During the Risk Committee election process, OAK Research argued that activating ENA buybacks today would be a mistake. We believe that criticism is not enough and while we were not selected for the fourth term committee, we worked on the modeling and analysis if multiple buyback scenarios for Ethena in a case where the fee switch would be turned on.
By posting this topic on the governance forum, we seek to get feedback from community members as well as the Risk Committee in charge of activating the fee switch for Ethena.
Context and Objective of the Analysis
Ethena’s Fee Switch Mechanism
Ethena generates revenue through its USDe stablecoin, whose staked version (sUSDe) distributes yield to holders. This yield comes from three sources: funding rates from delta-neutral positions on CEXs, staking rewards from underlying collateral, and minting fees.
In November 2024, a governance proposal was discussed and accepted by the Ethena community regarding the activation of a “fee switch.” The idea is to capture a portion of sUSDe’s yield to fund ENA token buybacks in the market, and potentially redistribute them to sENA holders.
Following this initial vote, Risk Committee members established several success criteria before triggering the fee switch:
- USDe circulating supply: >$6B
- Protocol cumulative revenue: >$250M
- CEX adoption: USDe integrated on 4 of the 5 largest exchanges by derivatives volume
- sUSDe APY superior by 5 to 7.5% compared to the competitive benchmark (Sky’s sUSDS)
On September 15, 2025, the Ethena Foundation announced that the protocol now met the criteria to activate the fee switch. This mechanism has not yet been implemented and was one of the determining factors in governance discussions during the recent election of new Risk Committee members.
It’s important to understand that activating a fee switch on Ethena is not trivial. The revenue generated by USDe belongs to sUSDe holders, and any protocol capture would mechanically reduce yield and therefore the stablecoin’s competitiveness.
This requirement introduces a major constraint that makes buybacks unpredictable: when protocol revenues decline or sUSDS yield increases, buybacks can stop abruptly. An intermittent buyback may be more harmful to ENA than no buyback at all.
The objective of this analysis is to model and study different fee switch activation scenarios to determine which might be most suitable for Ethena while maintaining sUSDe’s competitiveness alongside a consistent buyback program.
Study Assumptions
- Period: September 1, 2024 to February 13, 2026 (531 days), excluding Q2 2024 (launch phase with abnormally high APYs, notably due to very low TVL)
- Two distinct market regimes will be studied during this period: the bull phase (63 days between November 2024 and January 2025) and the normal phase (the remaining 468 days)
- Competitiveness benchmark: sUSDS (Sky), the primary reference in the yield-bearing stablecoin market
- Competitiveness criterion: sUSDe APY post-buyback ≥ 1.075× sUSDS APY (i.e., USDe must offer at minimum 7.5% additional yield compared to the benchmark after capture)
- ENA price used: actual daily historical data
Reference Data for the Period
The data covers 531 days (September 1, 2024 to February 13, 2026). During this period, sUSDe’s average APY was 8.2% while sUSDS’s average APY was 6.39%.
We identified two distinct market regimes: the bull phase (63 days from November 5, 2024 to January 6, 2025) and the normal phase (the remaining 468 days).
- The bull phase corresponds to Donald Trump’s election, which generated a bullish rally across most cryptos. sUSDe’s yield reached an average of 19.7%, significantly outperforming sUSDS (10.8%) despite incentives implemented to encourage migration from sDAI.
- The normal phase, corresponding to current conditions (from January 2025 to today), shows a very compressed APY delta vs sUSDS (+0.2 to 0.5pp), making the competitiveness criterion difficult to maintain if a fee switch is activated.
The competitiveness criterion (sUSDe APY > 1.075x sUSDS APY) was active for 305 days, or 57% of the study period. Total yield generated by USDe was approximately $368 million, and ENA’s average price was $0.457 within a range of $0.118 to $1.305.
Scenario Presentation
Preamble
We modeled six fee switch activation scenarios, ranging from the simplest approach (fixed capture) to more sophisticated mechanisms (progressive brackets and counter-cyclical reserve). All scenarios respect a fundamental criterion: buyback only activates when the competitiveness ratio allows maintaining sUSDe’s attractiveness.
