The Solana Thesis: Internet Capital Markets

Multicoin Capital has been investing in Solana’s native asset, SOL, and the broader Solana ecosystem since Solana’s seed round in May 2018. We have previously published four theses about Solana in that time. The first two were published about nine months before mainnet produced its first block in March 2020. As the Solana network has evolved, so has our frame of reference on how to think about the Solana network, and SOL the asset.

Now that Solana is a $100B asset, has the fastest-growing developer ecosystem, and has surpassed Ethereum on most major on-chain metrics (trading volumes, daily active addresses, REV, TEV, DePIN payments, etc.), we wanted to share our thoughts on how we’re underwriting SOL for strong returns even above $100B market cap.

This essay is the 5th in our evolving Solana thesis. The first four are:

1. The Separation of Time and State
2. The World Computer Should be Logically Centralized
3. Technical Scalability Creates Social Scalability
4. The Hidden Costs of Modular Systems

In this essay, I’ll make the case that Solana is the leading chain to power Internet Capital Markets. Moreover, I’ll argue that Solana as a technology can outperform the major traditional finance (TradFi) players—including NYSE, NASDAQ, CME, JPM, GS, and MS on the financial markets side, and Visa and Mastercard on the payments side - on core performance metrics such as latency, while simultaneously preserving the core properties of blockchains that TradFi never offered (atomic composability, and permissionless access for users, developers, and validators). Most importantly, I’ll argue that the Solana ecosystem has the ability to concurrently accomplish both of the following objectives, even though they seem at odds with one another:

1. Reduce fees for financial services to end users by 90-99%
2. Capture more aggregate market cap than TradFi incumbents

While TradFi players like NYSE and NASDAQ offer one narrow slice of value in the finance stack, Solana already powers a superset of the features of these systems via the unique DeFi protocols that have been in production on Solana for years. Solana not only expands the total addressable market (TAM) for trading by increasing access and performance, it also captures value from more layers of the finance stack.

Broadly speaking, one can say that all financial services fit into one of two broad categories: payments and finance. I’ll start this post by explaining how payments are a loss leader for blockchains; after that, the bulk of this post will focus on the core plumbing that powers finance on Wall Street.

Delivering The Best Global Payments Experience

There are lots of ways to move money. Apple Pay is an awesome UX. Using a physical credit card is great. Using Venmo or PayPal or Square Cash is good. Everything else is meh, or worse—ACH, Wires, Zelle, Bill Pay, remittances, etc.

But even despite their UX, the fees of these legacy systems are nonsensical. Wires are $25, and credit cards can be more than 2%. It is insane that updating ledger entries can impose this scale of cost on consumers and merchants. This fails basic common sense, and directly violates the natural intuition that electronic transactions should be cheaper than analog transactions.

Solana trivializes payments and makes the UX incredible. And fees are almost zero. See this video from Sling Money, which is built exclusively on Solana. This is the future of money movement.

The market cap of payments companies around the world is around $1.4T. Solana aims to collapse that by 90%. The only cost that Solana itself imposes on users is gas, which is roughly one tenth of a penny per transaction, or $.001 per transaction. Even if the Solana network grows to sustain 50,000 transactions per second on average for an entire year, that’s only $1.5B in cost to users. For context, Visa sustains a few thousands transactions per second.

Payments are a loss leader for blockchains. They are important for driving adoption, and they provide real utility to users and companies, but they are not the primary source of profits for a blockchain or its ecosystem.

However, payments are paramount for the growth of blockchains. The beauty of payments is that they are inherently viral. When Alice sends money to Bob, and then Bob sends money to Carol, that leads to a natural growth in the adoption of wallets.

The primary source of profits for a blockchain is not payments, which effectively round to $0. Instead, the primary source of profits for a blockchain comes from the volatility that naturally occurs between asset prices, which manifests in the form of maximum extractable value (MEV). My cofounder Tushar explains this concept further in his presentation from the 2022 Multicoin Summit.

The rest of this essay will focus on how and why Solana can outperform TradFi on traditional performance metrics, and how that will allow SOL and the Solana ecosystem to capture profits.

Market Efficiency in CeFi and DeFi

Solana is a decentralized network of thousands of nodes that reach consensus on a series of financial transactions on a rolling 400ms basis (and will hopefully decrease to 120ms in the years ahead).

