in the satellite chain where they are needed.
After using the received query value, the result is sent to the home chain through the ICQ module and IBC module.
(Please note that since ICQ is an open-source protocol, there are many variations and many different ways to get data, so consider this example to be just one of many ways to get data using ICQ).
After all, since Stride is the liquid staking chain of the Interchain ecosystem, we need to know how much staking rewards are being received by the assets staked through Stride. This is where ICQ can be used to read staking rewards. However, Stride doesn't rely on ICQ as much as other chains (especially QuickSilver, which it competes with).
Another feature that is arguably as important component of Stride as ICA and ICQ is Replicated Security (RS or ICS), formerly known as Interchain Security. Along with Neutron, Stride will be one of the first chains to adopt RS. So, what is RS and how will it change the Cosmos ecosystem?
3.3.1. A PoS ecosystem where token value is security
First of all, all chains in the Cosmos ecosystem, including the Cosmos Hub, are inherently using PoS consensus. This means that the security of the chain itself is directly related to the value of the token. For example, if the native/governance token issued by an app chain has a market capitalization of $100, it would only take a little over $50 to buy the chain's majority tokens and attack the chain. Therefore, there are structural limitations that make newly launched PoS chains less secure than chains with high market capitalization, and RS is a concept that came up to resolve this problem.
3.3.2 Getting to know RS
To understand RS, you need to understand two concepts first: Provider-Chain and Consumer-Chain. A Provider Chain is literally a chain that provides security to a Consumer Chain, and a Consumer Chain rewards the Provider Chain with the Consumer Chain's native token in exchange for the security that provider chain provides. While many people understand RS as "Cosmos Hub sharing security with other App-chains," any App-chain can become a Provider Chain in the future. However, since Cosmos Hub has the highest value in the Cosmos ecosystem to date, it is likely that Cosmos Hub will be the leading provider chain and therefore the only provider chain for the time being (other Cosmos ecosystem appchains are already adopting different security sharing methods, so it is already clear that Cosmos Hub is not the only one sharing security).
Once you understand the provider and consumer chains, let's dive into how RS works. First, the provider chain passes its own set of validators to the consumer chain via the inter-block communication (IBC) protocol. The consumer chain then applies the provider chain's validator set to the consumer chain's validator set. In this way, the provider chain's validator set and the consumer chain's validator set can be configured identically. These validators will produce blocks and validate transactions for the consumer chain. The problem is that just because the same validators produce the consumer chain's blocks and validate transactions, it doesn't mean that the consumer chain has the same level of security as the provider chain. In the end, it's the connectivity of the consumer chain to the provider chain that matters, which is why when performing RS, the provider chain's validators also act as validators for the consumer chain, using the provider chain's native token (e.g. ATOM). Using the native token of the provider chain for RS means that if validators fail to validate and produce blocks for the consumer chain, they lose the token of the provider chain, not the token of the consumer chain, and because of this structure, they share the security level of the provider chain.
Now that Stride is also adopting RS, Stride will have the same set of validators as Cosmos Hub, which will allow it to enjoy the same level of security as Cosmos.
Now that you know how Stride works with ICA, ICQ, and RS. Then what makes Stride different as a Liquid Staking Derivative? And how does it compare to other competitors like Lido, pStake, and Quicksilver?
First of all, Stride aims to provide liquid staking services across multi-chain (or inter-chain) ecosystems, something that hasn't been possible before due to technical limitations. Until now, the fact that the Cosmos ecosystem was composed of independent layer 1 chains made it difficult to organically liquid stake (since independent chains could not communicate with each other to send and receive contracts), but with the advent of technologies such as ICA and ICQ, the multichain structure of the Cosmos ecosystem has made it a more viable ecosystem for liquid staking. This is because in a traditional single-chain ecosystem, the only asset that could be used to liquid stake was the chain's governance token (in the case of Ethereum, only Ethereum, and in the case of Solana, only Solana), whereas in the Cosmos ecosystem, with an infinite number of layer-1 chains running in parallel, the number of tokens that could be liquid staked could be infinite.
