Summary
Ramp A of the following DOLA pools over 1 week:
| Pool | Current A | Proposed A |
|---|---|---|
| DOLA/USR | 200 | 750 |
| DOLA/sUSDe | 200 | 750 |
| DOLA/sUSDS | 200 | 750 |
| sDOLA/scrvUSD | 500 | 1000 |
Abstract
Included in this proposal is an order flow analysis and simulations that indicate optimal A parameter for each respective pool. Our goal is to first determine if parameter optimization could be beneficial, based on its competitive order flow standing within COW Swaps, considering metrics such as volume, trade frequency, and trade size. Secondly, we run simulations to identify a target liquidity concentration that optimizes pool performance.
The amplification factor determines the concentration of liquidity at the traded price. A lower amplification factor would turn the AMM into a constant product function, and a higher amplification factor would make the AMM resemble a constant sum function. In other words, if the subjected pair is expected to have a constant or slow-moving exchange rate, it is ideal to ramp up the liquidity concentration to get maximum capital efficiency.
For optimizing the amplification factor, statistical attributes (min, max, mean, quartiles) are computed from the trades between the tokens in the subjected pool. All the trades originating from the pool, dex aggregators, arbitrageurs, etc, are considered while finding the statistical attributes. This would help to create synthetic trades.
The simulator has arbitrageur agents who capitalize on profit opportunities via rebalancing the pool and trading agents who execute the synthetic trades. With n trades and m parameter combinations, the simulation is concluded.
With simulated data for n trades, we compute the following:
- Cumulative loss due to fee & slippage
- Cumulative fees earned towards the protocol and the LPs
For m parameter combinations, a linear regression gives the slope of both of the above factors. Maximising the A should be constrained by the diminishing returns; thus, the following ratio is plotted against A to check this behavior and pick a suitably high A.
maximizeRatio = slope_fees/slope_loss
We identify the ratio that maximizes the gains for the fees accumulated and minimizes the losses generated from slippage, and elect for the most consevative A value that most greatly improves expected pool performance.
Analysis
See in the dropdowns below a thorough order flow and simulations analysis for each pool pair.
DOLA/USR
DOLA/USR Order Flow Analysis
This section presents an empirical analysis of the pool’s performance in COW Swaps over the past year. We first examined direct competitors for DOLA/USR trades, then broaden our scope to include pools competing for either DOLA or USR volume. We observed the DOLA/USR pool captured all COW Swap volume for this specific pair, indicating no direct competition.
Next, we examined COW Swap flow for pools competing for either DOLA or USR volume. The DOLA/USR Curve pool ranked 13th by volume over the past year for COW Swaps among pools trading either DOLA or USR.
Source: Dune
We observe this pool facilitating relatively large trades (typically ~$18k, if we look at median volume) but with relatively low trade frequency compared to pools that accumulated more volume over the period.
Note: In the below chart, blue shows trade frequency of other DEX pools, gray shows trade frequency of this DOLA/USR pool, and orange shows trade frequency of other curve pools.
Subsequent charts should be interpreted similarly.
Source: Dune
This pool most frequently facilitated trade sizes between $10k and $208k. Trade frequency was relatively low, however.
Source: Dune
This pool accumulated most of its volume from trade sizes between $10k and $208k. Curve pools dominated volume for all buckets, particularly between $10k and $208k.
Source: Dune
This pool made up a small proportion of overall volume for swaps involving either DOLA or USR; there was a lot of competition from other protocols for DOLA or USR volume, and other Curve pools exhibit dominance for trading these tokens.
This chart shows the historical TVL for curve pools which either trade DOLA or USR.
Conclusion
- This pool typically facilitated trade sizes ~$18k, but relatively infrequently.
- It has no direct competition for DOLA/USR flow.
If we suggest that retail trade activity tends to have a trade size <$10k our analysis suggests arbitrage (or whale) activity where trade sizes tend to be large (but retatively infrequently). Increasing the fee may be advisable to increase welfare for LPs.
