My aim here is to provide a comprehensive introduction to Solana liquid staking. I just realized in doing research for this Phantom has a somewhat similarly structured article (https://phantom.app/learn/crypto-101/solana-liquid-staking), and I’m going to cover some of the same territory, but with a lot more opinions and editorializing. I want you to know not just what liquid staking is and how to do it, but why. What would possess a man to spring out of bed in the morning and say to himself, “I’m going to liquid stake my SOL today”? Come, join me on my magic carpet, and I will show you a whole new world.
I will try to make it accessible to newcomers without completely boring the reader to death with over-explanation (you can judge whether I’ve succeeded). I’ve also tried to structure this rather long article so you can skip certain sections if you’re already familiar with the topic. With that in mind, here is the table of contents:
I’ll also give the obligatory disclaimer that none of this should be construed to be financial, legal, or in Haseeb Qureshi’s wonderful addendum, life advice. This article is meant to educate, and hopefully, delight and entertain.
Before we jump right into liquid staking, let’s talk about regular staking. Staking in a delegated proof-of-stake network like Solana is the act of committing your tokens to a validator who must promise to faithfully verify transactions on the network or risk penalties. This is what creates fundamental alignment between the validators and the users of the network, without which, double spending, censoring, and all other manner of abuse would be possible. When you stake “natively” you choose a specific validator and delegate your tokens to them. You can perform this operation through an array of wallet software or with the Solana command line interface.
Since the launch of Solana mainnet-beta in February 2020, Solana has been following this proposed inflation schedule:
As I’m writing this in February 2024, we’re four years from that launch date, so it’s relatively easy to see the current inflation is about 5%. The precise inflation amount is governed by three parameters: initial inflation rate (8%), disinflation rate (-15%), and the long-term inflation rate (1.5%). Inflation started at 8% and decreased 15% on an annualized basis at every epoch boundary until it comes to rest at 1.5%. This could be changed in the future, but this is the schedule that has been observed since launch.
Who gets the SOL tokens that are created through inflation? It’s pretty straightforward: stakers. This means that every epoch, stakers are increasing their relative ownership of the total amount of SOL tokens at the expense of non-stakers. Nothing more complicated is really happening. If all SOL were staked, no one would be increasing the total value of their holdings. This leads to a high stake rate on the Solana network; about 2/3 of all SOL is staked at time of writing. The rate of liquid staking, however, is still under
With a proof-of-work blockchain, validators have a high burden of equipment and energy costs that force them to sell some (or maybe all) tokens they earn merely to break even. With a proof-of-stake network like Ethereum, these costs are very, very low, so there isn’t much sell pressure at all. With Solana, operating expenses for validators are slightly higher than Ethereum because validators must execute transactions as a part of consensus which imposes a cost and validator equipment is slightly more expensive because it needs higher performance than Ethereum. So validator sell pressure is still very low compared to Bitcoin, but a bit higher than Ethereum.
I’ve taken some pains here to detail this because people keep getting confused about this. I occasionally rant about this on there interwebz. Ok, I admitted above there is a tiny bit of sell pressure because running a validator isn’t cost free, but the crypto Twitter doesn’t much appreciate nuance, so I went a little more aggressive. Anyway, with this essential background on regular staking in hand, let’s talk about liquid staking.
Since there is a high incentive to stake on Solana so as to not be diluted by the other stakers, one of the few reasons not to is that it locks your capital up each epoch. With liquid staking you contribute your tokens to a stake pool that manages the distribution of the stake across a set of validators, and tokenizing the fact that you have committed your tokens to the pool. This action returns a new asset to the staker that represents that fact and allows the user to redeem it for the original staked SOL. Thus, it can be used as a functional equivalent to SOL in many contexts.
Most popular liquid staking tokens (LSTs for short) on Solana are “reward-bearing” tokens. Almost all of the SOL in the pool is staked to validators chosen by the pool operators (sometimes minus a small buffer which facilitates quick redemptions), so this staked SOL in the pool accrues rewards in the form of more SOL. Thus, the LST doesn’t change in quantity over time, but comes to represent more SOL each epoch and so it appreciates in price relative to SOL. Another approach is “rebasing” whereby holders of the LST are awarded more of the liquid token and each liquid token is redeemable 1:1 (with a delay), but this is not much used on Solana.
Each epoch takes about 2.5 days on Solana. If a user wants to withdraw vanilla staked SOL, they must submit a request to withdraw their delegated stake wait until the end of the epoch and then it becomes redeemable. In traditional financial markets this is framed as taking on duration risk. You are making a bet that the reward of locking up the capital for 2.5 days will outweigh the risk that you will need it immediately. 2.5 days is much less of a risk, duration-wise, than a 10-year US Treasury note. On the other hand, US Treasuries are… a bit less volatile than crypto, generally speaking.
So, when a holder of a liquid staked token wants to access the underlying SOL token, they can redeem it from the pool of staked tokens controlled by the liquid staking protocol, opting either to wait until the end of the epoch when it can be unstaked by the underlying pool, or by using existing liquidity pools to trade the mSOL for SOL on the open market. Here is an example from Marinade Finance:
Marinade is telling you it’s using the Jupiter decentralized exchange aggregator to trade the mSOL for SOL. I checked unstaking 10,000 mSOL and the price impact was 0.01%, but at 100,000 mSOL, the price impact shown above is 100,000 (I don’t actually have 100,000 or even 10,000 mSOL, but it lets you simulate it). What does this mean?
It means you’re getting 8.162% less SOL than you would if you waited until the end of the epoch and unstaked directly. The price impact above is based on current market conditions, i.e. how much liquidity exists. It could be more or less if you check again. This highlights an important fact about LSTs though. I said above they can be used as the functional equivalent of SOL in many contexts, but there are important differences. This is one of them: you’re carrying just a touch of duration risk. If you need the capital mega now, and you need a lot of it, you’re going to have to discount its value to get it.