Of course, the fact that more tokens can be liquid staked doesn’t mean it is always better, but more tokens available to be liquid staked could mean more opportunities for service providers (more options for tokens to succeed, and risk hedging for Stride because Stride can provide liquid staking services to other Layer 1s even if a particular Cosmos-based Layer 1 token doesn't succeed). Also, while there aren't many high-market-cap chains in the Cosmos ecosystem right now, it would be very meaningful to provide liquid staking services to the multi-chain ecosystem of Cosmos, assuming it continues to grow in the future. This is why you should look at STRIDE differently than other liquid staking services like marinade and others.
The fact that Stride uses the RS mentioned above is of course a huge differentiator. The use of RS means that Stride has the same level of security as the Cosmos Hub from the start. Especially for a blockchain like Stride that provides liquid staking for the entire Cosmos ecosystem, security is very important as it deals with a lot of assets, so being able to secure them from the start with RS is very important.
Currently, Cosmos has a market capitalization of about $3 billion. This means that in a PoS ecosystem where the value of a token is directly related to its security, STRIDE will enjoy the same level of security as a blockchain with a market capitalization of around $3 billion. One could argue that this is a reason for users to trust STRIDE. However, if ATOM, the token of Cosmos Hub, is used to secure Stride, what is the utility of the Stride token itself?
4.3.1 Tokenomics before RS
First of all, unlike Neutron, which has been using RS from the beginning, Stride chose to have its own validator set at first and then incorporate it into RS. Therefore, there will be some changes in the tokenomics before and after the implementation of RS. First of all, let's take a look at the current tokenomics before RS.
(Stride token distribution table : Source: Stride)
The total supply of Stride is 100 million, with roughly half of the total supply being distributed in the middle of year 2, and 95% of total supply will be in circulation by the end of year 3. Stride tokens will be released to the market in three different ways: 1) block rewards (18,900,000 ST tokens in total), 2) Stride incentives (31,000,000 ST tokens in total), and 3) rewards for core contributors and partners (40,900,000 ST tokens in total). It is very common for blockchains to distribute tokens as block rewards and tokens to early contributors and investors, but another token distribution channel called Stride Incentives is a structure that is rarely found in other blockchains. So what are Stride Incentives and how are they distributed?
The Stride Incentive Pool will be operated by the Stride DAO, composed ST token holders. The holders will be the ones who decide how the incentive pool is distributed. However, to make the operation of the incentive pool more predictable, there is an emissions quota that the DAO must spend. According to Stride, a total of 20,000,000 ST tokens will be available in year one and 9,000,000 ST tokens in year two and beyond. The Stride Incentive Pool can be used to provide liquidity for shadow tokens issued by Stride, as well as to fund the order book.
In addition, Stride token holders will be able to decide which validators to delegate original tokens deposited in Stride, adjust Stride's incentives, and change Stride's structure through governance. Ultimately, the value of the Stride token will be proportional to the value Stride provides, but it will also depend on how large the Cosmos ecosystem becomes as Stride provides liquid staking services to the Cosmos ecosystem.
4.3.2 Tokenomics after RS Adoption
As I explained above, adopting RS is a huge win for Stride. However, for RS to be sustainable, Cosmos Hub must also be paid for their security. Just as there's no free lunch, there's no free security. So, in exchange for the benefits it receives from RS, Stride will share 15% of the revenue from liquid staking(as explained in the ICA section, Stride reinvests the rewards of assets staked on other chains every 6 hours, with 90% of the rewards going to the delegation ICA to be re-staked, and 10% going to the Revenue ICA as a fee for providing liquid staking. Think of this 15% as 15% of the 10% of the total rewards sent to the Revenue ICA), 15% of the Stride token inflation, 15% of the MEV revenue, and 15% of the transaction fees with Cosmos Hub. This means that the total rewards after RS will be lower than they were before RS, at least in the short term, but it would be very strategic to dedicate 15% of the revenue if Stride is able to provide liquid staking to more app-chains because of RS.