Before detailing each scenario, it’s essential to define several concepts that will help you understand:
- Gross sUSDe APY: The actual sUSDe yield without any buyback capture
- Post-buyback sUSDe APY: The effective sUSDe yield after applying the buyback mechanism
- Gross ratio: Gross sUSDe APY / sUSDS APY, measuring initial outperformance versus the benchmark
- Post-buyback ratio: Post-buyback sUSDe APY / sUSDS APY, must remain ≥ 1.075× to maintain the competitiveness criterion (+7.5%)
- Yield surplus: The yield difference between gross sUSDe APY and the minimum threshold required for competitiveness (1.075× sUSDS APY)
Beyond daily compliance with the competitiveness criterion (post-buyback ratio ≥ 1.075×), the scenario must also ensure that the annual average APY respects this same ratio over the entire period. Based on analyzed historical data (average sUSDS APY = 6.39%), this imposes a minimum average sUSDe APY of 6.87% for the complete year.
Scenario 1 (S1) — 25% Fixed Capture Under Condition
Logic: S1 applies a simple 25% capture of total yield, but only when the competitiveness margin is very comfortable. The activation criterion is deliberately conservative to ensure that after capture, sUSDe maintains the competitiveness criterion.
For post-buyback APY to exceed sUSDS by 7.5% after a 25% capture, the gross ratio must be at least 1.075/0.75, or 1.433×.
Operation:
- Inactive if post-buyback ratio < 1.075× (equivalent to gross APY < 1.433× sUSDS APY)
- Active if gross ratio ≥ 1.433×
- Capture: 25% of daily total yield allocated to buyback
- Result: Post-buyback APY = 75% × gross APY, guaranteeing a post-buyback ratio ≥ 1.075×
Concrete example: If sUSDS shows 5% and sUSDe generates 8% (gross ratio = 1.6×), the buyback captures 2% (25% of 8%), leaving 6% for sUSDe holders (post-buyback ratio = 1.2×, above the 1.075× criterion).
Philosophy: “All or nothing” approach prioritizing simplicity and maximum yield preservation, at the cost of high buyback intermittency.
Scenario 2 — 50% of Surplus Beyond Criterion
Logic: S2 adopts a more nuanced approach by capturing only part of the available surplus beyond the competitiveness threshold. This allows buyback activation more frequently (as soon as the gross ratio exceeds 1.075×) while leaving a comfortable margin for sUSDe holders.
Operation:
- Inactive if gross ratio < 1.075×
- Active if gross ratio ≥ 1.075×
- Surplus calculation: Gross yield − (1.075× sUSDS APY)
- Capture: 50% of this surplus for buyback, 50% remains with holders
Concrete example: If sUSDS shows 5% and sUSDe generates 7%, the threshold is met (1.075× 5% = 5.375%) and available surplus is 1.625% (7% − 5.375%). This allows 0.8125% (50% × 1.625%) for buyback and post-buyback APY of 6.1875% (7% − 0.8125%).
Philosophy: Balanced sharing of surplus between yield preservation and buyback generation. More frequent activation than S1, but moderate captures.
Scenario 3 — 100% of Surplus (Maximalist Approach)
Logic: S3 maximizes buyback volume by capturing the entire surplus beyond the competitiveness threshold. sUSDe’s yield is systematically fixed at exactly 1.075× sUSDS APY when buyback is active.
Operation:
- Inactive if gross ratio < 1.075×
- Active if gross ratio ≥ 1.075×
- Capture: 100% of surplus (Gross yield − 1.075× sUSDS APY)
- Result: Post-buyback APY = exactly 1.075× sUSDS APY (no more, no less)
Concrete example: If sUSDS shows 5% and sUSDe generates 8%, this means post-buyback APY is fixed at exactly 5.375% (1.075× 5%) and buyback corresponds to 2.625% (100% of 8% − 5.375%).
Philosophy: Absolute buyback maximization under competitiveness constraint. Sacrifices all yield volatility for maximum volume. sUSDe becomes a yield product “capped” at 1.075× sUSDS.
