Figure 1: Gas Fees and Ratio to Transaction Value.
Figure plots weekly Ethereum gas fees and the ratio of the gas fee to
transaction value for USDC transfers from 2021 through 2025. The gas fee
paid for a single transaction equals the gas price (the
per-unit cost of blockspace, denominated in ETH) multiplied by the
gas used (a unitless measure of the computational work the
transaction requires). The resulting fee is converted to U.S. dollars at
the prevailing ETH/USD exchange rate. Each series shows the 25th, 50th,
and 75th percentiles across transactions within a day, averaged to the
weekly frequency. The top-left panel plots the Ethereum gas fee in U.S.
dollars. The top-right panel plots the gas-fee-to-transaction-value
ratio for all USDC transfers on Ethereum. The bottom-left panel
restricts to USDC transfers with values below the daily median
transaction size. The bottom-right panel restricts to USDC transfers
with values above the daily median transaction size.
Figure Type: Four-panel time series chart Four-panel figure showing weekly Ethereum gas fees and gas-fee-to-transaction-value ratios for USDC transfers from 2021 through 2025. Each panel displays the 25th, 50th, and 75th percentiles. Top-left panel: Ethereum gas fees in U.S. dollars. Gas fees show high volatility with peaks exceeding $50 in 2021-2022, followed by a declining trend. By 2025, median gas fees have fallen to approximately $3-5. Top-right panel: Gas-fee-to-transaction-value ratio for all USDC transfers on Ethereum, shown as a percentage. The ratio exhibits high volatility, with the 75th percentile reaching above 20 percent during 2021-2022 congestion periods, declining to around 5 percent by 2025. Bottom-left panel: Gas-fee-to-transaction-value ratio for USDC transfers below the daily median transaction size. Small transactions bear disproportionately high costs, with the 75th percentile exceeding 40 percent during peak congestion periods in 2021-2022, declining to ap- proximately 10 percent by 2025. Bottom-right panel: Gas-fee-to-transaction-value ratio for USDC transfers above the daily median transaction size. Larger transactions show much lower ratios, with the 75th percentile staying below 10 percent throughout the sample period and declining to under 2 percent by 2025.
Figure 2: Gas Fees vs. Number of Transactions.
Figure plots a binned scatter of Ethereum gas fees against the number
of daily transactions. The left panel shows the gas fees for all
Ethereum transactions; the right panel shows gas fees for USDC
transactions.
Figure Type: Two-panel binned scatter plotTwo-panel binned scatter plot showing the relationship between daily transaction volume and average gas fees on Ethereum. Left panel: Gas fees for all Ethereum transactions. The x-axis shows the number of daily transactions in millions, ranging from approximately 0.5 to 2.0 million. The y-axis shows average gas fees in U.S. dollars, ranging from $0 to $60. The relationship is positive and convex: gas fees increase with transaction volume, and the rate of increase accelerates at higher volumes. At low transaction volumes (0.5-1.0 million transactions), gas fees remain below $10. As volume approaches 1.5 million transactions, gas fees rise steeply to $20-40. At the highest volumes near 2.0 million transactions, gas fees can exceed $50. Right panel: Gas fees for USDC transactions specifically. The pattern mirrors the left panel, showing the same positive and convex relationship. This indicates that USDC users face con- gestion costs driven by network-wide activity, not just stablecoin-specific demand. The convexity demonstrates that congestion costs accelerate non-linearly as the network approaches capacity.
Figure 3: Possible Payoff differential \(\pi(\gamma,\lambda)\) as a function of
\(\lambda\) for \(\gamma\in
(\underline{\gamma},\overline{\gamma})\).