Exploring optimizations for the ‘A’ parameter can also be explored. Such adjustments could potentially increase trade frequency, especially for higher trade sizes. This would make volume “stickier,” which in turn could make a fee increase viable.
DOLA/USR Simulations
{
"chain": "ethereum",
"address": "0x38De22a3175708D45E7c7c64CD78479C8B56f76E",
"data": [
{
"token_supply": 33886797.861491635,
"balances": [
18030596.310635258,
15923695.481096948
],
"token_prices": [
0.9999864930372862,
1.0001703510199142
],
"token_price": 1.0024726044390129,
"usd_price": 1.0027620272293662,
"tvl_usd": 33980394.57394384,
"block_number": 22662430,
"timestamp": 1749340800
}
]
}
{
"chain": "ethereum",
"address": "0x38De22a3175708D45E7c7c64CD78479C8B56f76E",
"data": [
{
"timestamp": 1749412800,
"a": 200,
"fee": 4000000,
"admin_fee": 5000000000,
"virtual_price": 1001982034474545000,
"xcp_profit": null,
"xcp_profit_a": null,
"base_daily_apr": 0.003940606575437489,
"base_weekly_apr": 0.001349521949510546,
"offpeg_fee_multiplier": 30000000000,
"gamma": null,
"mid_fee": null,
"out_fee": null,
"fee_gamma": null,
"allowed_extra_profit": null,
"adjustment_step": null,
"ma_half_time": null,
"price_scale": null,
"price_oracle": [
1000604915652127600
],
"block_number": 22662112
}
]
}
A=50: losses_slope=194.733308, fee_slope=196.171470, ratio=1.0074
A=100: losses_slope=125.535933, fee_slope=179.825457, ratio=1.4325
A=250: losses_slope=89.949553, fee_slope=162.942489, ratio=1.8115
A=500: losses_slope=74.505801, fee_slope=143.672477, ratio=1.9283
A=750: losses_slope=72.237536, fee_slope=140.954487, ratio=1.9513
A=1000: losses_slope=68.357588, fee_slope=134.080064, ratio=1.9615
A=2000: losses_slope=67.141146, fee_slope=133.142427, ratio=1.9830
DOLA/sUSDe
DOLA/sUSDe Order Flow Analysis
This section presents an empirical analysis of the pool’s performance in COW Swaps over the past year. We first examine direct competitors for DOLA/sUSDe trades, then broaden our scope to include pools competing for either DOLA or sUSDe volume. The DOLA/sUSDe pool captured all COW Swap volume for this specific pair, indicating no direct competition. Trade frequency was low, but the average trade size being significantly greater than $10k implies that whale or large arbitrage trades are typically being facilitated on this pool.
Source: Dune
Next, we examine COW Swap flow for pools competing for either DOLA or sUSDe volume. The DOLA/sUSDe Curve pool ranked 6th by volume over the past year for COW Swaps among pools trading either DOLA or sUSDe.
Source: Dune
Compared to other top 10 pools in this category, it exhibited a higher average trade size (i.e. median volume) than the rest and reduced trade frequency. Despite this relatively low trade frequency, a median volume of $25k implies trade activity primarily from large arbitrageurs or whales. However, its infrequent utilization suggests that parameter optimization could be beneficial.
Note: Blue shows trade frequency of other pools, gray shows trade frequency of this DOLA/sUSDe pool, and orange shows trade frequency of other curve pools.
Subsequent charts should be interpreted similarly.
Source: Dune
This pool most frequently facilitated trade sizes between $10k and $208k. It least frequently facilitated trade sizes larger than this.
Source: Dune
This pool accumulated most of its volume from trade sizes between $10k and $208k, whereas many other Curve pools typically amassed volume from larger trade sizes.
Source: Dune
This pool struggled to capture COW Swap volume compared to its competitors.
This chart shows the historical TVL for curve pools which either trade DOLA or sUSDe.
