Friday Is the Best Day to LP Memecoins. Saturday Is a Tax.

The conventional wisdom on weekend LPing is wrong.

I've heard it a hundred times: "weekends are dead, don't bother LPing." When I started looking at this, I expected to find Saturday and Sunday were both bad. Take both days off, simple article.

What the data actually shows is sharper. Friday is by far the best day to LP trending memecoins on Meteora — the median wallet earns 19x more on Friday than Saturday. Saturday is genuinely a tax. Sunday is fine, actually quite good.

Win rate isn't the headline. Dollar gains are. You can have a 50% win rate and still be very profitable if your wins are big. So the question I cared about was: how much does the median active wallet actually take home on each day of the week?

Before any of those numbers land, I need to walk through how I built this. Every number in this article was verified via direct SQL against the full dataset (no truncation, no shortcuts).

| Stat | Value | |---|---| | Median Friday wallet PnL | $2.48 | | Median Saturday wallet PnL | $0.13 | | Closed positions analyzed | 374K | | Memecoin pools | 1,114 | | Wallets per day-cell | ~2,200–2,700 | | Window | 28 days (April 1–28, 2026) |

How I Picked the Pools

The goal: find the pools where memecoin LPing actually happens. Not the pools where you'd LP SOL/USDC for stable yield. Memecoin pools, where the action is.

In plain English:

1. Every day for the last 28 days, I asked Dune: "which memecoins had the most SOL traded against them today?"
2. I took the top 10 memecoins for each day. Some showed up multiple days, some only once. They all went on one big list.
3. For each memecoin on the list, I found every Meteora pool where you can LP that memecoin against SOL. One memecoin can have 5, 20, even 90+ separate pools at different fee tiers and bin steps — wallets LP across all of them.
4. Final pool count: 1,114 pools.

What I excluded: anything that's not actually a memecoin. SOL/USDC, USDT, cbBTC (Coinbase BTC), JitoSOL, mSOL (liquid staking), JUP, JTO. Stablecoin and major-token LPing is a different game.

Why volume, not fees? If I'd ranked pools by LP fees earned, a pool where LPs earned $0 because no traders showed up would never make the list. We'd never see the dead pools — which are the whole point of studying weekends. Volume captures trader attention regardless of whether LPs profited.

How I Picked the Wallets

The goal: find regulars. People who LP these pools repeatedly. Not whales who deposited once, not single-position lurkers.

1. For each of the 1,114 pools, I asked Dune: "which 20 wallets opened the most positions here in the last 28 days?" Position count, not money earned.
2. After removing duplicates (the same wallet often shows up as a top LPer in many pools): 1,381 unique wallets.
3. I ranked the 1,381 by total activity across all pools and picked the top 300 for the deep dive. Those 300 cover ~89% of all LP positions in the universe — going deeper adds wallets that barely participate.

Critical: I did NOT pick wallets based on whether they made money. Selection is based on activity, not profit. A wallet that LPs 100 times and loses on most of them is in our sample. That's the point — we want to study what actually happens, not just the survivors.

What Data I Pulled

For all 300 wallets, I pulled every LP position they ever closed in any of our 1,114 pools over the 28-day window.

Per position, I got:

• Pool address — which exact pool the position was in
• Opened at / Closed at — the exact dates and times
• Hold duration — how long the position was held, in hours
• Fees earned — how much they collected in trading fees
• PnL (profit/loss) — net result of the position, including fees, IL, and price moves
• Impermanent loss — how much the LP lost to price moves vs holding
• Token symbols — names of the tokens, e.g. "BURNIE", "WOJAK"

That gave us 210,684 closed positions from those 300 wallets.

Then I cross-checked with a totally separate group of wallets — 5,149 wallets I already had data on from a previous study, picked by completely different criteria. I filtered their positions to the same 1,114-pool universe and same 28-day window. Another 156,918 positions from a different population.

And a third group: TrackLP's own quality-scored wallets — 588 wallets with quality_score ≥ 45 from our intelligence engine. These are the wallets we'd flag as worth watching or copying.

If all three groups (Group A: my fresh 300, Group B: the older 5,149, Group C: the 588 scored) show the same day-of-week pattern, it's a real effect — not an artifact of how I picked wallets.

One More Filter: Bots

Some wallets cycle 100, 200, 500 positions per day — clearly automated. Bots don't have a day off. They fire constantly regardless of weekday or weekend. Including them dilutes the human-LPer signal we're trying to study.

I flagged any wallet that opened more than 50 positions on any single day as automated. Conservative threshold — a human rarely opens 50+ positions in 24 hours.

Across all three groups combined, 315 unique wallets were flagged as bots and excluded.

I sensitivity-checked at 30/day and 100/day. The Friday-best, Saturday-worst pattern holds either way.

Do We Have a Good Sample?

The honest answer: yes, by every measure that matters.