A simple incident can illustrate the point. Consider the case of the “mSOL depeg” that occurred on 12 December, 2023. This wallet address 85b5jKkgSuopF3MUA9s4zsBhRANrererBLRx689PqTPA, over the course of about 20 minutes, swapped about 68,536 mSOL for SOL on the open market using 9 transactions. This caused the price of mSOL to devalue from about $78 to $66. Here is a 5 minute chart from birdeye.so:
You can see the price came back up relatively quickly to the price it was before. Why? Arb bots and assorted other opportunists who noticed jumped in to buy the mSOL back up because they didn’t need the capital within the next epoch. The price of the underlying SOL didn’t really change, so they were essentially buying discounted SOL. For more analysis of this scenario you can refer to @y2kappa’s excellent X thread: https://twitter.com/y2kappa/status/1735104884138258896.
This scenario is not unique to mSOL, it applies equally to every liquid staking token (with some caveats we’ll explore later). The liquidity of any specific LST will inevitably be less than the liquidity of the SOL token. However, this is mostly something to worry about if you are blessed to have a lot of liquid staked SOL. More money, more problems.
So the lesson here is that liquid staked tokens carry some of the duration risk of regular staked SOL, but only insofar as there is a lack of liquidity in the market. The deeper the LST’s markets, the less the duration risk matters. If your stack is relatively modest, you don’t really feel this duration risk at all. Still, it’s important to know there is a limit to how liquid the token can be.
There are three LSTs on Solana with > $100 million total value locked that highlight different approaches to incentivizing adoption: Jito (jitoSOL), Marinade (mSOL), and BlazeStake (bSOL). This is obviously also subject to change, but you can see rough numbers here: https://defillama.com/protocols/Liquid%20Staking/Solana. Jito is in first place at time of writing so we’ll start there.
Jito’s Value Proposition
In Jito’s blog post on the subject they claim four reasons why you should stake with them (https://www.jito.network/blog/an-invitation-to-stake-with-the-jito-protocol/), but one is really the differentiator: Maximum extractable value (MEV). MEV is a deep topic, but in short, traders can take advantage of situations where there is a profitable way to order transactions such that they derive value from trades. Solana’s design (mainly fast block times, transactions executed in parallel) actually make MEV much more difficult to extract than in a system like Ethereum with a public mempool, but it’s still there. Some object to MEV entirely because it usually results in worse prices for other users and it is considered parasitic. Not all MEV is strictly harming other users, and in any case, these are open permissionless systems, so no one can be stopped from doing it anyway. You might as well complain that carnivorous animals eat other animals; this is just now nature is.
So the best thing to do is to stop complaining and start profiting. That’s what Jito is offering more or less: a share of the take. They have modified the original Solana Labs validator software to add the capability for validators to accept ordered transactions in exchange for a fee. This creates a more orderly and accessible market for MEV. The validators can then pass some of this extra SOL through to stakers who delegate to them in the form of yield. So this is what you’re getting when you delegate your SOL to Jito’s liquid staking pool. Importantly, other LSTs can also delegate their tokens to validators running Jito’s modified validator software though, so it’s not completely exclusive to jitoSOL. Marinade and BlazeStake both also benefit to some degree from the fact that validators they delegate to leverage Jito’s MEV module.
Marinade’s Value Proposition
Marinade was Solana’s first liquid staking protocol and has pioneered many best practices for the category. As they point out in their blog post, when you stake directly, you have to choose a specific validator to delegate to (https://marinade.finance/blog/introducing-protected-staking-rewards/). How do you know which is the right one? Liquid staking pools handle this for you and also distribute your stake across numerous validators, reducing the risk that one of them will fail to maintain their validator system or otherwise behave sub-optimally. Last month Marinade announced an initiative to hold validators accountable for their performance by requiring them to create what is basically an insurance fund. If they drop the ball and fail to maintain optimal performance, they will have to reimburse those who have delegated to them out of the fund.
Marinade also employs a feature they call directed stake, whereby stakers do choose a validator to support, and in return the validator may offer extra incentives. “Directed stake” actually only applies to 20% of the staker’s total and the rest is distributed based on Marinade’s delegation strategy algorithm.
Finally, Marinade is governed by the MNDE token. MNDE holders can vote on decisions regarding how the protocol operates and has a value independent of mSOL. Marinade sometimes runs campaigns, as it is currently doing, whereby certain actions generate additional yield in the form of MNDE tokens that are distributed to stakers. This is intended to incentivize stakers to perform actions that the protocol deems beneficial, like providing liquidity to ensure that mSOL is instantly redeemable. These incentives are a bit complicated but we’ll save some of that for later. For more details on their current campaign you can refer to their blog post here: https://marinade.finance/blog/let-marinade-earn-season-2-begin/.
BlazeStake’s Value Proposition
BlazeStake’s main discriminator is tied to the BLZE governance token. Like MNDE, BLZE can be used to vote on how incentives are distributed, but is also independently valued. BlazeStake is a relatively young project, so they’ve only distributed about 80% of the total supply of tokens whereas Marinade has fully distributed their token and has to repurchase them to create MNDE token incentives. That could be good or bad depending on your perspective and your investment time horizon. Jito also has a governance token JTO but it is not currently used to incentivize staking with Jito. Let’s quickly compare the top three governance tokens across TVL in the protocol, market cap, circulating supply, and growth trends:
Numbers retrieved on 16 February, 2024 from Defillama, Coingecko, and Birdeye
Focusing on the ratio of total value locked (assets the protocol controls) to the fully-diluted market cap (FMDC) of the governance token that exercises control over said assets, BLZE looks pretty good. Now, there are plenty of caveats here. One of the reasons JTO’s FDMC is so high is that circulating supply is so low on a comparative basis. There are reasonable arguments why FDMC is “just a meme” in the short term. No one cares until they do, which is when the next tranche of tokens is unlocked. BLZE also has only 2/3 of the token supply unlocked so there are still new tokens flowing into the market as emissions. To understand the schedule by which they’re being unlocked you can refer to this thread: https://twitter.com/solblaze_org/status/1688480225255161856.
What can you do with BLZE that would make it valuable? Some similar things to what you can do with MNDE, namely, vote on governance proposals that direct incentives to accomplish specific goals that help with bSOL adoption. However, BlazeStake’s does this via a mechanism they call stake gauges that gives users much more constant and fine-grained control. The two basic options are to vote in their Realms.today decentralized autonomous organization (DAO) to direct extra stake to specific validators, or to direct more BLZE rewards to specific liquidity pools in DeFi. BLZE owners can also lock their BLZE for a period of time up to 5 years in order to boost their voting power within the DAO. Here is a screenshot of the stake gauge UI with some sample options where you can direct your votes after you’ve deposited your BLZE in the DAO. This is all pretty unique as far as I can tell among LSTs.
