This blog is a preview of our forthcoming report, “The New Rails: How Digital Assets Are Reshaping the Foundations of Finance.” Reserve your copy!
TL;DR
- No single blockchain is the right answer for every asset. The “best” chain depends entirely on what’s being tokenized: a money market fund has fundamentally different infrastructure needs than a high-frequency trading application.
- Low fees matter less than predictable fees. For financial institutions running daily operations, a network that occasionally spikes to hundreds of dollars per transaction (like Bitcoin during congestion) can pose more operational risk than one with slightly higher, but stable baseline costs.
- Speed can have two definitions, and institutions need to know which one they’re prioritizing. Raw throughput (how many transactions per second) and time to finality (when a transaction is truly irreversible) are not synonymous. For high-value settlement, only the latter counts.
- Custody concentration is a hidden systemic risk. When one exchange dominates the custody of assets on a given network, a shock to that entity — a hack, a run, an insolvency — can become a shock to the entire ecosystem.
- Compliance infrastructure should scale with illicit exposure. Every network carries some degree of illicit activity. The critical variable is whether robust monitoring tools exist to manage it. Tools like Chainalysis KYT and Address Screening are designed to give institutions the on-chain visibility to expand across networks without outpacing their compliance controls.
The tokenization of Real World Assets (RWAs) is no longer a speculative concept. It is a rapidly maturing sector of traditional finance (TradFi). But as banks, asset managers, and financial institutions transition from pilot programs to live deployments, they face a critical infrastructure dilemma: Where should we build?
The “perfect” general-purpose blockchain does not exist. Instead, building on-chain is an exercise in trade-offs across five axes: speed, cost volatility, contagion risk, illicit exposure, and governance. Using our data, we mapped the competitive landscape of nine major networks to show the tradeoffs among the networks without normative judgments. Financial institutions cannot rely on name brands alone. The underlying network architecture must suit the specific use cases of the asset being tokenized. On-chain data can act as a guide.
Here is what the on-chain data tells us about the current tokenization infrastructure landscape.
The three architectural archetypes
Before diving into individual metrics, it helps to look at these networks’ macro footprints. When plotting these blockchains on a multi-axis radar chart—where stretching toward the outer edge represents the superior business outcome—three distinct archetypes emerge:
- The ‘Lindy’ Institutional Anchors (Bitcoin & Ethereum): These networks are crypto’s oldest, with the longest proven track record. They heavily anchor the structural security and liquidity axes, but largely abandon the cost and throughput axes. They are not designed to execute high-volume, low-margin TradFi operations directly. They are the ultimate settlement layers, with the largest implicit strength being their tried and true nature. This is already playing out in practice. JPMorgan’s Onyx platform, for instance, uses a private Ethereum fork for institutional repo and intraday settlement, leaning on Ethereum’s architecture and familiarity while maintaining permissioned control.
- The “Goldilocks” Scaling Swarm (Arbitrum, Base, and Polygon): These Layer-2 (L2) networks represent the most balanced “sweet spot” (hence ‘Goldilocks’) for general-purpose TradFi development, maximizing performance and cost efficiency with low current volumes of illicit flows.
- The High-Frequency Engine (Solana, BNB, XRP Ledger, and TRON): Pushing the absolute outer boundaries for extreme throughput and negligible fees, Solana in particular is highly specialized. However, it pulls sharply inward on market concentration and historical fee volatility, reflecting the realities of its monolithic architecture.
With that macro perspective in mind, we now examine each axis in detail, starting with the one most directly affecting P&L: transaction costs.
Operational costs: predictability vs. tail risk
For financial institutions, low fees are important, but predictable fees are critical. To measure this, we analyzed daily average network fees and calculated their kurtosis — a statistical measure of “tail risk.” A high kurtosis score means a network is highly susceptible to violent, unpredictable pricing shocks during periods of congestion. From this analysis, several dynamics emerge.
