Solana: Light At The End Of The Tunnel?
by Ryan Shea
- It has been a rough couple of years for Solana, with its native token SOL having slumped more than 90% from its all-time high hit during the peak of the last crypto cycle.
- Part of the reason why SOL was hit hard during the cryptowinter was its strong association with SBF, the co-founder of bankrupt crypto exchange FTX, who is currently in court faces charges of wire fraud and money laundering.
- While the FTX fallout has overshadowed SOL for many months, sentiment has started to take a more positive turn recently.
- In early September, Rune Christensen, the co-founder of MakerDAO, proposed adopting the Solana tech stack for the defi lender’s back-end. The news went down well with Solana fans, but was less well-received by ETH bulls.
- MakerDAO is not the only entity interested in Solana. Visa – the giant retail payments processor – also recently announced it is running a pilot test on Solana to expand its stablecoin settlement capabilities.
- Does this mean the post-SBF hangover is finally over and it is now UpOnly for Solana? Read on to find out.
Solana – the ninth largest cryptocurrency by market cap1 - has had a rough couple of years. Having hit an all-time high in November 2021 at the peak of the last crypto cycle, Solana’s price has slumped by a staggering 92%.
Hero To Zero
There are a number of reasons why Solana ( SOL) was hit particularly hard by the cryptowinter but much of it has to do with Sam Bankman-Fried2, the disgraced founder of the misleadingly labelled crypto hedge fund Alameda Research and the crypto exchange FTX. SBF, as he was colloquially known, was a major supporter of Solana not only vocally but financially – see images below.
SBF’s spectacular fall from grace was triggered by the leaking of Alameda Research’s balance sheet in a Coindesk news story3 published on November 2, 2022. The article showed the hedge fund was stuffed full of FTX issued FTT tokens. For two companies founded by the same person, but supposedly segregated and operationally independent, this high degree of financial interconnectedness unsettled many in the crypto world and rightly so as it transpired, given that SBF is currently on trial for wire fraud and money laundering4. Soon after the news story broke, Changpeng Zhao (CZ) announced that Binance, another crypto exchange which had been an early investor in FTX and had accumulated substantial holdings of the FTT token as part of its 2021 divestment, would be selling their stash of FTT. This proved to be a tipping point. Just nine days after the story broke, both of SBF’s firms were forced into bankruptcy.
While Solana was not the catalyst for the problems that took down SBF’s crypto empire, his strong support and promotion of the SOL token – it was widely dubbed “Sam coin” - soured investor sentiment towards it. Even worse, it quickly came to light that Solana foundation had millions of dollars worth of cryptocurrencies held in accounts at FTX and, as a result of the Chapter 11 bankruptcy, these assets were effectively stranded. Little wonder that during this period the price of Solana nosedived more than 50%.
Unfortunately, the problems didn’t stop there. Solana’s price continued to be overshadowed by the fallout from FTX’s downfall. Last month a judge in the US Bankruptcy Court for the District of Delaware approved a motion allowing the sale of cryptocurrencies on FTX’s books5. Given FTX had over $1bn in SOL, worries that the market price might be adversely affected by the liquidation caused its price to slump more than 6%. This pessimism, however, proved to be misplaced because the liquidation order included a cap6 on the pace by which FTX’s crypto assets could be liquidated in order to optimize the recovery of funds for its creditors.
With this ruling out of the way, does this mark the end of the “SBF hangover” and a more constructive outlook for Solana? Recent events would suggest so.
Meet Your Maker
In early September Rune Christensen, co-founder of the decentralized crypto lender MakerDAO, posted a proposal on an internal forum outlining the fifth and final phase of his – let me be kind, complex7 - Endgame plan8. In the post9 he outlined why the project needs a “new, standalone blockchain (codename NewChain)” and he went on to recommend Solana as the best tech stack for the project.
This proposal marks quite a radical shift for Maker because since inception it has operated as an open-source project on Ethereum. Given Ethereum is the second largest cryptocurrency by market cap with a thriving ecosystem, the decision to transition to a forked version of a smaller blockchain might seem odd at face value.