Some Defi protocols require you to claim the BLZE rewards within their UI, but otherwise, your BLZE rewards are airdropped directly to you on an ongoing basis, about every two weeks. You can check your wallet’s current SolBlaze score at rewards.solblaze.org. There is no real advantage to trying to sybil this; more wallets doesn’t improve your score. The basic formula is 1 bSOL in your wallet = 1 point, 1 bSOL in a supported lending protocol = 1.5 points, and 1 bSOL in supported bSOL LP positions = 2 points. This is roughly in line with the risks you’re taking with the capital, so more risk granting more reward makes sense.
Also worth noting is that BlazeStake offers an option to stake your SOL with a single validator they call Custom Liquid Staking. Unlike with Marinade, 100% of the stake goes to that validator. Reminder from the above that with Marinade, only 20% of the stake is directed to the validator of your choice and the remaining 80% is distributed via their delegation strategy algorithm. Marinade outlines this in their documentation, but it’s not immediately apparent in their UI which I think is a little suboptimal.
BlazeStake is doing some other interesting things as well. They provide a real SOL (not testnet or devnet SOL) faucet you can use to get a little SOL for transaction fees when you’ve screwed up and accidentally staked all your SOL with them and don’t have enough to cover the transaction fee to unstake. This is really nice as it prevents you from having to onramp new funds through a centralized exchange. They also have a simple token mint UI that helps you launch a SPL token yourself, an RPC status page, and a SOL Pay SDK. They’re all good initiatives that help spread the SOL and liquid staking gospel.
If BlazeStake’s main value proposition is derived from its governance token’s emissions, price action on BLZE is a driver of value here. BLZE price shot up starting at the end of November 2023 and has somewhat stabilized in the $0.002 — $0.004 zone, but based on the above stats, there is likely a good deal of room to run. If it normalizes at Marinade-like levels, there is another easy 2.8x.
In my opinion, a similar or even higher valuation of BLZE compared to MNDE is warranted. I’m not sure how to justify JTO’s valuation relative to the other two. I like the project, what they’re doing with MEV on Solana is fundamentally unique and I expect new innovations in the future, but with almost 90% of token supply still to be issued it seems high. But fully diluted market cap doesn’t matter… until it does. Regardless, I like the chances of all three to appreciate relative to USD because these LST protocols are creating value. Simple thesis: LST good.
Concluding this section on the top three LSTs on Solana, they all tout the benefits to Solana decentralization. Liquid staking pools distribute stake across a wide validator set rather than setting up a winner-take scenario. They all credibly tout their security. There is smart contract risk with every LST, the risk that there is a bug in the smart contract that operates the stake pool, but they’re all credibly security conscious. These things are good but they don’t really differentiate the three approaches, which is why I’ve tried to highlight the unique features above.
There are many other LSTs on Solana and there are about to be a lot more thanks to the liquidity backstop pool sanctum.so is working on. Their aim is to make any LST as liquid as possible by accepting all of them and providing the exchange for SOL. They charge a fee less than 0.03% and usually 0.01% for what basically amounts to taking on the duration risk for you. You get the SOL in exchange for your LST instantly, and the pool provides a buffer that spreads that duration risk across a much wider surface area.
Sanctum supports twelve different LSTs at time of writing, but they aim to solve the liquidity bootstrapping problem for basically all LSTs. LSTs are only as liquid as there is available liquidity sitting in liquidity pools on protocols like Orca and Radium, so new LSTs would normally have to come up with a strategy for boosting this to deliver on the liquid part of the LST promise. Sanctum provides a huge extra buffer to make it possible for new and low market share LSTs to be liquid instantly.
The LSTs you can exchange for SOL instantly using sanctum at time of writing are: bSOL, cgntSOL, daoSOL, eSOL, jitoSOL, JSOL, laineSOL, LST (Marginfi’s liquid staking token), mSOL, riskSOL, scnSOL, and stSOL. Now that you know something about the various strategies the top three LSTs are following to make their approach attractive, you should be able to research and more critically evaluate some of the above smaller ones. Sanctum’s universal LST liquidity pool makes possible greater experimentation in the field of LSTs and maybe there is an even better fit for your goals and interests with one of the above.
I’ll leave you with one more resource if you want to dive deeper. In December 2023, a cofounder of Sanctum who goes by “J” did an interview with George Harrap that is with a listen: https://open.spotify.com/episode/2z6ykkF14oQ3nqKiGcNm2J?si=5d4276003c454cca.
Now that you know something about the options, let’s briefly look at simply how to do it. Using bSOL as the example, simply go to stake.solblaze.org, click on Stake tab, and you’re presented with this UI.
Remember that bSOL like most LSTs on Solana is reward-bearing, so you get back less bSOL than you submitted in SOL. Don’t let this scare you. In the image you see 0.8993 bSOL = 1 SOL. This is because 0.8993 bSOL represents a claim on the BlazeStake stake pool that is equivalent to 1 SOL, so you’re not losing any value. Over time as the stake pool’s SOL holdings grow, this number will continue to go down, meaning the amount of SOL you get back per bSOL will continue to grow. Right now 1 bSOL = 1.11 SOL and this number will continue to go up over time.
Choose the amount, click the button, approve the transaction, and you’re done.
Now that we’re armed with a decent understanding of how LSTs work, what the options are for LSTs and the benefits of different approaches, let’s look at Defi options. What is the point of all of this? The reason to have your staked tokens be liquid is so that you can do stuff with it.
I spent a good part of last year exploring different DeFi options on Solana. A lot depends on your appetite for risk and how you evaluate those risks compared to the rewards. That is a personal calculation so there is no right answer, but it’s important to understand what the risks and rewards are so you can even do such a calculation. Let’s start with something relatively simple, lending.
Lending
One of the simplest, lowest risk things you can do with an LST is to simply lend it. Platforms like MarginFi, Solend, and Kamino allow users to deposit collateral and borrow other assets of their choice. Crypto generally is pretty volatile, all these offer only over-collateralized lending positions. This means that your counterparty has to provide more USD value as collateral than they are borrowing. This generally varies based on some judgement of the quality of the asset used as collateral. If the value of the deposited collateral assets falls below a certain threshold, this collateral will be liquidated and used to repay the lenders.