Scenario 4 — Progressive Brackets (5% to 25%)
Logic: S4 introduces progressivity in the capture rate to avoid S1’s “cliff” effects. The buyback rate increases progressively in brackets from 5% to 25% as the gross ratio rises, while mathematically guaranteeing the post-buyback ratio remains ≥ 1.075×.
Operation:
- Inactive if gross ratio < 1.132×
- 5% bracket if 1.132× ≤ gross ratio < 1.195×, capture of 5% of total yield
- 10% bracket if 1.195× ≤ gross ratio < 1.265×, capture of 10%
- 15% bracket if 1.265× ≤ gross ratio < 1.342×, capture of 15%
- 20% bracket if 1.342× ≤ gross ratio < 1.433×, capture of 20%
- 25% bracket if gross ratio ≥ 1.433×, capture of 25%
Concrete example: If sUSDS shows 5% and sUSDe generates 6.5% (gross ratio = 1.3×), we’re in the 15% bracket. Buyback will be 0.975% (15% × 6.5%) and post-buyback APY 5.525% (with final ratio = 1.105×, above 1.075×).
Philosophy: Buyback smoothing according to market conditions. Progressive activation as soon as possible, yield volatility preservation (unlike S3), but at the cost of significant intermittency (active only when ratio > 1.132×).
Scenario 5 (S5) — Enhanced S2 with Reserve
Logic: S5 builds on S2’s principle (50% of surplus) but introduces a reserve mechanism. Instead of allocating everything to direct buyback, the surplus is divided: 25% to immediate buyback, 25% to a reserve that will be deployed later.
This reserve accumulates during market periods when sUSDe’s APY is particularly high compared to the benchmark. It’s then redistributed at 0.5% per day to maintain active buyback even when conditions no longer naturally allow it.
Operation:
- Accumulation if gross ratio ≥ 1.075×
- 25% of surplus (Gross APY − 1.075× sUSDS APY) to direct buyback and 25% to reserve
- Deployment if gross ratio < 1.075×
- No direct buyback possible, deployment of 0.5% of reserve per day to maintain buyback
Concrete example: If sUSDS shows 5% and sUSDe generates 7%, threshold is met (1.075× 5% = 5.375%) and available surplus is 1.625% (7% − 5.375%). This allows 0.4075% (25% × 1.625%) for buyback and the same amount to reserve. However, if sUSDS = 5% and sUSDe = 5% (gross ratio = 1.0×, below criterion), no direct buyback is possible and reserve deploys 0.5% per day.
Philosophy: Buyback continuity through the reserve. Accepts lower direct volume (25% instead of 50% of surplus) to guarantee quasi-permanent activity. Smooths market cycles by accumulating in favorable periods and deploying in difficult periods.
Scenario 6 (S6) — Enhanced S4 with Reserve
Logic: S6 combines S4’s progressive brackets with the reserve mechanism. Following the same idea as the previous scenario, each bracket allocates 50% of the capture to direct buyback and 50% to reserve. This preserves S4’s advantages (progressivity, yield volatility preservation) while adding the continuity provided by the reserve.
Operation:
- Accumulation if gross ratio ≥ 1.132×, according to appropriate bracket (5%, 10%, 15%, 20%, or 25%)
- Captured amount is divided in 2: 50% for direct buyback, 50% for reserve
- Deployment if gross ratio < 1.132×
- No direct buyback possible, deployment of 0.5% of reserve per day to maintain buyback
Concrete example: If sUSDS = 5% and sUSDe = 6.5% (gross ratio = 1.3×, 15% bracket), total capture is 0.975% (15% × 6.5%). Thus, we have 0.4875% (50% of 0.975%) for direct buyback and 0.4875% for reserve. Post-buyback APY is 5.525% (6.5% − 0.975%).
Philosophy: Best compromise among all objectives. Maintains progressivity and yield volatility (like S4) while guaranteeing buyback continuity (like S5). S6 sacrifices 50% of direct volume to obtain quasi-permanent activity.
Quantitative Results
Overview
The following table presents key metrics for each scenario over the entire analyzed period (September 1, 2024 – February 13, 2026). The objective is to simultaneously evaluate three dimensions: preserving sUSDe’s competitiveness (average APY), buyback efficiency (volume and price), and mechanism continuity (active days).