(Table comparing liquid staking derivatives in the Cosmos ecosystem | Source: Stride)
To be honest, each liquid staking service does not have a unique technological methodology (just looking at the table above right now, you can see that they all use the same ICA in the Liquid Staking Technology section). Since they all have very similar liquid staking mechanisms, it's hard to tell the difference in their technologies, and it's more relevant to compare them in other areas. Let's take a look at the non-technical differences one by one.
4.4.1. Economic security aspects of the chain
First of all, in terms of the economic security of the chain, Stride will have a similar level of security to Lido Finance. First of all, since Stride is a chain that directly uses RS (labeled ICS in the table above) as I mentioned above, it will have the security level of Cosmos, and since Lido Finance is implemented on the Neutron chain(this means that Lido in Cosmos is not itself a separate layer 1 chain but deployed on an App-chain called Neutron), it already shares the security of the Cosmos, so we can say that Stride and Lido have a similar level of security.
On the other hand, pStake and Quicksilver have not yet announced which security model they will adopt. If they do not use RS and provide services on their own validator set, they will probably have a lower level of security than Stride and Neutron in terms of economic security.
4.4.2. Code Security
When we talk about code security, we're referring to the degree to which a liquid staking protocol has a process for changing the code. In this regard, Stride, pStake, and Quicksilver have the same level of code security. This is because they're all proprietary blockchains, and changes to their code must go through a process called governance.
On the other hand, Lido is not a standalone chain, but rather a smart contract on an app-chain called Neutron, so it's likely that they are controlled by a multisig wallet, which means that from a code security perspective, any chain other than Lido is more secure.
4.4.3 Audits
While it's hard to compare smart contract audits based on the number of times they've been done, we can see that Stride and Quicksilver have been audited three times, indicating that they've taken a lot of care in securing their smart contracts. However, in the case of Quicksilver, it was exploited for $30,000 in December of 2022, and even chain itself briefly halted after exploit. Of course, it's hard to say that this incident was Quicksilver's fault (Quicksilver said it was caused by the Cosmos Hub not upgrading its security in time (it was running IBC-GO v3.0.0 when it should have been running IBC-GO v3.3.0)). Still, it seems clear that Stride is a secure protocol that hasn't had any security issues so far.
4.4.4 Rate Limiting
Rate limiting is a security feature that allows the blockchain to continuously monitor the number of tokens and automatically halt blockchain when it falls below or exceeds the protocol's set limit. For example, If the total amount of OSMO in circulation on Osmosis is 100 million, and Osmosis imposes a 25% rate limit on OSMO withdrawals and a 50% rate limit on deposits (you can also set the time period for which this limit applies, for example, 25% and 50% for one day or 25% and 50% for seven days), then only 25 million OSMO can be withdrawn from Osmosis and only 50 million OSMO can be deposited. The reason for implementing this rate limiting features is to minimize the impact of bad things happening in other ecosystems due to the interconnected nature of the Cosmos ecosystem. If we look at what happened to Osmosis after the collapse of Terra, we can see that rate limiting is a necessary security feature in the interchain ecosystem. Stride uses rate limiting in two ways: 1) to ensure that when shadow tokens are created, they can only be issued for the value of the tokens they are backed by, and 2) to limit the number of simple IBC deposits and withdrawals. Currently, Stride seems to be the only liquid staking service that applies rate limiting at the protocol level.