After all the filters, here are the active human LPers contributing data for each day of the week (all groups combined, bots excluded):

| Day | Different wallets | Closed positions | |---|---|---| | Monday | 2,172 | 19,933 | | Tuesday | 2,424 | 23,031 | | Wednesday | 2,651 | 25,301 | | Thursday | 2,595 | 24,259 | | Friday | 2,720 | 27,707 | | Saturday | 2,178 | 19,236 | | Sunday | 2,270 | 23,222 |

For context:

• Most published trading research uses samples of a few hundred wallets. We have ~2,200–2,700 different humans per day-of-week.
• Reliable medians need 30+ data points per cell. We have thousands.
• We cross-checked against three independently-selected groups of wallets. All three agreed on the extremes.

Active wallet count is itself a finding: 2,720 wallets open positions on Fridays, only 2,178 on Saturdays. Even active LPers self-select away from Saturday — they sense it's not the day to deploy.

Where the sample falls short:

• Only 28 days. A different month would give different numbers.
• Only trending memecoin pools. Doesn't tell you about LPing SOL/USDC, stables, or established memecoins like BONK and WIF.
• Only closed positions. If a wallet opened a position on Saturday and is still holding it (because they can't get out cleanly), it's not in our data. So if anything, the data slightly under-counts the worst Saturday outcomes.
• UTC days. A position opened 11pm UTC Saturday is "Saturday" here, even if it was Sunday morning where you live.

Three independently-selected groups of wallets, picked by different criteria, all agreed Friday is best and Saturday is worst. That's strong enough to publish a finding. It's not strong enough to predict next Saturday.

The Headline Result

Each wallet's daily haul, median across all active wallets per day-of-week. All three groups combined, bots excluded:

| Day | Wallets | Med positions | Med fees | Med PnL | Daily WR | |---|---|---|---|---|---| | Sunday | 2,270 | 6 | $9.80 | $1.55 | 63.2% | | Monday | 2,172 | 4 | $6.14 | $0.47 | 56.5% | | Tuesday | 2,424 | 5 | $7.93 | $0.48 | 57.0% | | Wednesday | 2,651 | 5 | $6.31 | $0.94 | 60.8% | | Thursday | 2,595 | 5 | $6.52 | $1.36 | 65.6% | | Friday | 2,720 | 6 | $10.47 | $2.48 | 68.8% | | Saturday | 2,178 | 4 | $8.34 | $0.13 | 52.8% |

The median Friday wallet earns $2.48 in PnL plus $10.47 in fees on 6 positions. The median Saturday wallet earns $0.13 in PnL plus $8.34 in fees on 4 positions. Friday's net haul is 19x larger. Same wallet population, same pools, different days.

It's Not About More Trades

Wallets aren't opening way more positions on Friday than Saturday. The median is 6 vs 4 — Saturday is only 33% lighter on opportunity count. So why is the PnL gap so much bigger?

The quality of each position:

| Day | Med PnL/position | Med fee/hour held | Top 10% wallet day | |---|---|---|---| | Sunday | $0.055 | $0.18/hr | $71 | | Monday | $0.043 | $0.11/hr | $58 | | Tuesday | $0.032 | $0.15/hr | $68 | | Wednesday | $0.052 | $0.13/hr | $74 | | Thursday | $0.053 | $0.11/hr | $78 | | Friday | $0.121 | $0.28/hr | $103 | | Saturday | $0.015 | $0.15/hr | $35 |

Friday's median position pays $0.12. Saturday's pays $0.015. That's an 8x ratio per trade. Same wallet, same pool universe, just running positions on different days.

And the ceiling is dramatically lower: the top 10% of Friday wallets had a $103 day. The top 10% of Saturday wallets capped out at $35. The day's best haul on Saturday is roughly a third of what Friday delivers — even for the lucky few.

It's not that Saturday LPers can't win. 1,151 wallets had a profitable Saturday — 53% of those active. The problem is the wins are smaller, the misses are heavier, and the day's best haul is a third of what Friday delivers.

All Days Have Winners

Even on the worst days, plenty of memecoins were profitable to LP. The day-of-week effect is about average outcomes, not a hard rule.

Share of memecoins with positive median PnL per day (min 30 positions):

| Day | Tokens with sample | Profitable | % profitable | |---|---|---|---| | Thursday | 37 | 31 | 84% | | Friday | 38 | 31 | 82% | | Sunday | 38 | 28 | 74% | | Monday | 39 | 28 | 72% | | Tuesday | 40 | 26 | 65% | | Saturday | 40 | 21 | 53% | | Wednesday | 38 | 20 | 53% |

Saturday and Wednesday tie for worst at 53% — but even on those days, more than half of memecoins had positive median PnL. Picking the right token still matters.