By Scenario
Scenario 1 (S1): Prudence First
S1 stands out for its conservative approach: it best preserves sUSDe’s APY (7.19% average) because its activation criterion is deliberately strict. Buyback only starts when the sUSDe APY / sUSDS APY ratio reaches 1.433×, guaranteeing that after 25% capture, sUSDe broadly maintains its competitive advantage.
This prudence has a cost: buyback is only active 30.5% of the time (162 days out of 531), mainly during the late 2024 bull phase. Total volume remains modest ($45.3M) and average acquisition price is the highest of all scenarios ($0.602/ENA). Under current market conditions (relatively calm), S1 would be virtually invisible, changing almost nothing from the no-fee-switch situation.
In summary: Excellent for preserving yield, poor for buyback visibility. An unsuitable choice if the objective is to establish structural buying pressure on ENA.
Scenario 2 (S2): Simple Balance
S2 offers an interesting compromise by capturing only 50% of available surplus beyond the competitiveness threshold. The result is surprising: average APY (7.18%) is almost identical to S1, as is total volume ($45.3M), but with a major difference in temporal distribution.
Unlike S1, S2 is active 57.4% of the time (305 days out of 531) thanks to its more accessible activation criterion (ratio ≥ 1.075× vs 1.433× for S1). This means that in normal periods, S2 maintains visible buyback presence where S1 completely disappears. Average acquisition price is slightly better ($0.571/ENA, -5% vs S1).
In summary: Same yield and volume as S1, but with significantly superior continuity. A defensible option if mechanism simplicity is priority.
Scenario 3 (S3): Maximalist Buyback
S3 adopts the most aggressive approach: it captures 100% of surplus beyond the competitiveness threshold, systematically fixing post-buyback yield at exactly 1.075× sUSDS APY. The volume result is spectacular ($90.6M buyback), double that of S1/S2/S4 scenarios.
However, it’s immediately disqualified because average APY (6.17%) falls below the minimum annual criterion (6.87% corresponding to 1.075× average sUSDS APY of 6.39%). This means a user holding sUSDe obtains an annualized return below the required competitiveness criterion.
In summary: Record buyback volume, but at the cost of progressive competitiveness erosion. Eliminated outright despite impressive buyback performance.
Scenario 4 (S4): Progressivity
S4 improves upon S1 by introducing progressive brackets (5% → 10% → 15% → 20% → 25%) that adapt to market conditions. This progressivity allows buyback activation earlier than S1 (as soon as ratio > 1.132×) while maintaining a comfortable safety margin.
The result is an interesting balance: average APY slightly lower than S1 and S2 but broadly qualified (7.08%) and superior buyback volume ($50.9M). However, S4 remains handicapped by its intermittency: nearly half the time (267 days out of 531), buyback is inactive. This is better than S1 but insufficient to establish truly impactful buying pressure.
In summary: Best basic scenario (without reserve) in terms of yield/volume balance, but still penalized by insufficient continuity.
Scenario 5 (S5): Continuity at Volume’s Cost
S5 transforms S2 by adding a reserve mechanism: instead of allocating 25% of surplus to direct buyback, it divides this amount in two (12.5% buyback, 12.5% reserve). This accumulated reserve is then deployed at 0.5% per day when conditions no longer allow natural buyback.
The impact on buyback continuity is radical (active 95.3% of the time, 506 days out of 531). Average APY remains excellent (7.18%, identical to S2), fully preserving competitiveness. Final reserve stands at $12.0M, offering an additional buffer for approximately 60 days of buyback.
The compromise lies in volume ($33.3M total, or -26% compared to S2). However, average acquisition price ($0.479/ENA) is significantly better thanks to the reserve’s counter-cyclical effect: dollar accumulation in bull when ENA is expensive, deployment in normal period when it’s potentially cheaper.
In summary: Sacrifice of immediate volume (-26% vs S2) to obtain quasi-total continuity (95%). A strategic choice prioritizing long-term ENA price stability.
Scenario 6 (S6): Optimal Compromise
S6 applies the same logic as S5 (buyback/reserve split) but on S4’s progressive brackets. Each bracket allocates half its capture to immediate buyback and the other half to reserve. This combination captures advantages from both approaches: progressivity and continuity.