Conclusion
- This pool typically facilitated larger trades, often between $10k and $208k.
- It was infrequently selected for COW Swap trades.
- It has no direct competition for DOLA/sUSDe flow.
Based on these points, one could argue for increasing the current fee:
- It already has the same or a higher fee tier than its competitors, but is less frequently selected for COW Swap trades.
- despite this, it typically facilitates trade sizes larger than its competitors
Exploring optimizations for the ‘A’ parameter is also advisable, and may help mitigate drawbacks from increasing the fee tier.
Such adjustments could potentially increase trade frequency for larger trade sizes, making the volume “stickier.” This, in turn, could make a fee increase more viable, ultimately increasing LP welfare (but fee increases should be explored after monitoring pool activity post-param change).
DOLA/sUSDe Simulations
{
"chain": "ethereum",
"address": "0x744793B5110f6ca9cC7CDfe1CE16677c3Eb192ef",
"data": [
{
"token_supply": 17179065.192482807,
"balances": [
9597231.880147973,
6791143.988240621
],
"token_prices": [
1.0003090024101882,
1.1780846245860592
],
"token_price": 1.024432506997157,
"usd_price": 1.0247007261750805,
"tvl_usd": 17603400.58871608,
"block_number": 22662427,
"timestamp": 1749340800
}
]
}
{
"chain": "ethereum",
"address": "0x744793B5110f6ca9cC7CDfe1CE16677c3Eb192ef",
"data": [
{
"timestamp": 1749412800,
"a": 200,
"fee": 4000000,
"admin_fee": 5000000000,
"virtual_price": 1023792437285359700,
"xcp_profit": null,
"xcp_profit_a": null,
"base_daily_apr": 0.02953363342680415,
"base_weekly_apr": 0.027576740762845775,
"offpeg_fee_multiplier": 20000000000,
"gamma": null,
"mid_fee": null,
"out_fee": null,
"fee_gamma": null,
"allowed_extra_profit": null,
"adjustment_step": null,
"ma_half_time": null,
"price_scale": null,
"price_oracle": [
1000862287161083300
],
"block_number": 22662112
}
]
}
A=50: losses_slope=714.257532, fee_slope=358.265184, ratio=0.5016
A=100: losses_slope=424.688922, fee_slope=353.360043, ratio=0.8320
A=250: losses_slope=239.543144, fee_slope=329.435954, ratio=1.3753
A=500: losses_slope=176.843893, fee_slope=302.384763, ratio=1.7099
A=750: losses_slope=154.561653, fee_slope=281.222048, ratio=1.8195
A=1000: losses_slope=147.724344, fee_slope=269.522239, ratio=1.8245
A=2000: losses_slope=135.638132, fee_slope=255.924771, ratio=1.8868
DOLA/sUSDS
DOLA/sUSDS Order Flow Analysis
This section presents an empirical analysis of the pool’s performance in COW Swaps over the past year. We first examine direct competitors for DOLA/sUSDS trades, then broaden our scope to include pools competing for either DOLA or sUSDS volume. The DOLA/sUSDS pool captured all COW Swap volume for this specific pair, indicating no direct competition.
Source: Dune
Next, we examine COW Swap flow for pools competing for either DOLA or sUSDS volume. The DOLA/sUSDS Curve pool ranked 13th by volume over the past year for COW Swaps among pools trading either DOLA or sUSDS.
Source: Dune
Compared to other top 10 pools in this category, it was one of the pools which exhibited higher average trade size than the rest (again, we use median volume to measure this) and reletively low trade frequency (lowest in top 13). This pattern might indicate it primarily serves whale traders or facilitates larger arbitrage trades.
Note: Blue shows trade frequency of other DEX pools, gray shows trade frequency of this DOLA/sUSDS pool, and orange shows trade frequency of other curve pools.
Subsequent charts should be interpreted similarly.