Figure Type: Three-panel conceptual diagram (TikZ) Three-panel theoretical diagram illustrating the payoff differential function π(γ, λ) as a func- tion of the network externality parameter λ for intermediate congestion levels γ ∈ (γ, γ). The x-axis in each panel represents λ, the measure of network externalities (benefits from using the stablecoin). Two critical points are marked: λˆ (the threshold where users are indifferent between holding and redeeming) and m (the total mass of users). The y-axis represents π(γ, λ), the payoff differential between holding and redeeming the sta- blecoin. The function exhibits a downward-sloping segment (higher network externalities reduce redemption incentives) and features a discrete upward jump at λˆ of magnitude n(v_(h) − v_(l)), rep- resenting the coordination benefit when crossing the network externality threshold. Case I (left panel): The payoff function is positive at λ = λˆ and remains positive through λ = m. In this case, even low-network-externality users prefer to hold the stablecoin, resulting in no equilibrium with redemptions. ˆ Case II (middle panel): The payoff function crosses zero between λ and m. The function starts positive at λˆ but becomes negative before reaching m. This creates multiple equilibria: one where all users hold (if they coordinate on high network externalities) and one where users with λ above the zero-crossing hold while others redeem. ˆ Case III (right panel): The payoff function is negative at λ = λ and remains negative. Even at the threshold λˆ, the payoff differential is negative, implying that redemption dominates. This case features a unique equilibrium with widespread redemptions.
Figure 4: Redemptions under strategic complementarities in our model
and pre-emptive redemptions in Bernardo and Welch, 2004.
Figure Type: Two-panel conceptual diagram (TikZ) Two-panel conceptual diagram contrasting predicted redemption patterns under strategic complementarities (our model) versus strategic substitutabilities (pre-emptive redemptions as in Bernardo and Welch, 2004). Both panels have congestion on the x-axis (measured by the parameter γ) and redemptions on the y-axis. A critical congestion threshold γ^(∗) is marked on the x-axis. Left panel: Strategic Complementarities. The redemption function exhibits a discrete jump at the threshold γ^(∗). For congestion below γ^(∗), redemptions are zero. At γ^(∗), redemptions jump discontinuously to a high positive level. For congestion above γ^(∗), redemptions remain at the elevated level. This creates a step function: flat at zero, then a vertical jump, then flat at a high redemption rate. Right panel: Strategic Substitutabilities (pre-emptive redemptions). The redemption function is smooth and continuous, increasing linearly from the origin. As congestion rises, redemptions increase gradually and proportionally. There is no discrete threshold or jump.
Figure 5: Empty Slots Rate on Ethereum. Figure plots
the daily share of empty slots on the Ethereum blockchain since the
Merge in September 2022.
Figure Type: Single-panel time series Time series plot showing the daily share of empty slots on the Ethereum blockchain from September 2022 (the Merge) through December 2025. The x-axis spans from September 2022 to December 2025. The y-axis shows the empty slots rate as a percentage, ranging from 0 to 10 percent. An empty slot occurs when a validator fails to propose a block in their assigned slot. The time series shows that the empty slots rate typically remains below 2 percent. The series exhibits occasional spikes, with several episodes reaching 4-6 percent and one extreme event exceeding 9 percent. These spikes are rare and transient, typically lasting only a day or two before returning to the baseline near-zero rate. The distribution is heavily right-skewed: the median rate is 0.6 percent, but the maximum reaches 9.6 percent. The average rate over the full sample is 0.7 percent.
Figure 6: USDT Transfers from Ethereum to Tron.
Figure plots the daily net transfer amount of USDT from Ethereum to
Tron, in millions of U.S. dollars. Transfers are identified by matching
USDT transfer events on Ethereum to USDT transfer events of the same
dollar amount on Tron occurring within 60 minutes of each other.
Positive values indicate net flows from Ethereum to Tron; negative
values indicate net flows in the opposite direction.