The rules for how this happen are somewhat complicated and different projects take different approaches. If you’re going to put a lot of capital at risk, you should absolutely understand what they are. After the mSOL “depeg” event mentioned above, the founders of Solend and MarginFi had a debate on the Lightspeed podcast where they compared their risk management philosophies, if you want a full 90 minute intro (https://open.spotify.com/episode/6VJqS9cvHZjQJRaLLVGVuU?si=cb184be3fa8b4949.)
Usually APRs on lending for LSTs are relatively low because the demand isn’t that high. Since it’s already staked, the largest, lowest-risk yield opportunity has already been claimed. Still, you can get a little more yield, or maybe some protocol rewards by lending it in a relatively safe manner, so it’s an option worth considering.
Looping
Now that we’ve talked about lending, let’s talk about what lending enables. Say you have 10 bSOL and you’re getting their currently advertised APY of 7.34% (remember this changes slightly with literally every epoch), plus 0.81% APY in the form of BLZE emissions (and this is dependent on BLZE price). How can you generate more yield? One way is to deposit that bSOL in a lending protocol, borrow more SOL, then stake that SOL with BlazeStake too. Drift Protocol and Kamino Finance offer a straightforward product that executes this with one click as well as variable leverage you can configure. You could also do this manually via a protocol like Solend. The important variable to pay attention to is the SOL borrow rate. The higher this is, the lower your overall APR will be. Why? Because you have to deduct this as a cost. Here is a quick example from Drift’s SuperStakeSOL UI.
In this example, you’re borrowing 5 SOL using 10 bSOL as collateral. You must pay 0.6% APR to borrow the SOL, but you can then restake that SOL with bSOL again and earn that extra yield again. As long as the borrow rate is lower than the rewards, this makes sense. Remember that you’re taking on risk here though. The primary risk is a “depeg” event like the above one outlined for mSOL. This is always possible with LSTs because the mechanism which ties their value to SOL has a duration component.
Providing Liquidity
Another way to generate more yield on your LSTs is by providing liquidity. The thing that underlies decentralized exchanges is liquidity pools that enable you to trade pairs of tokens. If there are enough of these pools, you can trade any asset for any other asset by daisy chaining swaps through a list of pools. Since USDC and SOL usually have the deepest liquidity, any route usually comes back through those, so for example if you wanted to trade WIF for WHALES, you’d probably exchange the WIF for SOL, and then the SOL for WHALES.
The pools charge a fee for the swap and this fee goes back to the liquidity providers (LPs). When trades are in balance and the same amount of WIF is being swapped for WHALES as WHALES is for WIF, then the prices stay stable and LPs just extract fees. The reason this fee is justified though is the risk of impermanent loss. If demand starts to tilt in solidly in the direction of WHALES, the amount of WHALES in the pool diminishes and the amount of WIF in the pool constantly increase until the LPs are holding nothing but WIF and the WIF is now worth less than it was before. This trend could obviously always reverse, which is why it’s called impermanent loss, but it’s something to keep in mind. Here is a video to provide a deeper explanation with visuals (https://www.youtube.com/watch?v=_m6Mowq3Ptk).
So what pools are attractive for LSTs? First, the SOL-LST pool is going to see a lot of use because it supports the instant stake/unstake action. Second, LST1-LST2 type pools are nice because you’re holding two LSTs which are both earning their normal staking yield, and also collecting a small trading fee on top of that, while knowing that their prices should be highly correlated because they in turn are both correlated to SOL price (though again, not pegged, as we’ve discussed). The risk of impermanent loss is about as low as it can get. Third, for a higher risk, higher reward option, you may want to consider pools that allow you to trade an LST for their protocol’s governance token, like jitoSOL-JTO or bSOL-BLZE. These are often incentivized with extra governance token rewards by the protocol to ensure there is reasonable liquidity.
I can’t emphasize enough that there is a lot more nuance worth understanding with LPing. There are differences between pools in how they distribute liquidity over the pool, and how much control over this the LP has. Different approaches will be better or worse depending on how the tokens trade against one another, so you need a good mental model for how you think the price action will unfold, and therefore, where you want your liquidity in the pool in order to meet with consistently good results. The first two options above (SOL-LST, LST1-LST2), your mental model is basically “they’re going to continue to be highly correlated” which makes this really simple. If you’re going further afield than simple options like that, I’d recommend starting with very small amounts and observing how the price action unfolds. Notice how the pool balances change as they assets trade, how the fees accrue, and decide whether it’s worth it to go bigger.
Some DeFi protocols like Kamino offer vaults that automatically manage the LP strategy so you don’t have to do all that much. They come with a pre-built strategy for deploying the capital into liquidity pools they follow. You should understand what it is before you do it, but it means you don’t have to go into the pool manually and rebalance your ranges. They of course charge a small fee for this.
The trio of lending, looping, and LPing doesn’t exhaust your options with LSTs but I don’t want to turn this into a dissertation. If you haven’t tried all three of these, you may want to check them out before you go deeper.
Ok, if you read this whole thing: I’m sorry that happened, or congratulations. I realize that was quite a journey, but hopefully it’s equipped you with a solid understanding of the what, the why, and the how of liquid staking on Solana. As you can probably tell, I really like this stuff and I’m excited for future of LSTs and DeFi on Solana. There are so many good options available right now and pace of innovation in this space has been breathtaking.
If you spot any errors in the article you think need correction, or you think I’ve mischaracterized anything, I’m happy to take the feedback and consider changes. I could have delved much deeper into every one of these sections with more charts, more data and analysis, more description of the various features of different LSTs, but I tried to balance conciseness with comprehensiveness. All Solana LSTs and DeFi protocols are good boys. I have yet to discover a project I thought was simply misguided, predatory, or dumb. They all have strengths and weaknesses and serve slightly different goals and target customers, which is amazing for the future of the Solana ecosystem.
The amazing feature of Solana is that you can try pretty much all of this with $10. If you’re worried about the risks involved and whether your grasp of them is sufficient to justify giving this a try, you can just start small and get a feel for it. The only one that can evaluate your risk appetite, and whether any given opportunity is worth it, is you. I hope this article has helped inform your choices.