The correct way to measure market efficiency is not via latency for transaction inclusion but via spreads that’re offered by market-makers (MMs). At the end of the day, what buyers and sellers experience are prices. Human users—meaning non-bots—cannot experience the difference between 50ms, 100ms, and 200ms financial transactions. For context, the average time for a human eye blink is 100-150ms.

Market making in centralized finance (CeFi) is nearly-deterministic. Most MMs have a server that is co-located with the CeFi exchange, and each MM has a fiber cable of exactly the same length connecting their server to the exchange. The exchange finalizes transactions on a microsecond timescale, so MMs know their risk exposure with an extremely high degree of precision in real-time.

Decentralized Finance (DeFi) exchanges—such as Drift, Phoenix, Clearpools, Raydium, and Orca—by contrast are far less deterministic than CeFi exchanges because:

1. Solana’s network leaders are constantly rotating
2. The time to finality increases due to the need to reach consensus amongst validators around the world

As such, MMs cannot know their risk exposure in real-time with the same degree of precision. In many cases, MMs may leave stale prices on the blockchain order book that can get picked off by others.

Consequently, DeFi spreads are typically wider than CeFi spreads.

Let’s examine how these systems are changing to produce a better experience for both makers and takers.

Makers - Tightening Spreads via Conditional Liquidity

Things are changing. DFlow just quietly launched Conditional Liquidity (CL) on Solana. As the name implies, conditional liquidity is liquidity that is only available to take if the taker order meets some predefined condition. For the purposes of this essay, the condition that matters is toxic vs non-toxic order flow.

How does CL work? CL specifies that a given unit of liquidity can only be taken if the taker order is endorsed by a known front-end application. That includes wallets such as Phantom, Backpack, Solflare, and Fuse, and apps such as Drift, Kamino, Jupiter, and DFlow’s own front end. This mechanism guarantees that bots cannot consume CL because bot orders are not endorsed by an endorser. This is a huge improvement for MMs, as it virtually guarantees that they will not get picked off even if their quotes are several seconds delayed.

Although CL is mechanically a new concept, it is directly inspired by widely adopted practices in TradFi. Robinhood is the pioneer on this front. Robinhood consistently offers its customers better prices than the national best bid and offer (NBBO) that are quoted on NYSE and NASDAQ. They have demonstrated this pricing improvement empirically for trillions of dollars of trades over the last decade. This makes sense, because MMs have strong statistical reasons to believe that Robinhood users are less toxic on average than firms who are directly trading on NYSE or NASDAQ. Put simply, who would you rather face in a trade: Joe, who watches YouTube videos, or Citadel?

CL allows MMs to know that they are not facing the proverbial Citadel.

For more context on how order flow segmentation leads to better prices for retail traders, read this.

The beauty of DFlow’s CL is that it offers the best of both TradFi and crypto. It provides the ability to offer tighter spreads to retail traders such as Robinhood’s customers, and the real-time permissionless access and open auditability of a blockchain.

CL is a nascent concept. However, we expect it will become the dominant paradigm for quoting liquidity on-chain over the next few years because MMs hate getting picked off with stale quotes. Market making is fundamentally about quoting prices based on the maximum available information. There is simply no reason for market makers—both passive and active—not to incorporate more information (i.e., conditional liquidity) into their pricing.

DFlow’s CL implementation on Solana is 100% open source today and imposes no fees or taxes of any kind. Here is the GitHub repo.

Conditional liquidity is the most important functional improvement in DeFi since Uniswap launched the xyk automated market maker (AMM) in late 2018. As it is adopted, it will reshape all discussions in DeFi about UX, spreads, MEV, and more.

To re-iterate, CL will enable MMs to quote tighter for normal users. We expect this to be mutually beneficial for MMs, users, SOL, and the Solana ecosystem.

Takers - Capitalizing on Alpha via Lower Latency

Financial markets are supposed to incorporate all public information into asset prices. They generally do this. However, price discovery for most assets happens on one server in one location, while the information that impacts prices is produced all over the world.

The TradFi market microstructure is designed around low-latency traders who want to be co-located with exchanges’ matching engines.

If you as a retail trader observe an event in Singapore that will impact the price of TSLA, you still have to send a message to New Jersey, which is next to the market maker. This is fundamentally unfair to takers, and unnecessarily gracious to makers.