The above chart is a density distribution (ridge plot) shown on a logarithmic scale, tracking daily average network fees since April 2024. The “peaks” or “mountains” show where a network’s daily fee sits on a normal day. The long tails stretching to the right represent days when network congestion caused fees to spike. The “kurtosis” score quantifies this tail risk: a higher number means the network is highly susceptible to violent, unpredictable pricing shocks (tail risk fee events).
On the one hand, with a kurtosis of 0, TRON is the most predictable. Its tightly clumped distribution makes its operational costs highly predictable, explaining its dominance as a reliable rail for global stablecoin payments. Similarly, thanks to the Dencun upgrade offloading data to L2s, Ethereum Mainnet is currently experiencing highly stable daily average fees without the extreme gas spikes of previous cycles.
In contrast, Bitcoin carries extreme tail risk, with a kurtosis of 246, likely related to fee expansion during the runes and ordinal mania. Runes and ordinals are protocols allowing users to inscribe data (e.g., images, text, and tokens) directly onto the Bitcoin blockchain, and have triggered periodic surges in network demand that have little to do with financial settlement activity. While baseline fees are manageable, network congestion can cause fees to spike violently into the dozens or hundreds of dollars, posing massive unpredictability for daily operations at scale. While L2 networks are broadly cheap, their volatility profiles differ. Base and Optimism showcase relatively tight predictability, making them ideal for retail-facing tokenization. Conversely, Arbitrum still occasionally experiences proportional fee spikes. The next operational constraint is throughput, because a cheap network that can’t clear volume at scale is still a bottleneck.
The need for speed: throughput vs. finality
When evaluating network speed, TradFi must distinguish between how many transactions a network can handle (throughput) and how fast those transactions actually settle (time to finality).
Solana is the undisputed leader in raw throughput, measured in transactions per second (TPS). It processes over twice as much on-chain volume as the second-highest chain (TRON). Its growing market share over the last year indicates an unparalleled ability to absorb massive transaction bursts, making it the natural home for high-frequency trading applications.
But throughput is meaningless if settlement lags. Arbitrum consistently holds the top spot for fastest block completion (time to finality), followed closely by BNB (bolstered by its recent Fermi upgrade) and TRON. Franklin Templeton expanded its OnChain US Government Money Fund (FOBXX) to Solana in February 2025, a decision rooted in the network’s throughput profile. The firm had flagged its rationale months earlier, stating that “Solana has shown major adoption and continues to mature, overcoming technological growing pains and highlighting the potential of high-throughput, monolithic architectures.”
One important caveat is that our charts measure ‘soft finality,’ which refers to when a transaction is accepted, ordered, and processed by a network node (often on an L2), but not yet finalized on the base layer. This could be appropriate for some forms of transactions and inappropriate for others, such as transactions in the million-dollar range and beyond. For RWAs, true finality is paramount. Institutions building on L2s like Arbitrum must account for “L1 processing risk.” That is to say, the additional time required for the L2 to communicate and finalize its transaction batch on the Ethereum Mainnet, which can take between 15 minutes and several hours, depending on proof generation. This resembles a regional clearing house that batches trades intraday, but only books final settlement with the central bank overnight: fast for most purposes, but not instant.
Contagion risk: institutional dependency
To assess systemic risk, we looked at centralized exchange (CEX) dependency. Within the crypto ecosystem, CEXs routinely provide one another with crypto assets in order to facilitate the proper, liquid function of markets. For example, CEX 1 might send 100 BTC to CEX 2 in order to allow CEX 2 to cover withdrawals and, in the future, CEX 2 might reciprocate and send crypto assets back to CEX 1. This dynamic makes the crypto ecosystem more robust and agile, because it ensures that prices do not seize up on particular exchanges and lowers the risk that an exchange will experience the on-chain equivalent of a bank run and need to freeze withdrawals.
However, the proper functioning of this system can generate fragility, similar to just in time logistics and shipping where minor disruptions can cause massive ripple effects. In particular, if you take large direct inter-exchange flows by asset as a measure of liquidity provision, then you can determine how crucial that inter-exchange liquidity is relative to the exchange’s on-chain liabilities (i.e., withdrawals by users). When that ratio gets too high, the system is building up systemic risk, because absent others, on-chain liabilities cannot be covered and price can collapse.