My aim here is to provide a comprehensive introduction to Solana liquid staking. I just realized in doing research for this Phantom has a somewhat similarly structured article (https://phantom.app/learn/crypto-101/solana-liquid-staking), and I’m going to cover some of the same territory, but with a lot more opinions and editorializing. I want you to know not just what liquid staking is and how to do it, but why. What would possess a man to spring out of bed in the morning and say to himself, “I’m going to liquid stake my SOL today”? Come, join me on my magic carpet, and I will show you a whole new world.
I will try to make it accessible to newcomers without completely boring the reader to death with over-explanation (you can judge whether I’ve succeeded). I’ve also tried to structure this rather long article so you can skip certain sections if you’re already familiar with the topic. With that in mind, here is the table of contents:
I’ll also give the obligatory disclaimer that none of this should be construed to be financial, legal, or in Haseeb Qureshi’s wonderful addendum, life advice. This article is meant to educate, and hopefully, delight and entertain.
Before we jump right into liquid staking, let’s talk about regular staking. Staking in a delegated proof-of-stake network like Solana is the act of committing your tokens to a validator who must promise to faithfully verify transactions on the network or risk penalties. This is what creates fundamental alignment between the validators and the users of the network, without which, double spending, censoring, and all other manner of abuse would be possible. When you stake “natively” you choose a specific validator and delegate your tokens to them. You can perform this operation through an array of wallet software or with the Solana command line interface.
Since the launch of Solana mainnet-beta in February 2020, Solana has been following this proposed inflation schedule:
As I’m writing this in February 2024, we’re four years from that launch date, so it’s relatively easy to see the current inflation is about 5%. The precise inflation amount is governed by three parameters: initial inflation rate (8%), disinflation rate (-15%), and the long-term inflation rate (1.5%). Inflation started at 8% and decreased 15% on an annualized basis at every epoch boundary until it comes to rest at 1.5%. This could be changed in the future, but this is the schedule that has been observed since launch.
Who gets the SOL tokens that are created through inflation? It’s pretty straightforward: stakers. This means that every epoch, stakers are increasing their relative ownership of the total amount of SOL tokens at the expense of non-stakers. Nothing more complicated is really happening. If all SOL were staked, no one would be increasing the total value of their holdings. This leads to a high stake rate on the Solana network; about 2/3 of all SOL is staked at time of writing. The rate of liquid staking, however, is still under
With a proof-of-work blockchain, validators have a high burden of equipment and energy costs that force them to sell some (or maybe all) tokens they earn merely to break even. With a proof-of-stake network like Ethereum, these costs are very, very low, so there isn’t much sell pressure at all. With Solana, operating expenses for validators are slightly higher than Ethereum because validators must execute transactions as a part of consensus which imposes a cost and validator equipment is slightly more expensive because it needs higher performance than Ethereum. So validator sell pressure is still very low compared to Bitcoin, but a bit higher than Ethereum.
I’ve taken some pains here to detail this because people keep getting confused about this. I occasionally rant about this on there interwebz. Ok, I admitted above there is a tiny bit of sell pressure because running a validator isn’t cost free, but the crypto Twitter doesn’t much appreciate nuance, so I went a little more aggressive. Anyway, with this essential background on regular staking in hand, let’s talk about liquid staking.
Since there is a high incentive to stake on Solana so as to not be diluted by the other stakers, one of the few reasons not to is that it locks your capital up each epoch. With liquid staking you contribute your tokens to a stake pool that manages the distribution of the stake across a set of validators, and tokenizing the fact that you have committed your tokens to the pool. This action returns a new asset to the staker that represents that fact and allows the user to redeem it for the original staked SOL. Thus, it can be used as a functional equivalent to SOL in many contexts.
Most popular liquid staking tokens (LSTs for short) on Solana are “reward-bearing” tokens. Almost all of the SOL in the pool is staked to validators chosen by the pool operators (sometimes minus a small buffer which facilitates quick redemptions), so this staked SOL in the pool accrues rewards in the form of more SOL. Thus, the LST doesn’t change in quantity over time, but comes to represent more SOL each epoch and so it appreciates in price relative to SOL. Another approach is “rebasing” whereby holders of the LST are awarded more of the liquid token and each liquid token is redeemable 1:1 (with a delay), but this is not much used on Solana.
Each epoch takes about 2.5 days on Solana. If a user wants to withdraw vanilla staked SOL, they must submit a request to withdraw their delegated stake wait until the end of the epoch and then it becomes redeemable. In traditional financial markets this is framed as taking on duration risk. You are making a bet that the reward of locking up the capital for 2.5 days will outweigh the risk that you will need it immediately. 2.5 days is much less of a risk, duration-wise, than a 10-year US Treasury note. On the other hand, US Treasuries are… a bit less volatile than crypto, generally speaking.
So, when a holder of a liquid staked token wants to access the underlying SOL token, they can redeem it from the pool of staked tokens controlled by the liquid staking protocol, opting either to wait until the end of the epoch when it can be unstaked by the underlying pool, or by using existing liquidity pools to trade the mSOL for SOL on the open market. Here is an example from Marinade Finance:
Marinade is telling you it’s using the Jupiter decentralized exchange aggregator to trade the mSOL for SOL. I checked unstaking 10,000 mSOL and the price impact was 0.01%, but at 100,000 mSOL, the price impact shown above is 100,000 (I don’t actually have 100,000 or even 10,000 mSOL, but it lets you simulate it). What does this mean?
It means you’re getting 8.162% less SOL than you would if you waited until the end of the epoch and unstaked directly. The price impact above is based on current market conditions, i.e. how much liquidity exists. It could be more or less if you check again. This highlights an important fact about LSTs though. I said above they can be used as the functional equivalent of SOL in many contexts, but there are important differences. This is one of them: you’re carrying just a touch of duration risk. If you need the capital mega now, and you need a lot of it, you’re going to have to discount its value to get it.