The first principles correct view of this problem is that the observer of this information should be able to place an order based on this new information with a validator not in New Jersey, but in Singapore. This market participant should earn this alpha for observing this information first, and getting an order to the global order book the fastest.

Today, Solana, like other leading chains, has a single leader at any moment in time. But this is changing soon, as Solana is moving towards Multiple Concurrent Leaders (MCL, see footnote).

Under MCL, there won’t just be two leaders at any moment in time, but dozens. With MCL, participants who observe information in the physical world can and will incorporate that information into asset pricing faster.

The key to optimizing price discovery is not lowering latency on a single matching engine by an incremental nanosecond. Instead it’s to empower people around the world with information to update prices by pushing price discovery to the edge.

Counterintuitively, decentralization enables takers to minimize latency for time-to-transaction inclusion, and therefore maximize information propagation in financial markets.

Decentralized price discovery is by definition better than centralized price discovery. The world is a big, heterogenous place.

Expanding TAM Horizontally…

Most of the major exchanges around the world, from the London Stock Exchange to the Chicago Mercantile Exchange to the Tokyo Stock Exchange, trade one type of asset (e.g. equities or commodities). But blockchains have laid bare the reality that all units of value—currencies, commodities, equities, derivatives positions, debt obligations, memecoins, governance tokens, utility tokens, NFTs, and more—can be represented as standardized tokens on a permissionless blockchain.

Most of the assets that trade on blockchains today are blockchain-native assets. That is, they are natively created and issued on-chain. This includes DeFi tokens, DePIN tokens, NFTs, and more. But there is a growing cohort of assets being issued on-chain that represent TradFi assets, including US equities, bonds, real estate, US treasuries, mezzanine debt, and more.

Virtually all assets will trade on inherently global and permissionless systems like Solana, eventually. That doesn’t necessarily mean that people stop trading on NYSE, NASDAQ, and CME, but rather that an increasing percentage of trading volumes will take place on-chain instead of on TradFi venues. This is only natural because blockchains are inherently global, permissionless, 24/7, far easier for retail traders to access, and far easier for developers to integrate than TradFi.

It is trivial to integrate a private key and a token into any app, whether that app is a telegram bot or light weight Android app or WeChat mini-app. It is exponentially more difficult to interface with the plethora of heterogeneous systems that represent TradFi systems around the world. Their APIs are far more complex, settlement times are slow and non-uniform, and in many cases TradFi institutions simply will not face retail traders at all.

By virtue of being public and permissionless, blockchains explicitly increase participation in financial markets of all forms. Ultimately, asset issuers don’t care about the rails their assets trade on. Asset issuers just want to ensure that anyone who wants to buy their asset can do so. Today, most CEOs don’t believe that natively issuing stock on-chain is increasing their universe of potential stockholders, but that will change in the next few years as global crypto ownership grows from an estimated 500M to a few billion.

Not only do we believe that crypto will support all TradFi assets, we also expect it to support many kinds of new assets that simply couldn’t exist before. One of my favorite examples is Parcl, which offers perpetual contracts for the average price per square foot for real estate transactions that have closed in a given market on a rolling 30-day basis. Parcl allows you to literally long Austin and short San Francisco—and use the equity value of each position to collateralize the other!

There are even teams building products that issue NFTs that represent individual bottles of whiskey, wine, and watches on-chain!

The TAM for Solana is expanding in all directions. Wall Street is slowly moving on-chain, and developers are building all kinds of new financial markets on-chain.

…And Capturing Value From Innovation Up The Stack

So far, everything in this post has discussed Solana as just a matching engine. But with DeFi protocols such as Drift, Jupiter, Kamino, marginfi, and many others, the Solana ecosystem can offer:

1. every financial service imaginable
2. to everyone around the world
3. with more transparency and auditability that meaningfully reduce the risk of contagion cascades
4. with higher capital efficiency than TradFi.

Today, the largest DeFi primitives on Solana are 1) spot trading, 2) borrow/lend, and 3) trading perpetual futures. These correspond approximately to 1) NYSE/NASDAQ, 2) the big banks that offer consumer and prime lending services along with FCMs, and 3) CME. And those are just for the USA. Solana competes to offer financial services for everyone in the world.

Although many Solana proponents including Anatoly (Cofounder and CEO of Solana Labs) and myself have called Solana decentralized NASDAQ, the TAM for Solana and its ecosystem is vastly larger than NASDAQ’s. Solana is trying to power all financial services around the world; it is far more than a matching engine.