Source: Dune
This pool most frequently facilitated trade sizes between $10k and $208k. But trade activity was notably infrequent. Curve pools dominated the <$10k bucket.
Source: Dune
This pool accumulated most of its volume from trade sizes between $10k and $208k, Curve pools typically accumulated volume in this bucket, as well.
Source: Dune
Curve dominated volume for DOLA and/or sUSDS.
This chart shows the historical TVL for curve pools which either trade DOLA or sUSDS.
Conclusion
- This pool typically facilitated larger trades (>$10k), but did so relatively infrequently.
- It has no direct competition for DOLA/sUSDS flow.
Our analysis suggests arbitrage (or whale) activity where trade sizes tend to be large, and consequently, potentially high LVR. Increasing the fee could increase welfare for LPs, but optimizations of other params can be sought as well.
Exploring optimizations for the ‘A’ parameter is advisable. Such an adjustment could potentially increase trade frequency, especially for higher trade sizes. This would make volume “stickier,” which in turn could make a fee increase potentially advisable further down the line.
DOLA/sUSDS Simulations
{
"chain": "ethereum",
"address": "0x8b83c4aA949254895507D09365229BC3a8c7f710",
"data": [
{
"token_supply": 15621342.152536208,
"balances": [
7791942.688728311,
7675458.618712658
],
"token_prices": [
1.0002103009425212,
1.0544180349982735
],
"token_price": 1.0170663613894289,
"usd_price": 1.0173921218860178,
"tvl_usd": 15893030.439276306,
"block_number": 22662431,
"timestamp": 1749340800
}
]
}
{
"chain": "ethereum",
"address": "0x8b83c4aA949254895507D09365229BC3a8c7f710",
"data": [
{
"timestamp": 1749412800,
"a": 200,
"fee": 4000000,
"admin_fee": 5000000000,
"virtual_price": 1016953490992117200,
"xcp_profit": null,
"xcp_profit_a": null,
"base_daily_apr": 0.02447987495748083,
"base_weekly_apr": 0.023701262940889922,
"offpeg_fee_multiplier": 20000000000,
"gamma": null,
"mid_fee": null,
"out_fee": null,
"fee_gamma": null,
"allowed_extra_profit": null,
"adjustment_step": null,
"ma_half_time": null,
"price_scale": null,
"price_oracle": [
999822705414225000
],
"block_number": 22662112
}
]
}
A=50: losses_slope=225.033492, fee_slope=184.658009, ratio=0.8206
A=100: losses_slope=140.535881, fee_slope=176.606328, ratio=1.2567
A=250: losses_slope=94.421492, fee_slope=164.606693, ratio=1.7433
A=500: losses_slope=75.512622, fee_slope=143.002859, ratio=1.8938
A=750: losses_slope=69.679343, fee_slope=135.836818, ratio=1.9495
A=1000: losses_slope=66.533073, fee_slope=129.839280, ratio=1.9515
A=2000: losses_slope=64.756895, fee_slope=127.856603, ratio=1.9744
sDOLA/scrvUSD
sDOLA/scrvUSD Order Flow Analysis
This section presents an empirical analysis of the pool’s performance in COW Swaps over the past year. We first examine direct competitors for DOLA/scrvUSD trades, then broaden our scope to include pools competing for either DOLA or scrvUSD volume. The DOLA/scrvUSD pool captured all COW Swap volume for this specific pair, indicating no direct competition. It also infrequently facilitated trades, average trade size was a modest $12k.
Source: Dune
Next, we examine COW Swap flow for pools competing for either DOLA or scrvUSD volume. The DOLA/scrvUSD Curve pool ranked 12th by volume over the past year for COW Swaps among pools trading either DOLA or scrvUSD.
Source: Dune
Despite this pool having no direct competitors, its infrequent utilization compared to others in this list suggests that parameter optimization could be beneficial.
Note: Blue shows trade frequency of other DEX pools, gray shows trade frequency of this DOLA/sUSDS pool, and orange shows trade frequency of other curve pools.