Results reflect this balance: average APY of 7.08% (identical to S4, broadly qualified), exceptional continuity at 95.3% like S5, but with slightly superior volume ($38.7M, +16% vs S5). Bracket progressivity allows better capture of bull market opportunities while maintaining sustained activity in normal periods. S6 also achieves the best average acquisition price ($0.451/ENA).
In summary: The only scenario that checks all boxes without critical weakness: preserved yield (7.08%), maximum continuity (95%), optimal purchase price ($0.451), decent volume ($38.7M), and proper reserve.
Results by Market Regime
Market regime analysis reveals radically different behaviors across scenarios. This granularity is essential to understand how each mechanism adapts to actual conditions and to anticipate their performance in similar future contexts.
Bull Period (63 days, November 5, 2024 – January 6, 2025)
This phase is quite exceptional as it corresponds to the post-Trump election bull period in crypto markets, characterized by very high funding rates that allowed sUSDe’s APY to reach extremely high levels (20-30%). Despite an APY set above 13% by Sky to support sDAI to sUSDS migration, the competitiveness criterion was broadly exceeded.
S1 and S4 are strictly identical in bull period. When APY/sUSDS ratios broadly exceed 1.433×, S4’s progressive brackets naturally converge toward the maximum 25% bracket, producing exactly the same results as S1. Both offer high average yield (15.76%) and accumulate approximately $20M in buybacks at $0.773/ENA. sUSDe holders fully benefit from upside: 15.75% vs 10.63% for sUSDS benchmark.
S2 offers slightly lower yield for the period (15.35% average) because it starts capturing as soon as ratio exceeds 1.075×, while S1/S4 only activate maximum capture beyond 1.433×. In return, S2 generates +8% volume ($21.5M vs $19.9M) through earlier activation. This is an acceptable trade-off: -0.41pp median yield for +8% buyback.
S3 is the massive accumulator with $43.0M buybacks (+116% vs S1/S4), but at the cost of compressed yield at 10.96% median. Users lose 40% of potential upside (16.75% → 10.96%). Certainly, 10.96% remains very competitive (above sUSDS at 10.63%), but sUSDe’s value proposition is weakened: why choose sUSDe if it only marginally outperforms sUSDS in bull market? S3 transforms the product into a “slightly better sUSDS” instead of a “high variable yield product.”
S5 and S6 distinguish themselves by their moderation: they accumulate only $11-12M in direct buyback (about half of S1/S2/S4), but preserve yield at the same level as S2 (exactly 15.35% for S5, 15.76% for S6). More importantly, these scenarios accumulate $10.7-11.0M in reserve during this phase. Herein lies their strategic advantage: avoiding massive ENA purchases at high prices ($0.77-0.79/ENA) during euphoria, to accumulate capital that will be deployed later when ENA is cheaper and buybacks aren’t active.
Continuity in bull: S5 and S6 are active 100% of the time (63/63 days) because their reserve mechanism intervenes even on days when ratio would be slightly below thresholds. Other scenarios are active 69.8% of the time (44/63 days), representing 19 days without buyback even during this favorable phase.
Normal Period (468 days, or 88% of total period)
This phase reflects current market conditions (January 2025 – February 2026): average sUSDe APY at 6.6%, with compressed delta versus sUSDS (5.8% vs 6.6%, or only +0.8pp). It’s under these conditions that structural differences between scenarios become critical.
S1 becomes virtually invisible with only 25.2% active days (118 days out of 468). Its strict criterion (ratio ≥ 1.433×) is rarely met when spreads compress. For ENA holders, this means 350 days out of 468 (75%) without any buyback signal. This is too long an absence to maintain positive token dynamics. Average yield remains correct (6.04%), but extreme intermittency disqualifies S1 as a viable solution.
S2 becomes more active (55.8% of the time, 261 days) and generates interesting volume ($23.8M) at a correct price ($0.467/ENA). However, approximately 45% of days remain without buyback (207 inactive days), creating 1 to 3-week gaps where no buying pressure is visible. These repeated interruptions create uncertainty for ENA investors. S2 is better than S1 but remains fragile in terms of continuity.