4.4.5 Chain minimalism
Chain minimalism means having the least amount of codes on the blockchain. Whether or not chain minimalism is necessarily a good thing is a matter of personal preference, so we'll leave that aside and just compare the degree to which a protocol is chain minimalist, and Stride is the most chain minimalist chain. This is because Stride has no plans to implement any applications on top of the Stride chain other than liquid staking feature. However, pStake and Quicksilver have stated that they will implement multiple applications on their chains, and Neutron is a multi-purpose smart contract platform from the ground up, so it's fair to say that Stride is the most chain minimal of the four.
4.4.6. Host-chain validator selection
How do you select a validator from the client chains that Stride provide liquid staking to? After all, liquid staking chains provide the service of staking users' tokens through ICA and selecting validators for them, so each chain has criteria for how to delegate tokens. In the case of Stride, users choose which validators to delegate their tokens to through governance; in the case of pStake, users set certain criteria and delegate to the validators that best meet those criteria; and in the case of Quicksilver, users can choose their own validators when issuing shadow tokens. In the end, this is where the comparison between Stride and Quicksilver comes in, and it's hard to say which methodology is better because there are clear advantages and disadvantages to having on-chain governance determine delegation versus letting users do it themselves. It's a matter of taste.
4.4.7 Participation in governance of customer chains (liquid staked token voting power)
Will holders of shadow tokens be given the right to participate in the governance of their chains? Both Stride and Quicksilver said "yes" to this question. I think it's a good thing that shadow token holders will be able to participate in governance, as it will increase the usability of the shadow token itself.
This is how Stride compares to other competing protocols. While all of them have their pros and cons, there are still many areas where Stride has a slight edge. However, a lot of this will depend on personal preference and taste. I hope you'll take a look and decide for yourself.
Currently, Stride offers liquid staking for Cosmos Hub (ATOM), Osmosis (OSMO), Juno (JUNO), Stargaze (STARS), EVMOS, Terra2 (LUNA), and Injective (INJ), and will soon offer liquid staking for dozens of assets in the Cosmos ecosystem, including Secret Network (SCRT) and Accelerator (AXL).
As a chain that strives for chain minimalism, it is interesting to see that Stride's ecosystem is not composed of all different applications built on chain, but chains that Stride provides liquid staking services instead.
After the development of technologies such as ICA, ICQ, and RS, many market players wanted to provide liquid staking services to the Cosmos ecosystem. This is because most of the technical difficulties have been solved. While Stride is now the market leader, offering liquid staking services for the largest number of assets, we don't know yet. We know that Stride’s competitors(pStake and Quicksilver) are actively building their products, and Lido, which has strong dominance on Ethereum and other ecosystems, prepares to launch on the Neutron blockchain. In the case of Lido in particular, it's the player with the best understanding of the nature of liquid stkaing services, having beaten out numerous competitors in the Ethereum staking market. Ultimately, for Stride to survive the liquid staking derivative war in the Cosmos ecosystem, it will need to not only aggressively increase the number of assets it supports, but also increase the usability of its shadow tokens so that they can be widely used in the Cosmos ecosystem. The key here will be whether or not Stride can leverage other ecosystems to increase the utility of its shadow tokens while sticking to chain minimalism. In the case of Lido, there are a number of applications that will be on Neutron (a multi-purpose blockchain), and Stride will need to have a specific and well thought out plan on how to increase the usability of the shadow token without creating its own applications. One thing is for sure, once Stride dominates the Cosmos ecosystem, its potential value will increase tremendously. We're excited to see where Stride goes from here.
Thanks to Kate for designing the graphics for this article.
We produce in-depth blockchain research articles
Let’s learn about Sei DB, one of the most important features of Sei v2.
This covers the Injective Network, which has undergone consistent development and change in the market over the past four years.
Solana is the leading integrated blockchain featuring parallel processing, low fees, and fast transactions. With their practical and developer-friendly technology stack, Solana is rapidly building a unique ecosystem for real-world adoption, overcoming a series of crises and regaining market share.
Explore the origins of the Sui blockchain and its technologies in detail, and investigate how Sui differs from other blockchains