The chart above maps this measure over time for 6 networks. For institutional dependency (direct CEX to CEX transfers >$1M as a share of on-chain liabilities), Solana tells the most dramatic story. High interconnectivity across a token’s ecosystem can be a source of strength, but excessively high dependence on other exchanges can also bring with it a risk of contagion. For example, if CEX 1 depends on CEX 2 for all or almost all of its liquidity to cover withdrawals, problems could quickly emerge if CEX 2 experiences a run or sudden surge of withdrawal demand. This dynamic is illustrated by the ramp-up of Solana before the FTX debacle, wherein excessive dependency between CEXs primed the pump for a collapse. The data suggest that tokens on some newer blockchains are prone to excessive degrees of dependency (e.g. >50% of liabilities), while older networks like Bitcoin or Ethereum tend to run below this majority threshold, suggesting a lower risk of contagion. Solana’s exchange outflows as a percentage of interexchange dependences have risen in 2025-2026, a trend we will continue to monitor.
The compliance mandate: liquidity vs. illicit exposure
For any regulated entity, compliance is an essential function and can protect revenue streams by mitigating and eliminating potentially high-risk and illicit exposure. We plotted total network liquidity against the percentage of that volume coming from known illicit entities to map networks to quadrants.
As depicted in the above chart, which measures liquidity against illicit exposure as a share of total activity, the “green zone” is clear: Ethereum, Solana, and Base sit perfectly in the TradFi sweet spot (high-volume and low illicit exposure, in the bottom-right quadrant). These networks provide the multi-trillion-dollar liquidity required by institutions while maintaining lower illicit share (below 1%). BlackRock’s BUIDL fund, the largest tokenized money market fund by AUM, was first deployed on Ethereum. This choice reflects this calculus: deep liquidity with a low illicit footprint.
BUIDL’s multi-chain expansion — now spanning Ethereum, Solana, Polygon, Arbitrum, Avalanche, and beyond — reflects a deliberate strategy to maximize investor optionality and ecosystem reach. As BlackRock’s CEO Larry Fink and COO Rob Goldstein wrote in The Economist, tokenization represents “the next major evolution in market infrastructure,” one designed to enable instantaneous settlement and expand access to investable assets across digital ecosystems. Fink has been explicit about this ambition: “We need to be tokenizing all assets, especially assets that have multiple levels of intermediaries,” a vision that demands presence across blockchains, not commitment to just one.
TRON occupies a unique position. It possesses massive institutional-scale liquidity, serving as the dominant global settlement layer for OTC desks moving popular stablecoins, but the proportion of that service inflow volume from illicit sources approaches 4%. This is not a reason to categorically avoid a network, but it does underscore the critical role played by compliance.
To be sure, every blockchain carries some degree of illicit exposure, and usually as overall volumes increase, so does illicit usage. The more consequential decision for regulated institutions is whether robust, reliable compliance tooling exists to monitor, identify, and mitigate it in real time. Solutions like Chainalysis KYT (Know Your Transaction) and Address Screening are designed precisely for this purpose, giving TradFi institutions the on-chain visibility necessary to expand across and between networks confidently while keeping pace with compliance controls.
Governance: who’s in charge during a crisis?
While charts perfectly plot quantitative on-chain realities, they cannot measure qualitative TradFi requirements. The final trade-off is governance. For financial institutions, a network’s governance structure dictates two critical realities: its agility in executing essential technical upgrades—such as migrating to quantum-resistant cryptography—and its protocol for handling severe adverse events.
Data alone cannot quantify the value of decentralization structures. From a TradFi perspective, institutions must weigh the immutability of Bitcoin—where centralized intervention is impossible, acting as a double-edged sword—against the governance structures of Proof of Stake (PoS) ecosystems. For example, the existence of an oversight authority that can intervene in case of a catastrophe, such as the Arbitrum Security Council or DAO structures that can theoretically pause or reverse transactions during a hack. Institutions must decide which governance model aligns with their internal risk and legal frameworks.