Figure Type: Single-panel time series Time series plot showing the daily net transfer amount of USDT from Ethereum to Tron, measured in millions of U.S. dollars, from January 2023 through December 2025. The x-axis spans from January 2023 to December 2025. The y-axis shows net transfer amounts in millions of dollars, ranging from approximately -$200 million to +$600 million. Positive values indicate net flows from Ethereum to Tron; negative values indicate net flows from Tron to Ethereum. Transfers are identified by matching USDT transfer events on Ethereum to USDT transfer events of the same dollar amount on Tron occurring within 60 minutes of each other. The time series shows high volatility with frequent spikes in both directions. The series oscillates around zero, but positive spikes (Ethereum to Tron) tend to be larger in magnitude than negative spikes (Tron to Ethereum), particularly in 2023 and 2024. Several episodes show net transfers from Ethereum to Tron exceeding $400 million in a single day, with the largest spike approaching $600 million. Negative transfers (Tron to Ethereum) are less extreme, rarely exceeding -$150 million. Over the sample period, there appears to be a gradual decline in both the frequency and magnitude of large transfer spikes, particularly in 2025. The series becomes somewhat less volatile in the latter portion of the sample, though large one-day transfers still occur occasionally.
Figure 7: USDT Transfers from Ethereum to Tron vs. Gas Fees.
Figure plots a binned scatter of the daily net USDT transfer amount
from Ethereum to Tron (in millions) against the demeaned Ethereum gas
fee, where the gas fee is demeaned by its 365-day trailing moving
average. The positive relationship indicates that days following
above-average congestion on Ethereum tend to coincide with larger net
transfers of USDT from Ethereum to Tron.
Figure Type: Binned scatter plot Binned scatter plot showing the relationship between demeaned Ethereum gas fees and net daily USDT transfer amounts from Ethereum to Tron. The x-axis shows the Ethereum gas fee demeaned by its 365-day trailing moving average, measured in U.S. dollars. The range is approximately -$4 to +$4, representing deviations from the long-run average gas fee. The y-axis shows net transfer amounts from Ethereum to Tron in millions of dollars, ranging from approximately -$50 million to +$150 million. The scatter plot shows a clear positive relationship: days with above-average Ethereum gas fees (positive x-axis values) tend to coincide with larger net transfers from Ethereum to Tron (positive y-axis values). When gas fees are $2-4 above their moving average, net transfers from Ethereum to Tron average around $100-150 million. Conversely, when gas fees are below average (negative x-axis values), net transfers are smaller or even negative, indicating flows back to Ethereum. The relationship appears approximately linear across the range of gas fee deviations. The dispersion around the fitted line indicates substantial day-to-day variation, but the upward slope is pronounced and persistent.
Figure IA.1: Gas Fees vs. Number of Transactions: Before and After Pectra.
Figure plots binned scatters of Ethereum gas fees against daily
transaction counts with quadratic fitted lines. The left panel uses data
before the Pectra hard fork (May 7, 2025); the right panel uses data
after the Pectra hard fork. Both panels use the same axis scales. The
convex relationship visible in the left panel flattens after the
capacity expansion.
Figure Type: Two-panel binned scatter plot Two-panel binned scatter plot comparing the relationship between daily transaction volume and gas fees before and after the Pectra upgrade on Ethereum. Left panel: Pre-Pectra period (data through March 2025). The x-axis shows the number of daily transactions in millions, ranging from approximately 0.5 to 1.8 million. The y-axis shows gas fees in U.S. dollars, ranging from $0 to $50. The relationship is positive and convex: gas fees rise with transaction volume, and the rate of increase accelerates at higher volumes. At low transaction volumes (0.5-1.0 million), gas fees remain below $10. At moderate volumes (1.0-1.5 million), gas fees rise to $15-25. At high volumes approaching 1.8 million transactions, gas fees spike to $40-50. A fitted quadratic curve shows clear upward curvature, demonstrating convexity. Right panel: Post-Pectra period (data from April 2025 onward). The axes have the same scales as the left panel. The post-Pectra relationship shows a notably different pattern: the convexityis reduced. Gas fees remain low (under $10) across a much wider range of transaction volumes. Even at volumes approaching 1.5-1.8 million transactions, gas fees do not spike as sharply as in the pre-Pectra period. The fitted curve is flatter and shows less pronounced curvature.