Subsequent charts should be interpreted similarly.
Source: Dune
This pool most frequently facilitated trade sizes between $10k and $208k. Trade frequency was relatively low, however.
Source: Dune
This pool accumulated most of its volume from trade sizes between $10k and $208k. Curve pools dominated volume for all buckets, particularly between $10k and $208k.
Source: Dune
This pool competed with other pools from competing protocols, with only a 1.9% difference in volume capture. Also shows Curve’s dominance for these tokens.
This chart shows the historical TVL for curve pools which either trade DOLA or scrvUSD.
Conclusion
- This pool typically facilitated trades just over $10k, but relatively infrequently.
- It has no direct competition for DOLA/scrvUSD flow.
If we suggest that retail trade activity tends to have a trade size <$10k our analysis suggests arbitrage (or whale) activity where trade sizes tend to be large (but relatively infrequent). Increasing the fee is therefore advisable to increase welfare for LPs.
Exploring optimizations for the ‘A’ parameter can also be explored. Such adjustments could potentially increase trade frequency, especially for higher trade sizes. This would make volume “stickier,” which in turn could make a fee increase viable, which would increase LP welfare.
sDOLA/scrvUSD Simulations
{
"chain": "ethereum",
"address": "0x76A962BA6770068bCF454D34dDE17175611e6637",
"data": [
{
"token_supply": 4490691.014391347,
"balances": [
2030494.4344083832,
2307009.402106748
],
"token_prices": [
1.1326581155153657,
1.0477804752177415
],
"token_price": 1.0078670166061903,
"usd_price": 1.0560233816159215,
"tvl_usd": 4742274.710809783,
"block_number": 22661492,
"timestamp": 1749405359
}
]
}
{
"chain": "ethereum",
"address": "0x76A962BA6770068bCF454D34dDE17175611e6637",
"data": [
{
"timestamp": 1749412800,
"a": 500,
"fee": 2000000,
"admin_fee": 5000000000,
"virtual_price": 1056025929511942700,
"xcp_profit": null,
"xcp_profit_a": null,
"base_daily_apr": 0.07192561287506494,
"base_weekly_apr": 0.07580229615490852,
"offpeg_fee_multiplier": 50000000000,
"gamma": null,
"mid_fee": null,
"out_fee": null,
"fee_gamma": null,
"allowed_extra_profit": null,
"adjustment_step": null,
"ma_half_time": null,
"price_scale": null,
"price_oracle": [
999580711491304600
],
"block_number": 22662112
}
]
}
A=100: losses_slope=191.216207, fee_slope=65.852992, ratio=0.3444
A=250: losses_slope=90.768626, fee_slope=66.016211, ratio=0.7273
A=500: losses_slope=57.324171, fee_slope=65.188708, ratio=1.1372
A=750: losses_slope=44.455438, fee_slope=62.816246, ratio=1.4130
A=1000: losses_slope=39.235852, fee_slope=60.925079, ratio=1.5528
A=2000: losses_slope=33.032970, fee_slope=55.354330, ratio=1.6757
Specification
RAMP_TIME = chain.time() + (86400 * 14) + 14400
ACTIONS = [
# DOLA/USR ramp A to 750 over 1 week
("0x38De22a3175708D45E7c7c64CD78479C8B56f76E", "ramp_A", 750, RAMP_TIME),
# DOLA/sUSDe ramp A to 750 over 1 week
("0x744793B5110f6ca9cC7CDfe1CE16677c3Eb192ef", "ramp_A", 750, RAMP_TIME),
# DOLA/sUSDS ramp A to 750 over 1 week
("0x8b83c4aA949254895507D09365229BC3a8c7f710", "ramp_A", 750, RAMP_TIME),
# sDOLA/scrvUSD ramp A to 1000 over 1 week
("0x76A962BA6770068bCF454D34dDE17175611e6637", "ramp_A", 1000, RAMP_TIME),
]