S3 maintains activity similar to S2 (55.8%) and generates the highest volume ($47.6M, +100% vs S2), but its average yield drops to 5.53%. Once again, it falls below sUSDS level and is thus disqualified for this reason.
S4 offers an interesting compromise: average yield of 5.91%, volume of $31.1M (+22% vs S2), active 47% of the time (220 days). This is better than S1 and S2 in terms of yield/volume balance, but continuity remains insufficient with 248 days without buyback out of 468. S4 is the “best without reserve,” but it still exposes ENA token to significant intermittency.
S5 and S6 completely transform dynamics with exceptional continuity of 94.7% active days (443 out of 468, or only 25 days without buyback over 15+ months), excellent acquisition price ($0.395/ENA and $0.385/ENA, or 25-33% cheaper than other scenarios), and especially a reserve accumulated in bull ($10.7-11.0M) progressively deployed during normal period.
Final reserves remain positive: Despite 468 days of normal period with active deployment, S5 and S6 end with $12.0-12.3M in reserve. This means that even after traversing 88% of the period under tight conditions, the mechanism remains sustainable. The reserve doesn’t deplete and continues oscillating between accumulation (favorable days) and deployment (difficult days), creating dynamic equilibrium.
Bull vs Normal Comparison and Final Observation
In bull market, all scenarios “function” in the sense that they generate buyback and maintain competitiveness. Differences mainly concern yield level offered to holders (S1/S4 offer best, S3 worst) and accumulated volume (S3 maximal, S5/S6 moderate).
Since normal period represents 88% of time (468 days out of 531), it’s normal period performance that determines optimal choice. A scenario excellent in bull but mediocre in normal is a poor choice. Conversely, a balanced scenario in both regimes (like S6) is structurally superior.
Analysis and Data Interpretation
Structural Trilemma and Scenario Design
Activating a fee switch on Ethena confronts governance with a constraint that all models must manage and cannot be simply resolved: preserving sUSDe’s yield to maintain its competitiveness, generating sufficiently high buyback volume to support ENA token, and ensuring mechanism continuity to avoid counterproductive intermittency.
This trilemma stems directly from Ethena’s nature. The protocol generates revenue through the reserve backing USDe and redistributes it to sUSDe holders. By taking a portion of this revenue to implement ENA buyback, the risk is imposing a constraint on sUSDe’s competitiveness versus competitors, notably sUSDS.
- To offer the best possible yield to sUSDe holders, revenue allocated to buyback must be minimized, which mechanically reduces its volume, interest, and impact
- To maximize this volume, a significant portion of revenue must be captured, which compresses APY and can even make it fall below the minimum annual criterion (as demonstrated by S3’s disqualification at 6.17%)
- To maintain buyback continuity over time, independent of market conditions, a reserve mechanism must be introduced that diverts capital that could have been used immediately for direct purchases
The design of the 6 models presented above reflects this trilemma. Scenarios S1, S2, and S4 prioritize different direct buyback approaches (strict condition, surplus sharing, and progressive brackets) without reserve mechanisms. They generally offer better average APYs and correct volumes, but result in low buyback activity rates.
Scenarios S5 and S6 introduce the reserve mechanism as a direct response to the continuity problem. By systematically dividing capture between immediate buyback and accumulation for later deployment, these scenarios accept slightly lower direct volume to guarantee quasi-permanent activity.
This approach partially resolves the trilemma by transforming a binary mechanism (buyback active or inactive according to conditions) into a continuous mechanism (buyback always active thanks to reserve). The price to pay is reduced immediate volume, but the gain is more important: continuity perceived by ENA investors and temporal purchase optimization (buying pressure in less euphoric market phases).
→ Given these elements, we believe Scenario 6 emerges as the optimal choice. It combines progressive brackets that preserve volatility and sUSDe yield quality with the reserve mechanism guaranteeing buyback continuity. Metrics confirm this: average APY of 7.08% (broadly qualified), 95.3% continuity, optimal acquisition price ($0.451/ENA), and total mobilized volume of $51M ($38.7M direct and $12.3M reserve).