Certainly, the suggested pivot by Maker away from Ethereum in favour of a forked Solana-blockchain for their back-end10 is not without its critics. Ethereum co-founder, Vitalek Buterin, in particular appears to have taken exception to the proposal given he dumped his entire MKR holdings just hours after Rune’s post was published – see image.
Most ETH bulls were critical of Rune’s proposal because in their view a forked Solana blockchain is no match for the security of the Ethereum blockchain which has over 700,000 validators who each have to put up a sizable stake (32ETH) to earn staking rewards and is highly decentralized11. While true, what such criticism fails to appreciate is that basing Maker’s back-end on a large, highly-decentralized, blockchain like Ethereum is actually a drawback given Maker’s ambitions. To understand and appreciate why we need to drill down into Rune’s post.
He kicks things off with the following statement:
“The most important reason for why NewChain is needed, is that it will allow the ecosystem to use hard forks to gracefully recover from the most severe form of governance attacks or technical failures.”
Court Not Code
Although not explicitly mentioned, it is imperative Maker is able to hard fork because of its successful Real World Asset (RWA) programme, which I discussed in an earlier research note. Since it was first launched back in 2022, RWA is now the single largest asset type held by Maker, totalling some $3bn. Moreover, it accounts for nearly two-thirds of the lending protocol’s overall revenue, meaning the RWA programme is quite a big deal for Maker – see images below.
RWA are an on-chain digital title of ownership of an off-chain asset such as real estate, commodities, equity (both public and private), fixed income and other high value collectibles. Because tokenization involves the transfer of legal rights to an asset that is not natively on-chain, the process has to comply with the laws of the jurisdiction where the asset resides and can be subject to potential court rulings.
While blockchain technology is far superior to the existing tradfi infrastructure – the reason why so many tradfi companies are exploring tokenized RWA – there is no way for a blockchain to guarantee that its state of the world matches that which is deemed legally valid. For example, hackers stealing RWA-backed tokens might be seen as legitimate from the perspective of the blockchain protocol, but it would not be compliant with the law. Similarly, a major technical failure of the protocol that saw RWA tokens locked up in a smart contract that its owners are unable to access would also be problematic from the perspective of the courts12.
The only way to mitigate this potential divergence is if the blockchain is hard-forkable to give Maker the ability to rollback transactions in order to comply with the decisions of the courts who, let us not forget, are the ultimate arbiter of truth in the off-chain world.
In short, if MakerDAO wants to continue to expand its successful RWA programme without running afoul of the regulators/legal system it will become necessary at some stage for Maker to have full sovereignty over its blockchain.
Achieving the necessary social consensus to hard-fork the entire Ethereum blockchain because of an issue that arises with MakerDAO, and doing so in a timely enough manner to assuage RWA-token holders and/or the courts, is a huge ask. Much better, therefore, for the protocol to transition to a smaller, bespoke, blockchain architecture.
A further advantage that Solana has over Ethereum is that it is a highly efficient blockchain, which is a characteristic Rune considers appropriate for Maker’s objectives13. Ethereum has, by contrast, very limited transactional bandwidth (12 tps in terms of Layer 1 transactions), whereas Solana averages 4000 tps and according to Solana foundation can handle up to 65,000 tps14.
Solana - Under The Hood
The reason why Solana has a vastly superior transactional bandwidth than Ethereum (and Bitcoin which maxes out at around 7 tps) is because of the way the protocol is designed. Like many other blockchains, including Ethereum after last year’s Merge, Solana uses Proof-of-Stake as part of its validation process. However, it combines this with Proof-of-History (PoH) – this is the clever innovation.
PoH provides a cryptographic method to verify that transaction X must have come before transaction Y, thereby enabling any validators to order transactions correctly regardless of the order in which they receive them. In Solana’s case, PoH relies on the fact that chaining hash functions is a computationally irreducible process. This is just a fancy way of saying that there is no way someone can skip ahead and calculate the hash value n steps in the future. The correct hash can only be found by repeatedly hashing, which necessarily takes time. Critically, this means chained hashing cannot be speeded up by using parallelized computing.