A simple incident can illustrate the point. Consider the case of the “mSOL depeg” that occurred on 12 December, 2023. This wallet address 85b5jKkgSuopF3MUA9s4zsBhRANrererBLRx689PqTPA, over the course of about 20 minutes, swapped about 68,536 mSOL for SOL on the open market using 9 transactions. This caused the price of mSOL to devalue from about $78 to $66. Here is a 5 minute chart from birdeye.so:
You can see the price came back up relatively quickly to the price it was before. Why? Arb bots and assorted other opportunists who noticed jumped in to buy the mSOL back up because they didn’t need the capital within the next epoch. The price of the underlying SOL didn’t really change, so they were essentially buying discounted SOL. For more analysis of this scenario you can refer to @y2kappa’s excellent X thread: https://twitter.com/y2kappa/status/1735104884138258896.
This scenario is not unique to mSOL, it applies equally to every liquid staking token (with some caveats we’ll explore later). The liquidity of any specific LST will inevitably be less than the liquidity of the SOL token. However, this is mostly something to worry about if you are blessed to have a lot of liquid staked SOL. More money, more problems.
So the lesson here is that liquid staked tokens carry some of the duration risk of regular staked SOL, but only insofar as there is a lack of liquidity in the market. The deeper the LST’s markets, the less the duration risk matters. If your stack is relatively modest, you don’t really feel this duration risk at all. Still, it’s important to know there is a limit to how liquid the token can be.
There are three LSTs on Solana with > $100 million total value locked that highlight different approaches to incentivizing adoption: Jito (jitoSOL), Marinade (mSOL), and BlazeStake (bSOL). This is obviously also subject to change, but you can see rough numbers here: https://defillama.com/protocols/Liquid%20Staking/Solana. Jito is in first place at time of writing so we’ll start there.
Jito’s Value Proposition
In Jito’s blog post on the subject they claim four reasons why you should stake with them (https://www.jito.network/blog/an-invitation-to-stake-with-the-jito-protocol/), but one is really the differentiator: Maximum extractable value (MEV). MEV is a deep topic, but in short, traders can take advantage of situations where there is a profitable way to order transactions such that they derive value from trades. Solana’s design (mainly fast block times, transactions executed in parallel) actually make MEV much more difficult to extract than in a system like Ethereum with a public mempool, but it’s still there. Some object to MEV entirely because it usually results in worse prices for other users and it is considered parasitic. Not all MEV is strictly harming other users, and in any case, these are open permissionless systems, so no one can be stopped from doing it anyway. You might as well complain that carnivorous animals eat other animals; this is just now nature is.
So the best thing to do is to stop complaining and start profiting. That’s what Jito is offering more or less: a share of the take. They have modified the original Solana Labs validator software to add the capability for validators to accept ordered transactions in exchange for a fee. This creates a more orderly and accessible market for MEV. The validators can then pass some of this extra SOL through to stakers who delegate to them in the form of yield. So this is what you’re getting when you delegate your SOL to Jito’s liquid staking pool. Importantly, other LSTs can also delegate their tokens to validators running Jito’s modified validator software though, so it’s not completely exclusive to jitoSOL. Marinade and BlazeStake both also benefit to some degree from the fact that validators they delegate to leverage Jito’s MEV module.
Marinade’s Value Proposition
Marinade was Solana’s first liquid staking protocol and has pioneered many best practices for the category. As they point out in their blog post, when you stake directly, you have to choose a specific validator to delegate to (https://marinade.finance/blog/introducing-protected-staking-rewards/). How do you know which is the right one? Liquid staking pools handle this for you and also distribute your stake across numerous validators, reducing the risk that one of them will fail to maintain their validator system or otherwise behave sub-optimally. Last month Marinade announced an initiative to hold validators accountable for their performance by requiring them to create what is basically an insurance fund. If they drop the ball and fail to maintain optimal performance, they will have to reimburse those who have delegated to them out of the fund.
Marinade also employs a feature they call directed stake, whereby stakers do choose a validator to support, and in return the validator may offer extra incentives. “Directed stake” actually only applies to 20% of the staker’s total and the rest is distributed based on Marinade’s delegation strategy algorithm.
Finally, Marinade is governed by the MNDE token. MNDE holders can vote on decisions regarding how the protocol operates and has a value independent of mSOL. Marinade sometimes runs campaigns, as it is currently doing, whereby certain actions generate additional yield in the form of MNDE tokens that are distributed to stakers. This is intended to incentivize stakers to perform actions that the protocol deems beneficial, like providing liquidity to ensure that mSOL is instantly redeemable. These incentives are a bit complicated but we’ll save some of that for later. For more details on their current campaign you can refer to their blog post here: https://marinade.finance/blog/let-marinade-earn-season-2-begin/.
BlazeStake’s Value Proposition
BlazeStake’s main discriminator is tied to the BLZE governance token. Like MNDE, BLZE can be used to vote on how incentives are distributed, but is also independently valued. BlazeStake is a relatively young project, so they’ve only distributed about 80% of the total supply of tokens whereas Marinade has fully distributed their token and has to repurchase them to create MNDE token incentives. That could be good or bad depending on your perspective and your investment time horizon. Jito also has a governance token JTO but it is not currently used to incentivize staking with Jito. Let’s quickly compare the top three governance tokens across TVL in the protocol, market cap, circulating supply, and growth trends:
Numbers retrieved on 16 February, 2024 from Defillama, Coingecko, and Birdeye
Focusing on the ratio of total value locked (assets the protocol controls) to the fully-diluted market cap (FMDC) of the governance token that exercises control over said assets, BLZE looks pretty good. Now, there are plenty of caveats here. One of the reasons JTO’s FDMC is so high is that circulating supply is so low on a comparative basis. There are reasonable arguments why FDMC is “just a meme” in the short term. No one cares until they do, which is when the next tranche of tokens is unlocked. BLZE also has only 2/3 of the token supply unlocked so there are still new tokens flowing into the market as emissions. To understand the schedule by which they’re being unlocked you can refer to this thread: https://twitter.com/solblaze_org/status/1688480225255161856.
What can you do with BLZE that would make it valuable? Some similar things to what you can do with MNDE, namely, vote on governance proposals that direct incentives to accomplish specific goals that help with bSOL adoption. However, BlazeStake’s does this via a mechanism they call stake gauges that gives users much more constant and fine-grained control. The two basic options are to vote in their Realms.today decentralized autonomous organization (DAO) to direct extra stake to specific validators, or to direct more BLZE rewards to specific liquidity pools in DeFi. BLZE owners can also lock their BLZE for a period of time up to 5 years in order to boost their voting power within the DAO. Here is a screenshot of the stake gauge UI with some sample options where you can direct your votes after you’ve deposited your BLZE in the DAO. This is all pretty unique as far as I can tell among LSTs.