Some of the memecoins from this 28-day window — most memecoins are short-lived, so each day had a different mix of hot tokens. Here are some of the ones that drove the most activity across the full window:

| Memecoin | Positions | Wallets | Median PnL/pos | Win rate | |---|---|---|---|---| | BURNIE | 17,819 | 1,989 | +$0.06 | 60% | | Iroha | 7,007 | 1,085 | +$0.06 | 61% | | ASTEROID | 6,255 | 895 | +$0.66 | 72% | | ALONSHOUSE | 5,581 | 888 | +$0.13 | 67% | | ROCKET | 4,679 | 859 | +$0.08 | 61% | | PEPE | 4,411 | 783 | +$0.12 | 63% | | DUMBMONEY | 4,401 | 678 | +$0.19 | 66% | | PIXEL | 3,298 | 644 | +$0.09 | 60% | | TripleT | 3,151 | 636 | +$0.15 | 59% | | 🪑 | 2,567 | 622 | +$0.13 | 63% |

Most delivered positive median per-position PnL with 60-70% win rates. The day-of-week effect is about aggregate outcomes — quieter days mean smaller per-position wins, more IL, fewer profitable wallets in absolute terms. It does NOT mean any specific memecoin was universally bad on Saturday or universally good on Friday. Picking the right token still matters more than picking the right day.

What About Top-Scored Wallets?

Group C — TrackLP's own list of high-quality LPers (588 wallets with quality_score ≥ 45). These are the wallets we'd flag as worth watching or copying. If our day-of-week finding is real, the elite should show the same pattern:

| Day | Wallets active | Med positions | Med fees | Med PnL | Daily WR | |---|---|---|---|---|---| | Sunday | 294 | 7 | $55.71 | $18.32 | 73.5% | | Monday | 268 | 5 | $29.29 | $9.22 | 68.7% | | Tuesday | 305 | 7 | $47.75 | $7.63 | 63.9% | | Wednesday | 322 | 7 | $31.49 | $12.47 | 72.7% | | Thursday | 320 | 7 | $32.34 | $18.44 | 76.6% | | Friday | 353 | 9 | $52.09 | $23.56 | 81.6% | | Saturday | 269 | 6 | $29.94 | $6.06 | 66.9% |

Top-scored wallets earn 30-100x more than the broader cohort (Friday median $23.56 PnL on 9 positions). But the day-of-week pattern is identical. Friday is their best day too. Saturday is their worst. The skill premium shows up as bigger wins, not as immunity to the day effect. Even elite wallets self-select away from Saturday — only 269 active vs 353 on Friday.

Three independently-selected wallet groups all point to the same answer: Friday best, Saturday worst, Sunday strong.

Why Friday?

Speculation now, not data. Friday's volume is high but not the highest — yet wallets earn the most. Best guess: Friday's volume is concentrated. Weekly close pumps, weekend positioning, retail logging in for the day, NY market close. The price action is pointed, not diffuse. LPs collect fees on directional moves, not random churn. That's why Friday's fee yield ($0.28/hr) is roughly double Tuesday's ($0.15/hr) despite Tuesday having more total volume.

Saturday inherits the worst of both worlds: enough liquidity is still deployed that competition for fees is high, but trader attention drops sharply (28% volume drop vs Friday). You're paying the IL cost of holding through a quiet day without the fee income to compensate.

Sunday recovers — Asian markets wake up, Monday positioning starts. The "weekends are dead" narrative ends up wrong because it's a Saturday narrative being applied to two different days.

What This Doesn't Tell You

It cannot tell you Friday will be the best day next week. The sample is one 28-day window. Different months, different SOL prices, different memecoin cycles will give different numbers.

It cannot tell you the day-of-week effect is universal. This is trending memecoin LPing on Meteora DLMM — memecoins that broke into the daily top-10 by SOL volume. It does NOT tell you about LPing established memecoins (BONK, WIF, POPCAT), SOL/USDC, stables, or major-token pairs.

It cannot tell you why Saturday is bad. The fee yield numbers point at "less concentrated price action" but I haven't quantified the price-action side. Could be retail, could be marketmakers, could be something else.

What it can tell you: in a recent 28-day window, on memecoin pools that captured trader attention, active LPers performed measurably better on Fridays and worse on Saturdays. Three independently-selected wallet groups all agree. That's enough to inform deployment decisions, not enough to bet the farm on.

How to Use This

• Friday is a deploy day. Median wallet earns $2.48 PnL + $10.47 fees on 6 positions. 69% chance of profitable day. Highest fee velocity.
• Sunday is underrated. Second-best PnL day at $1.55 median. Don't let "weekends are dead" stop you.
• Saturday is a skip day for most. $0.13 median PnL. 53% daily win rate. Top-10% day caps at $35 vs Friday's $103.
• If you can pick winners, every day works. Even Saturday has memecoins above 60% win rate. Top-scored LPers profit on every day.

If you only LP five days a week, the data points at Saturday as the day to take off — not Sunday.

Caveats

Single-month sample. UTC day attribution. Memecoin pools only. Closed positions only. Bot threshold at 50/day is heuristic. Sample bias toward wallets active on Meteora during the window. All numbers are descriptive of April 1 to April 28, 2026.

If anyone wants to cross-check the methodology or run their own queries, the entire pipeline is repeatable: Dune queries on lb_clmm_evt_Swap and lb_clmm_evt_PositionCreate, LP Agent for position-level data, wallet-level median aggregation. Same recipe, different month → different numbers, but the methodology should hold.

All data from tracklp.com using LP Agent and Dune Analytics.