A recent working paper from the Bank for International Settlements (BIS), published on September 15, 2026, challenges long-held assumptions about decentralized finance (DeFi) growth. The analysis revealed that while Tether’s USDT stablecoin saw significant issuance expansion on Ethereum through 2024, its proportion held in smart contract accounts actually stalled.
This trend suggests that a larger stablecoin supply doesn’t automatically translate to increased capital flowing into DeFi protocols.
International Settlements and Stablecoin Growth
On the Tron network, the share of USDT held in smart contract accounts consistently hovered around just 1% throughout the study’s historical series. These findings from the BIS provide a crucial recalibration for those assessing the true scale and nature of participation in the DeFi sector, especially concerning the dominant stablecoin.
The BIS study meticulously tracked where USDT tokens reside, offering a granular view beyond mere issuance figures. For Ethereum, the data shows that smart-contract accounts held more than 20% of the network’s USDT during parts of 2021 and 2022. This share then settled between 15% to 20% until late 2024.
Crucially, as USDT issuance continued to expand significantly through 2024, this proportion dropped to roughly 10% to 15%. This wasn’t a finding that the absolute balance of USDT in contracts necessarily fell, but rather that the torrent of newly issued tokens largely accumulated outside these smart contracts, diluting their overall share.
The Bank for International Settlements authors explicitly noted that the substantial growth in USDT issuance did not correspond with a sustained increase in contract holdings. This phenomenon suggests that much of the new stablecoin supply was absorbed by externally owned accounts, perhaps for purposes other than direct DeFi engagement, such as exchange custody or speculative holdings.
Dissecting Ethereum’s Contract Share
The distinction between a falling percentage share and a declining absolute balance is vital for understanding market dynamics. On Ethereum, the dollar value of contract-held USDT fluctuated in the low tens of billions, approximately $10 billion to $15 billion toward the end of the plotted period. Yet, even with these significant sums, the share dwindled.
This means that while the total volume of USDT on Ethereum grew, the portion actively locked into or managed by decentralized applications did not keep pace. Ethereum upgrades have aimed to enhance the network’s capacity and efficiency, but this data points to a more complex interaction between stablecoin supply and direct DeFi utility.
On Tron, the contract-held USDT remained a comparatively small slice, less than $1 billion, despite the network hosting a much larger total supply. These varying dynamics underscore that a one-size-fits-all approach to measuring DeFi adoption based solely on stablecoin issuance can be misleading.
New Metrics for Digital Asset Tracking
The BIS researchers reconstructed USDT holdings by meticulously analyzing transfer event logs from both Ethereum and Tron. Their methodology involved identifying smart-contract accounts through contract deployments and categorizing other addresses as externally owned accounts. This detailed approach allowed for cross-verification of token supply against mint, burn, and blacklist-destruction events.
This method offers a distinct advantage over traditional metrics like Total Value Locked (TVL), which sum up deposits reported by DeFi protocols. TVL can inadvertently count the same tokens multiple times across different applications or bridges, leading to inflated figures. The BIS reconstruction, conversely, tracks a single token’s journey across addresses on specific networks, including those outside explicit DeFi applications.
Such precision is crucial for asking more fundamental questions, like how USDT is truly distributed between various account types. While TVL can certainly describe the scale of selected venues, it doesn’t offer the same clarity on the fundamental allocation of capital. Understanding Ethereum’s influence on broader crypto asset behavior requires such detailed analysis.
Beyond Traditional DeFi Measures
The Bank for International Settlements approach shifts the focus from protocol-level reporting to a token-centric view. This provides a clearer, less ambiguous picture of stablecoin distribution. It’s a significant step toward developing more robust and accurate measurement frameworks for the burgeoning digital asset ecosystem, moving beyond superficial metrics.
This new perspective helps to peel back layers of complexity in how crypto and decentralized finance ecosystems are measured. It offers a more nuanced understanding, essential for both market participants and regulators, who seek reliable indicators of real economic activity rather than just speculative interest.
The Imperfect Guide of Account Types
While the BIS methodology represents a significant improvement in tracking token distribution, the nature of an account type remains an imperfect guide to its ultimate economic use.
A smart contract, for instance, might hold USDT not for direct DeFi lending or trading, but rather as a reserve for a bridge, a wrapper, or even a custodial service. These distinct uses have different implications for true DeFi demand.
Similarly, an externally owned address holding USDT could be doing so for a variety of reasons: facilitating payments, serving as personal savings, enabling cross-border remittances, or acting as exchange custody. The mere presence of USDT in such an account doesn’t automatically imply its deployment in active DeFi strategies or consumption through commercial transactions.
So, Tron’s consistently low 1% contract-held share doesn’t necessarily mean the remaining 99% was spent on payments. Nor does Ethereum’s falling share definitively prove a contraction in DeFi usage. Instead, these figures highlight the need for even deeper analysis into the intent behind token holdings to truly grasp the dynamics of the market.
The findings serve as a vital reminder that correlation does not equal causation. A burgeoning supply of stablecoins like USDT is often heralded as a sign of expanding DeFi adoption, but the BIS data suggests a more intricate reality. It forces market observers to look beyond simple growth figures and consider the underlying distribution mechanisms.
Current Market Context and Future Implications
Despite the historical data presented by the BIS, the current scale of USDT across the crypto landscape remains substantial. As of September 28, DefiLlama reported a total USDT market capitalization across all chains at approximately $183.7 billion. Ethereum alone accounted for about $73.3 billion of this, while Tron held an even larger share at roughly $92.5 billion.
These figures, however, are a later, third-party snapshot of total supply and don’t update the BIS breakdown of holder types. They cannot confirm whether today’s vast sums of USDT are predominantly held within identified DeFi contracts, centralized exchange wallets, or other types of accounts. This gap in real-time, granular data continues to present a challenge for accurate market assessment.
Moving forward, a precise understanding of true DeFi deployment would necessitate a continuously updated breakdown of balances. This would need to specifically identify funds within DeFi contracts on each chain, ideally separating those used for bridging or custody from those actively engaged in lending, borrowing, or trading protocols. Without this level of detail, assumptions about market growth remain speculative.
Ultimately, rising USDT totals, by themselves, are insufficient to establish either an increase in DeFi deployment or a surge in payment transactions. Furthermore, these figures offer no direct insight into the demand for native blockchain tokens such as ETH on Ethereum or TRX on Tron.
The BIS paper reinforces the need for more sophisticated analytical tools to truly interpret the evolving digital asset economy and its underlying drivers, offering vital context for those building on Ethereum’s financial rails and beyond.