Some Defi protocols require you to claim the BLZE rewards within their UI, but otherwise, your BLZE rewards are airdropped directly to you on an ongoing basis, about every two weeks. You can check your wallet’s current SolBlaze score at rewards.solblaze.org. There is no real advantage to trying to sybil this; more wallets doesn’t improve your score. The basic formula is 1 bSOL in your wallet = 1 point, 1 bSOL in a supported lending protocol = 1.5 points, and 1 bSOL in supported bSOL LP positions = 2 points. This is roughly in line with the risks you’re taking with the capital, so more risk granting more reward makes sense.
Also worth noting is that BlazeStake offers an option to stake your SOL with a single validator they call Custom Liquid Staking. Unlike with Marinade, 100% of the stake goes to that validator. Reminder from the above that with Marinade, only 20% of the stake is directed to the validator of your choice and the remaining 80% is distributed via their delegation strategy algorithm. Marinade outlines this in their documentation, but it’s not immediately apparent in their UI which I think is a little suboptimal.
BlazeStake is doing some other interesting things as well. They provide a real SOL (not testnet or devnet SOL) faucet you can use to get a little SOL for transaction fees when you’ve screwed up and accidentally staked all your SOL with them and don’t have enough to cover the transaction fee to unstake. This is really nice as it prevents you from having to onramp new funds through a centralized exchange. They also have a simple token mint UI that helps you launch a SPL token yourself, an RPC status page, and a SOL Pay SDK. They’re all good initiatives that help spread the SOL and liquid staking gospel.
If BlazeStake’s main value proposition is derived from its governance token’s emissions, price action on BLZE is a driver of value here. BLZE price shot up starting at the end of November 2023 and has somewhat stabilized in the $0.002 — $0.004 zone, but based on the above stats, there is likely a good deal of room to run. If it normalizes at Marinade-like levels, there is another easy 2.8x.
In my opinion, a similar or even higher valuation of BLZE compared to MNDE is warranted. I’m not sure how to justify JTO’s valuation relative to the other two. I like the project, what they’re doing with MEV on Solana is fundamentally unique and I expect new innovations in the future, but with almost 90% of token supply still to be issued it seems high. But fully diluted market cap doesn’t matter… until it does. Regardless, I like the chances of all three to appreciate relative to USD because these LST protocols are creating value. Simple thesis: LST good.
Concluding this section on the top three LSTs on Solana, they all tout the benefits to Solana decentralization. Liquid staking pools distribute stake across a wide validator set rather than setting up a winner-take scenario. They all credibly tout their security. There is smart contract risk with every LST, the risk that there is a bug in the smart contract that operates the stake pool, but they’re all credibly security conscious. These things are good but they don’t really differentiate the three approaches, which is why I’ve tried to highlight the unique features above.
There are many other LSTs on Solana and there are about to be a lot more thanks to the liquidity backstop pool sanctum.so is working on. Their aim is to make any LST as liquid as possible by accepting all of them and providing the exchange for SOL. They charge a fee less than 0.03% and usually 0.01% for what basically amounts to taking on the duration risk for you. You get the SOL in exchange for your LST instantly, and the pool provides a buffer that spreads that duration risk across a much wider surface area.
Sanctum supports twelve different LSTs at time of writing, but they aim to solve the liquidity bootstrapping problem for basically all LSTs. LSTs are only as liquid as there is available liquidity sitting in liquidity pools on protocols like Orca and Radium, so new LSTs would normally have to come up with a strategy for boosting this to deliver on the liquid part of the LST promise. Sanctum provides a huge extra buffer to make it possible for new and low market share LSTs to be liquid instantly.
The LSTs you can exchange for SOL instantly using sanctum at time of writing are: bSOL, cgntSOL, daoSOL, eSOL, jitoSOL, JSOL, laineSOL, LST (Marginfi’s liquid staking token), mSOL, riskSOL, scnSOL, and stSOL. Now that you know something about the various strategies the top three LSTs are following to make their approach attractive, you should be able to research and more critically evaluate some of the above smaller ones. Sanctum’s universal LST liquidity pool makes possible greater experimentation in the field of LSTs and maybe there is an even better fit for your goals and interests with one of the above.
I’ll leave you with one more resource if you want to dive deeper. In December 2023, a cofounder of Sanctum who goes by “J” did an interview with George Harrap that is with a listen: https://open.spotify.com/episode/2z6ykkF14oQ3nqKiGcNm2J?si=5d4276003c454cca.
Now that you know something about the options, let’s briefly look at simply how to do it. Using bSOL as the example, simply go to stake.solblaze.org, click on Stake tab, and you’re presented with this UI.
Remember that bSOL like most LSTs on Solana is reward-bearing, so you get back less bSOL than you submitted in SOL. Don’t let this scare you. In the image you see 0.8993 bSOL = 1 SOL. This is because 0.8993 bSOL represents a claim on the BlazeStake stake pool that is equivalent to 1 SOL, so you’re not losing any value. Over time as the stake pool’s SOL holdings grow, this number will continue to go down, meaning the amount of SOL you get back per bSOL will continue to grow. Right now 1 bSOL = 1.11 SOL and this number will continue to go up over time.
Choose the amount, click the button, approve the transaction, and you’re done.
Now that we’re armed with a decent understanding of how LSTs work, what the options are for LSTs and the benefits of different approaches, let’s look at Defi options. What is the point of all of this? The reason to have your staked tokens be liquid is so that you can do stuff with it.
I spent a good part of last year exploring different DeFi options on Solana. A lot depends on your appetite for risk and how you evaluate those risks compared to the rewards. That is a personal calculation so there is no right answer, but it’s important to understand what the risks and rewards are so you can even do such a calculation. Let’s start with something relatively simple, lending.
Lending
One of the simplest, lowest risk things you can do with an LST is to simply lend it. Platforms like MarginFi, Solend, and Kamino allow users to deposit collateral and borrow other assets of their choice. Crypto generally is pretty volatile, all these offer only over-collateralized lending positions. This means that your counterparty has to provide more USD value as collateral than they are borrowing. This generally varies based on some judgement of the quality of the asset used as collateral. If the value of the deposited collateral assets falls below a certain threshold, this collateral will be liquidated and used to repay the lenders.