Buyback Intermittency as Major Risk
Data analysis confirms an element we discussed in our previous analysis on Ethena’s fee switch: intermittent buyback can be more harmful to ENA token than no buyback at all. This assertion deserves development as it constitutes the central argument in favor of reserve scenarios.
In total absence of buyback (current situation), ENA’s price evolves according to natural supply and demand, protocol fundamentals, and general crypto market sentiment. Investors have no expectations and therefore no disappointment is possible. ENA price fluctuations aren’t amplified by buyback-related expectations.
The risk of intermittent buyback is that dynamics change. When buyback is active, it creates visible buying pressure that attracts opportunistic investors anticipating mechanism continuation. When buyback stops, these same investors may be tempted to sell. This creates volatility that wouldn’t exist in total absence of mechanism.
In other words, intermittent buyback can therefore be more harmful to ENA token than no buyback at all. Absence of buyback is neutral because it neither attracts nor disappoints anyone. However, buyback that starts then stops unpredictably creates volatility that impacts long-term holder confidence.
Scenarios S5 and S6, by maintaining activity 95.3% of the time (506 days out of 531, or only 25 days without buyback over 17+ months), virtually eliminate this risk. ENA investors know buyback is active practically every day and this predictability reduces volatility risks linked to the mechanism itself.
Volume Question and Buyback Materiality
Beyond the trilemma and intermittency risk, analysis raises a more fundamental question: are buyback volumes generated by these scenarios sufficiently significant to justify fee switch activation?
Scenario S6, which we consider optimal, generates $38.7M direct buyback over 531 days, or approximately $26.6M annualized (converting to 365 days). Final reserve of $12.3M brings total mobilized capital to approximately $51M over the entire period. In terms of daily buying pressure, this represents an amount of $73k/day that may seem modest relative to ENA token’s market cap and daily trading volumes.
OAK Research data on ENA indicates daily volume around $74M. This means buybacks would represent approximately 0.1% of daily trading volume, far from the 1-2% necessary to truly have impact. This finding is all the more important as ENA still faces several years of significant upcoming emissions (over $300 million in 2026 at current price).
Announcement of a fee switch by Ethena had been perceived by the community as a game changer for ENA, as confusion is common between revenue generated by USDe (several hundred million dollars distributed to sUSDe holders) and actual protocol revenue. Launching a buyback program with such low amounts can create disappointment that pushes price down rather than up.
This is all the more concerning in current context. We’re in “normal” market regime since January 2025, characterized by very compressed spreads between sUSDe and sUSDS (+0.2 to 0.5pp only). Under these conditions, even best scenarios struggle to accumulate capital (only $1.6M sent to reserve), proof that the mechanism works moderately under current conditions if not preceded by a bull period.
Conclusion and Recommendations
Current Context is Unfavorable for Fee Switch
Based on quantitative analysis of various scenarios, we consider S6 to be technically the most appropriate fee switch mechanism for Ethena’s situation. It preserves sUSDe’s yield (7.08% average vs 8.2% currently) and enables buyback continuity (95.3% of time active) while maximizing ENA purchase impact (average price of $0.451).
However, qualitative analysis of current context leads us to support the thesis discussed in our previous analysis: we recommend not activating a fee switch for now, despite S6’s intrinsic qualities. This position rests on three critical findings.
Compressed spreads drastically limit volumes. In the current “normal” market period (since January 2025), gaps between sUSDe and sUSDS are reduced (+0.2 to 0.5pp only). Even the best scenario generates derisory buyback: $21.6M annualized, or approximately $59k/day. This amount represents barely 0.1% of ENA’s daily trading volume, meaning buyback would have marginal impact, easily absorbed by normal supply and demand flow.
Reserve cannot fill under normal conditions. S6’s final $12.3M reserve comes almost entirely from bull phase ($10.7M over 63 days). In normal period, reserve only progressed by $1.6M over 468 days. Activating fee switch now would mean starting with an empty reserve that would struggle to fill, leading to either non-significant buyback or discontinuity that the mechanism is precisely supposed to avoid.