However, to ensure high performance, the verification that these hashes are correct needs to be done much faster. This is possible because while calculating a hash value n steps takes time - as it can only be done sequentially - verifying that a set of hash values are valid can be done in parallel, because all the hashes that need to be checked are known at the start of the verification process. And, GPUs, originally designed to power computer graphics for the gaming industry and more recently co-opted into AI inference, are the ideal architecture to carry out such highly parallelized hash verifications.
According to the Solana white paper15, the expected time to verify that a sequence of hashes is valid is a function of the number of cores able to verify the hash and modern GPUs have a lot of cores. For example, Nvidia’s RTX 4090 (the best desktop GPU currently available) has over 16,000 cores, which is more than sufficient compute power to make PoH work as a pseudo verifiable delay function (VDF)16.
Solana’s blockchain architecture is certainly elegant and, importantly, it loosens the constraints arising from the now famous blockchain dilemma. That said, it cannot completely overcome it.
In Solana’s case, and other cryptocurrencies that use PoS to ensure Sybil resistance to protect the integrity of the network, it is centralization that is sacrificed. To foster demand for the token in PoS systems, earning rewards from block creation/validation is determined probabilistically based on the number of tokens being staked. The greater the amount staked, the greater the probability of being chosen and hence the greater the expected staking rewards. The unfortunate side-effect of such tokenomics is that it also introduces centralizing forces within the protocol.
There is another centralizing force that often gets mentioned in relation to Solana and it arises directly because of its large transactional bandwidth. Supporting high tps rates means Solana nodes need to have quite high spec computing rigs. According to the Solana docs17, the hardware recommendations include a 12 core/24 thread CPU and 256GB of RAM. These are far and above what most people need, even gamers playing 4K games rarely need more than 32/64GB of RAM18. In addition, Solana also requires consistent internet speeds of at least 1GBbit/s symmetric (ie upload and download), something most home ISPs cannot provide. These fairly specialized compute requirements are more in keeping with servers/data centres than the desktops crypto users have at home.
While it is true that the requirements to run Solana nodes are significantly greater than say Bitcoin nodes, which can be easily spun up in a home environment, just like Bitcoin it is a permissionless process meaning anyone with sufficient compute capability can become a Solana node. At the time of writing, Solana has around 2,000 active validator nodes (and 925 non-block creating RPC nodes), so it is clear that the barriers of entry are not overly burdensome and the centralization risks arising from these tech requirements are exaggerated.
V For Visa
Rune is not alone in seeing the merits of Solana’s blockchain. At almost the same time as his proposal was published, Visa announced19 it was also looking to deploy the same tech stack for its stablecoin settlement pilot…
“...to test whether Solana has the ability to meet the demands of modern corporate treasury operations.”
Given Visa is the dominant player within the retail card payments system, processing 87 billion transactions worth more than $11.6bn last year, that they are even contemplating using Solana for its test pilot is considered by many in crypto as validation of the quality of the Solana project.
Low And Stable
Reading the Visa statement, it is clear that Solana is not only favoured because of its transactional bandwidth but also because of the fees associated with transferring value over its blockchain. Importantly, Solana fees are not only low, but stable as outlined in the following graphic included in their post – see image.
Average Transaction Fee
Source: Visa | Data: The Block and the Block Explorer
For the likes of Visa the stability of transaction fees is extremely important and Solana is able to ensure such stability because it has a localized fee market. Unlike Bitcoin and ETH, whose transaction fees can spike due to increased demand for block space, Solana has per-block limits and per-account limits that measure the amount of computation required by every transaction type. Once these limits are reached, users wishing to make similar transactions have to bid up fees to compete for the finite block space, but the fees for different transactions that are below their limits are unaffected. In this way, the spikes in transaction fees that have been witnessed in Ethereum e.g., due to a popular NFT mint can be greatly mitigated.