The rules for how this happen are somewhat complicated and different projects take different approaches. If you’re going to put a lot of capital at risk, you should absolutely understand what they are. After the mSOL “depeg” event mentioned above, the founders of Solend and MarginFi had a debate on the Lightspeed podcast where they compared their risk management philosophies, if you want a full 90 minute intro (https://open.spotify.com/episode/6VJqS9cvHZjQJRaLLVGVuU?si=cb184be3fa8b4949.)
Usually APRs on lending for LSTs are relatively low because the demand isn’t that high. Since it’s already staked, the largest, lowest-risk yield opportunity has already been claimed. Still, you can get a little more yield, or maybe some protocol rewards by lending it in a relatively safe manner, so it’s an option worth considering.
Looping
Now that we’ve talked about lending, let’s talk about what lending enables. Say you have 10 bSOL and you’re getting their currently advertised APY of 7.34% (remember this changes slightly with literally every epoch), plus 0.81% APY in the form of BLZE emissions (and this is dependent on BLZE price). How can you generate more yield? One way is to deposit that bSOL in a lending protocol, borrow more SOL, then stake that SOL with BlazeStake too. Drift Protocol and Kamino Finance offer a straightforward product that executes this with one click as well as variable leverage you can configure. You could also do this manually via a protocol like Solend. The important variable to pay attention to is the SOL borrow rate. The higher this is, the lower your overall APR will be. Why? Because you have to deduct this as a cost. Here is a quick example from Drift’s SuperStakeSOL UI.
In this example, you’re borrowing 5 SOL using 10 bSOL as collateral. You must pay 0.6% APR to borrow the SOL, but you can then restake that SOL with bSOL again and earn that extra yield again. As long as the borrow rate is lower than the rewards, this makes sense. Remember that you’re taking on risk here though. The primary risk is a “depeg” event like the above one outlined for mSOL. This is always possible with LSTs because the mechanism which ties their value to SOL has a duration component.
Providing Liquidity
Another way to generate more yield on your LSTs is by providing liquidity. The thing that underlies decentralized exchanges is liquidity pools that enable you to trade pairs of tokens. If there are enough of these pools, you can trade any asset for any other asset by daisy chaining swaps through a list of pools. Since USDC and SOL usually have the deepest liquidity, any route usually comes back through those, so for example if you wanted to trade WIF for WHALES, you’d probably exchange the WIF for SOL, and then the SOL for WHALES.
The pools charge a fee for the swap and this fee goes back to the liquidity providers (LPs). When trades are in balance and the same amount of WIF is being swapped for WHALES as WHALES is for WIF, then the prices stay stable and LPs just extract fees. The reason this fee is justified though is the risk of impermanent loss. If demand starts to tilt in solidly in the direction of WHALES, the amount of WHALES in the pool diminishes and the amount of WIF in the pool constantly increase until the LPs are holding nothing but WIF and the WIF is now worth less than it was before. This trend could obviously always reverse, which is why it’s called impermanent loss, but it’s something to keep in mind. Here is a video to provide a deeper explanation with visuals (https://www.youtube.com/watch?v=_m6Mowq3Ptk).
So what pools are attractive for LSTs? First, the SOL-LST pool is going to see a lot of use because it supports the instant stake/unstake action. Second, LST1-LST2 type pools are nice because you’re holding two LSTs which are both earning their normal staking yield, and also collecting a small trading fee on top of that, while knowing that their prices should be highly correlated because they in turn are both correlated to SOL price (though again, not pegged, as we’ve discussed). The risk of impermanent loss is about as low as it can get. Third, for a higher risk, higher reward option, you may want to consider pools that allow you to trade an LST for their protocol’s governance token, like jitoSOL-JTO or bSOL-BLZE. These are often incentivized with extra governance token rewards by the protocol to ensure there is reasonable liquidity.
I can’t emphasize enough that there is a lot more nuance worth understanding with LPing. There are differences between pools in how they distribute liquidity over the pool, and how much control over this the LP has. Different approaches will be better or worse depending on how the tokens trade against one another, so you need a good mental model for how you think the price action will unfold, and therefore, where you want your liquidity in the pool in order to meet with consistently good results. The first two options above (SOL-LST, LST1-LST2), your mental model is basically “they’re going to continue to be highly correlated” which makes this really simple. If you’re going further afield than simple options like that, I’d recommend starting with very small amounts and observing how the price action unfolds. Notice how the pool balances change as they assets trade, how the fees accrue, and decide whether it’s worth it to go bigger.
Some DeFi protocols like Kamino offer vaults that automatically manage the LP strategy so you don’t have to do all that much. They come with a pre-built strategy for deploying the capital into liquidity pools they follow. You should understand what it is before you do it, but it means you don’t have to go into the pool manually and rebalance your ranges. They of course charge a small fee for this.
The trio of lending, looping, and LPing doesn’t exhaust your options with LSTs but I don’t want to turn this into a dissertation. If you haven’t tried all three of these, you may want to check them out before you go deeper.
Ok, if you read this whole thing: I’m sorry that happened, or congratulations. I realize that was quite a journey, but hopefully it’s equipped you with a solid understanding of the what, the why, and the how of liquid staking on Solana. As you can probably tell, I really like this stuff and I’m excited for future of LSTs and DeFi on Solana. There are so many good options available right now and pace of innovation in this space has been breathtaking.
If you spot any errors in the article you think need correction, or you think I’ve mischaracterized anything, I’m happy to take the feedback and consider changes. I could have delved much deeper into every one of these sections with more charts, more data and analysis, more description of the various features of different LSTs, but I tried to balance conciseness with comprehensiveness. All Solana LSTs and DeFi protocols are good boys. I have yet to discover a project I thought was simply misguided, predatory, or dumb. They all have strengths and weaknesses and serve slightly different goals and target customers, which is amazing for the future of the Solana ecosystem.
The amazing feature of Solana is that you can try pretty much all of this with $10. If you’re worried about the risks involved and whether your grasp of them is sufficient to justify giving this a try, you can just start small and get a feel for it. The only one that can evaluate your risk appetite, and whether any given opportunity is worth it, is you. I hope this article has helped inform your choices.