Aligning assets to architecture
The data are clear: no single network dominates low illicit exposure, high speed, low fees, and perfect decentralization. Tokenization is the land of trade-offs. Take tokenized bonds: settlement finality and regulatory auditability make Ethereum the natural fit, as Société Générale’s Forge platform illustrates, having issued structured products and digital bonds directly on Ethereum Mainnet. A high-frequency trading application, by contrast, optimizes for throughput and near-instant finality, pointing toward an L2 like Arbitrum, which consistently performs well in time-to-finality. The goal for TradFi is not to find the “best” single blockchain or combination of blockchains, but rather to optimize for the specific requirements of the assets being deployed.
By analyzing the on-chain data alongside their own requirements and preferences, financial institutions can bypass the hype and build infrastructure that aligns with their long-term strategic goals. More than anything, these five dimensions — cost stability, throughput, contagion risk, illicit exposure, and governance — form a decision matrix, not a ranking. The right architecture depends entirely on the asset. As institutions determine where to build for tokenized assets, the payment rails running beneath them are transforming at equal scale, with stablecoin volume projected to reach $1.5 quadrillion by 2035, a figure that reframes the infrastructure decision from purely technical to existential.
FAQs
What is RWA tokenization and why does blockchain infrastructure matter?
Real World Asset (RWA) tokenization is the process of representing ownership of traditional financial assets (e.g., bonds, money market funds, real estate, private credit) as digital tokens on a blockchain. The choice of blockchain infrastructure matters, because different networks make fundamentally different trade-offs across speed, cost stability, regulatory risk, and governance. Picking the wrong chain for a given asset class can have direct implications for settlement risk, compliance costs, and counterparty exposure.
Which blockchain is best for institutional tokenization?
There is no single answer. Ethereum and its Layer-2 (L2) networks like Arbitrum and Base currently offer the strongest combination of institutional liquidity, low illicit exposure, and compliance tooling, making them a natural home for regulated assets like tokenized securities and money market funds. High-frequency or cost-sensitive applications may be better suited to Solana’s throughput. The right choice is always asset-specific, not chain-specific.
What is the difference between soft finality and hard finality on a blockchain?
Soft finality is the near-instant confirmation that a transaction has been accepted and ordered by the network, fast enough for most routine activity. Hard finality means the transaction is irreversible and fully settled on the underlying chain, which for Ethereum Layer-2 networks like Arbitrum can take anywhere from 12–24 hours to up to one week for withdrawals. For institutions processing high-value settlements, understanding which type of finality a network provides is critical.
How do financial institutions manage compliance risk when building on public blockchains?
The starting point is understanding the illicit exposure profile of the network being used. From there, institutions need real-time transaction monitoring and wallet screening tools that can flag high-risk counterparties before exposure accumulates. Chainalysis KYT and Address Screening are purpose-built for this use case, giving compliance teams the on-chain visibility to operate across multiple networks without assuming disproportionate regulatory risk.
What is the tokenization market expected to be worth?
When the growth of stablecoin payment rails is factored in alongside RWA tokenization, the broader on-chain financial economy is projected to reach into the hundreds of trillions over the next decade, making infrastructure decisions made today consequential for years to come.
This website contains links to third-party sites that are not under the control of Chainalysis, Inc. or its affiliates (collectively “Chainalysis”). Access to such information does not imply association with, endorsement of, approval of, or recommendation by Chainalysis of the site or its operators, and Chainalysis is not responsible for the products, services, or other content hosted therein.
This material is for informational purposes only, and is not intended to provide legal, tax, financial, or investment advice. Recipients should consult their own advisors before making these types of decisions. Chainalysis has no responsibility or liability for any decision made or any other acts or omissions in connection with Recipient’s use of this material.
Chainalysis does not guarantee or warrant the accuracy, completeness, timeliness, suitability or validity of the information in this report and will not be responsible for any claim attributable to errors, omissions, or other inaccuracies of any part of such material.