The incredible thing about Solana is that all of these different financial instruments can natively and atomically compose with one another without the explicit blessing or support of the app developers. This concept of using existing smart contracts as legos to build more useful services is what most in the industry call composability. This enables more rapid experimentation and growth as developers can build on a base set of contracts, integrations, and liquidity, all of which compound to create value for the stakeholders in the Solana ecosystem in a virtuous cycle. This means that Solana-based products can innovate more quickly and provide better consumer experiences.

Solana itself is not providing financial services. But Solana creates the stack that powers hundreds, soon to be thousands, of financial services that facilitate trillions of dollars of risk transfer each year. And while gas costs are near 0 and trending downwards, Solana directly profits from the growth of these financial services via maximum extractable value (MEV).

As my partner Tushar discussed in his presentations at the 2022 and 2024 Multicoin Summits, asset ledgers like Solana can be valued as a function of the MEV they capture. Each new financial service produces incremental MEV, some of which Solana can capture. Today individual applications are already producing more than $100M of MEV for Solana in addition to their own application-specific revenues, and everything here is still so nascent.

In Q4 2024, the Solana network earned over $800M in REV (this does not include SOL inflation), which is roughly $3.2B on an annualized basis. This is up from ~$0 a year prior. And this is despite the fact that there are almost no TradFi assets issued on Solana, and the relative immaturity of the major DeFi protocols on Solana, most of which are just a couple of years old.

The TAM for Solana is growing across three dimensions:

1. DeFi protocols continue to mature, adding new features and functions and creating more MEV opportunities.
2. Entrepreneurs are building novel financial markets natively on-chain (e.g. compute and telecom and energy markets, BECMs, and more).
3. More assets are being issued on-chain, from memecoins to US equities.

Each of these not only grow the TAM for Solana, they strengthen each other. For example, the more assets that are issued, the more collateral is available that can be borrowed against.

Solana is compounding at an increasingly faster rate.

Internet Capital Markets

The Solana ecosystem is pushing forward on all fronts to fulfill the vision of Internet Capital Markets. Solana is simultaneously improving execution for makers via conditional liquidity and takers via Multiple Concurrent Leaders. Furthermore, the Solana ecosystem is growing its TAM horizontally—by supporting a broader array of TradFi and crypto-native assets—and vertically—by capturing some MEV from each of the many financial services built on top of Solana.

There is an incredible opportunity to create a global and permissionless financial system that:

1. allows people with information advantages to capture alpha across every asset class
2. while crossing the tightest spreads
3. and with the lowest fees
4. with globally-sourced leverage that is transparent and auditable in real-time
5. with maximum capital efficiency via atomic composability across positions and protocols.

That is the vision for Internet Capital Markets. That is the vision for Solana.

Thanks to Tarun Chitra, Justin Saslaw, Jon Charbonneau, Max Resnick, Nitesh Nath, and Mert Mumtaz for providing feedback on this post.

Footnote — A quick explainer on how MCL works without getting too technical. When Anatoly was thinking about how to architect Solana for speed in 2017, he realized that all of the large permissioned distributed systems such as Google’s Spanner database are predicated on a trusted source of time. That trusted source of time can be used to resolve conflicts. So if two conflicting transactions arrive to servers in NYC and Singapore 1 microsecond apart, the system will execute the 1st transaction and ignore the 2nd based on the trusted timestamp associated with each transaction. The problem in all blockchains prior to Solana was that there was no trusted source of time on an intra-block basis. Anatoly solved this problem using a verifiable delay function (VDF), which is known as proof of history (POH) in Solana. Multicoin wrote about this insight in our first Solana essay, The Separation of Time and State. POH provides for a way to provably increment time intra-block, and without forcing the global network of validators to come to consensus. MCL uses POH to resolve conflicts across leaders. If a malicious actor tries to double-spend across validators, each validator’s POH acts as a clock. The POH nonces can be compared across nodes, and the transaction that has the lower nonce will be accepted, and the transaction with the higher nonce will be discarded. If by some chance the two conflicting transactions end up with the same POH nonce across validators, the system will need a rule to decide which transaction should be included in the canonical asset ledger, and which should be discarded. There are a number of ways to accomplish this; a simple way to do this is to compare the hashes of the conflicting transactions themselves, and to include the transaction with the lowest hash value, and discard the transaction with the higher hash value.