Bringing the likes of MakerDAO and Visa into the Solana ecosystem is clearly a big confidence boost after the mess of the last year as it shows increased confidence in their tech stack, but does that mean UpOnly for Solana’s price?
That is not quite so clear cut.
First, in terms of MakerDAO, if Rune’s proposal is accepted – which is not assured by any means because it must be voted on by the MKR delegates - and Maker adopts Solana’s tech stack for their back-end they will do so via a hard fork of the codebase. They will not be on Solana’s mainnet and it is also extremely unlikely they will use SOL as their native token. Given this, aside from the general confidence boost it gives crypto players in Solana’s tech stack, it is hard to see much of a direct benefit accruing to SOL holders from Maker pivoting in this way because it will not directly boost revenues via greater transactions or generate additional demand for SOL.
Of course, the same would not be true of Visa. If the pilot goes well and Visa starts to use Solana’s blockchain for processing stablecoin transactions it has the potential to significantly boost volumes. The question though is what would be the impact on overall revenues for Solana, which is what motivates people to purchase and own SOL?
At present, Solana generates an average daily fee revenue of $40,000. This is orders of magnitude smaller than Ethereum ($2.6mn), Bitcoin ($722,000) and considerably smaller than MakerDAO ($168,000)20. Clearly, this is not because of the inferiority of the Solana architecture. As I have argued, and as the likes of Maker and Visa considering using the Solana codebase confirms, the tech stack is good. It is instead a consequence of having low transaction fees. For Solana fee revenue to grow markedly and match those of other crypto projects, they need to substantially raise transaction volumes on their blockchain. Whether they can achieve such an outcome remains to be seen, but at least the recent omens are encouraging.
Until next time.
1At the time of writing Solana’s market cap was $256m – see: https://coinmarketcap.com/
2For any readers able to access the the BBC iplayer service, the UK broadcaster just published a 90 minute long documentary on Sam Bankman-Fried called Downfall of the Crypto King.
3The Coindesk article that broke the story has received many industry awards, most recently the Loeb Award – see: https://www.coindesk.com/business/2023/09/29/sam-bankman-fried-scoops-help-coindesk-win-major-journalism-award/
5 In it’s court filing FTX stated that it owned a total of $3.4 billion in cryptocurrencies, including $560 million in Bitcoin and $192 million in Ether – see: https://www.coindesk.com/policy/2023/09/11/ftx-holds-116b-in-sol-200m-in-bahamas-real-estate-court-filing-says/
6Initially the weekly limit was set at $50million, before increasing to the maximum weekly limit of $200m - see: https://storage.courtlistener.com/recap/gov.uscourts.deb.188450/gov.uscourts.deb.188450.2505.0.pdf
7Convoluted is probably more accurate.
10Rune made it clear that Solana fork would only apply to Maker’s back-end and that “Nothing will change on the user side, everything user facing like dai, spark protocol etc will remain on ethereum.” - see: https://twitter.com/RuneKek/status/1697764282144850180
11The decentralized crypto exchange dYdX has received similar criticism of choosing sovereignty over security when it decided to transition from Ethereum to Cosmos – see: https://www.coindesk.com/layer2/2022/06/29/a-major-crypto-exchange-abandons-ethereum-is-the-worlds-computer-falling-behind/
12This is very much a work-in-progress for Ethereum – see: https://ethresear.ch/t/recovering-from-smart-contract-hacks-forkable-reversible-roll-ups/16781
13See: As a counterpoint to this argument, some ETH fans suggested Maker deploys a Layer 2 Ethereum scaling solution. However, the issue there is, at least as of now, the majority of Layer 2’s rely on a single sequencer – or node - to relay information back to the Layer 1 blockchain, which is clearly more centralized than choosing to deploy a fork of the Solana codebase.
16Formally, chained hashing is not considered a verifiable delay function, so if you say this to a professional mathematician I suggest you stand well back.
17 Solana suggests that running a node a cloud may be possible, but may not be cost-effective over the long term – see: https://docs.solana.com/running-validator/validator-reqs
18I am a bit of a computer hardware nerd on the quiet.