Yield sacrifice can be counterproductive. Compressing yield from 8.2% to 7.08% (-1.12pp) in an environment of intense competition between yield-bearing stablecoins could result in TVL loss on sUSDe. As discussed in our previous analysis, Ethena’s growth is driven by USDe, not ENA. USDe is the economic engine, while ENA is an essential incentive layer but doesn’t generate sustainable demand for products. Thus, sacrificing USDe’s competitiveness for marginal-impact buyback would be illogical.
Buyback Lessons and S6 Validation
Although we recommend not activating fee switch now, analysis reveals important lessons for future activation when conditions are favorable.
Buyback remains the best theoretical alignment mechanism. It creates a direct link between protocol’s economic value and token value. This is an important mechanism for giving value to ENA, in a context where governance tokens and altcoins generally no longer attract investors.
But size and execution matter more than principle. Our observations converge with industry observations for several months: some buyback programs are too small to be effective and function more as communication operations than as genuine market support. It’s not that buybacks are fundamentally flawed, but that size and execution matter far more than teams often assume.
Timing is critical. Many programs failed by buying tokens when prices and revenues were already high rather than during downturns. Others lacked consistency, creating post-announcement bumps but no real revaluation over time. This is precisely the problem S6 intends to solve with reserve: it accumulates capital in bull phase (when ENA is expensive, at $0.77 average) and deploys it in normal phase (when ENA is more accessible, around $0.38 average).
Frequency and predictability inspire confidence. Frequent, predictable programs based on mechanical rules are more effective than one-off buybacks managed by the team. S6 checks these boxes: quasi-daily activity (95%), automatic rules (brackets + reserve), without human intervention. This is exactly the type of program that generates confidence when volumes are sufficient to have material impact.
ENA Tokenomics as Barrier
Beyond current market conditions, a deeper obstacle argues against immediate activation: ENA’s tokenomics remain problematic. As detailed in our previous analysis, the emission schedule creates structural selling pressure that far exceeds any realistic buyback’s capacity to counterbalance.
Significant ENA emissions are still scheduled in coming quarters (over $300 million in 2026 at current price). Thus, a $26M annual buyback (annualized S6) doesn’t truly compensate for these unlocks.
The optimal strategy would be to wait for the end, or at least significant attenuation, of unlocks before activating fee switch. At that point, each dollar of buyback would contribute to net supply reduction rather than counterbalancing emission. Impact would be structurally stronger, as buyback/emission ratio would become favorable.
This delay would also allow refining S6 parameters based on new data collected by then.
Alternatives to Consider
While awaiting favorable conditions, several alternatives deserve exploration to use excess revenue more effectively.
Direct distribution to sENA holders. The idea of transforming ENA into a cash-flow generating asset via redistribution to stakers could create measurable incentive to lock the token. However, this is a complicated mechanism from a regulatory standpoint and how the token might be perceived in different jurisdictions.
Maximizing sUSDe yield. Offering 8.2% (vs 6.39% for sUSDS, or +28% relative) is a direct competitive advantage for attracting TVL. Each percentage point can attract hundreds of millions additional, exponentially reinforcing revenue. ENA token then captures value via protocol growth rather than via buyback of uncertain efficiency.
Reinvestment in development. The annual $26M could fund new products, multi-chain expansion, or aggressive marketing versus sUSDS. These investments potentially have superior ROI to $60k/day buyback, as they structurally increase future revenues.
These alternatives aren’t exclusive and could be combined according to market phases: fee switch in bull (wide spreads), reallocation in normal (compressed spreads). This flexibility is preferable to a rigid mechanism functioning sub-optimally.
Final Recommendation and Decision Process
We recommend not activating a fee switch under current conditions, for four converging reasons: compressed spreads (insufficient daily buyback volumes), reserve that doesn’t fill much, problematic tokenomics, and more effective alternatives in our context.
This delay isn’t a rejection of fee switch, but recognition that timing is as important as mechanism. Activating today could even be counterproductive if created expectations exceed amount reality, generating disappointment that pushes price down.
If and when conditions evolve, then a mechanism similar to S6 should be privileged by governance. It’s conceptually aligned with best practices (frequency, predictability, temporal optimization) and presents the best balance.
Note that this analysis is in no way a study funded by Ethena, and constitutes solely research conducted by OAK Research teams with the aim of studying different possibilities available to the protocol should the fee switch activation be decided by the community.