Recommended settings, and how to get the most out of Whale Tail
Everything on this page comes from 11,464 positions that Whale Tail members have actually closed — not from theory, and not from what a whale earned for itself. Where we have tested something and it did not work, that is here too.
The short version
If you change nothing else, change these two.
| Setting | Use | Why |
|---|---|---|
| Stop loss | 15% | Caps what any single position can take back from the ones that worked. The biggest lever on this page. Now on by default. |
| Token-only entries | Off | They won only one time in four, and cost noticeably more to get in and out of than a SOL-side entry. Now off by default. |
| Position size | Start small | Your first positions on a new whale are buying information. Buy it cheaply. |
| Number of whales | 3–5 | Enough that one bad run does not define your month, few enough that you can see what each one is doing. |
Both defaults changed on 12 September and were applied to existing accounts. You can put either back from the settings panel on any whale you follow — we changed what happens by default, we did not take the choice away.
A loss floor is the setting that protects your upside
A floor protects the gains you have already made.
Liquidity provision is lopsided by nature. Most positions behave: they earn fees, the price stays near your range, and they close somewhere sensible. Then occasionally a token falls through the floor of its range and keeps going, and that one position can take back what a long run of good ones earned. That is true of DLMM anywhere, on any platform.
We looked hard for a way to spot those positions at the moment we entered them, because a filter that avoids one beforehand beats a stop that catches it after. We could not find one. They were not in identifiably bad pools. They did not come from identifiably bad whales — one of our strongest contributors also produces some of them. They were not held for an unusual length of time.
So the answer is not to predict them. It is to decide, in advance, the most you are willing to lose on any single position. A floor needs no foresight at all, and it never touches a position that is doing well.
| Floor | Positions it would touch | Left to run | Losses it would have prevented |
|---|---|---|---|
| −15% | 328 (3.4%) | 96.6% | about 65 SOL |
| −20% | 244 (2.5%) | 97.5% | about 48 SOL |
| −25% | 177 (1.8%) | 98.2% | about 36 SOL |
Being straight about that table: it is an upper bound. Replaying history cannot see a position that dipped below the floor and then recovered — those would have been cut short, and some of them would have come back. So we also ran the floor live, in shadow, before switching it on. Over the breaches that have closed so far, stopping out came out roughly 1.3 SOL ahead of holding on. Roughly half did recover; the ones that did not fell much further, which is why the floor still wins overall.
Why 15 and not 25
Three separate things landed on the same number: our own replay, our live shadow test, and the published guidance of the two biggest competitors in this space. When an argument from history, an experiment on live money and two rivals all pick 15, we stop arguing.
Getting your size right
There are two ways to size a copy, and the second one has a trap in it.
Fixed size
The same amount into every position, whatever the whale does. Simple, predictable, and the right choice until you know a whale well.
Scale with the whale
Copy a proportion of whatever they commit, so their conviction carries through to you. Your Position size then becomes a ceiling rather than a target.
The ratio is a multiplier, not a percentage. This catches people out, including us.
A ratio of 1 means one-for-one with the whale. A ratio of 0.1 means a
tenth of what they commit. A ratio of 100 does not mean 100% — it means a hundred
times, and every entry will be refused because you cannot afford it.
How to pick a ratio
your ratio = the SOL you have ÷ the SOL that whale runs
If you hold 5 SOL and the whale works with 100, your ratio is 0.05. Get this wrong in the other direction and you take the same losses they do while missing their bigger winners, because your ceiling caps the good trades and lets the bad ones through in full.
A quick way to tell your ratio is too high: if the whale is down 5% on a position and you are down 30% on the same one, your ratio is wrong — not the whale.
Choosing whales: one number matters
Every whale on our board shows what copying them has returned to Whale Tail members. That is deliberately not the same as what the whale earned for itself, and the gap between those two is the entire job.
One wallet in our history earned +337 SOL for itself while returning the people copying it almost nothing. Another was ranked the single best liquidity provider on a competitor's leaderboard — 99% wins, +1,714 SOL of its own profit — and was among the weakest whales we have ever measured for the people following it. Both look magnificent on any board that shows the whale's own number.
So when you look at a whale, look at the returned to members column and the number of closed positions behind it. A wonderful percentage over nine closes is not evidence. We treat about 40 closed positions as the point where a record starts to mean something.
New whales, and why we add them anyway
A whale with no history is not a bad whale — it is an unmeasured one, and somebody has to go first or no whale ever gets a record. When our automatic setup adds an unproven whale it does so in one slot, at half size, precisely so that finding out costs as little as possible. If you add one yourself, do the same.
Whales that have cost members money over a real sample get retired: they stop opening new positions, anything you already hold still exits normally, and they come off the board so nobody new picks them up. If you are following one when that happens, we message you and tell you what it did for you specifically — which is sometimes better than what it did for everyone.
Token filters: useful, and we will not oversell them
Every whale you follow has its own filters. The engine checks them before it commits any money, and it fails closed — if a token cannot be read at all, we skip rather than guess.
| Filter | A reasonable start | What it does |
|---|---|---|
| Min market cap | $500k – $2M | Keeps you out of tokens too small to exit without moving the price. |
| Min liquidity | $50k+ | The same idea, measured directly. Thin pools cost you on the way out, not the way in. |
| Jupiter organic score | 50+ | Jupiter's read on whether trading in a token is real. |
One we tried and removed
We ran a minimum token age of 2 hours for one night. It blocked entries at 89, 96 and 103 minutes — short of the line by under half an hour — and the positions it blocked did better than the ones it allowed. Early tokens are where a lot of the volatility lives, and the volatility is the point. It is off, and there is no age floor by default.
What we can honestly claim
These filters work — they are enforced on every entry and we have watched them refuse trades. What we cannot yet tell you is exactly where to set them, and we would rather say so than invent a number.
The reason is subtle and worth knowing: market cap and organic score are live figures. A token that has since collapsed reads near zero on both today, whatever it read when we entered. So checking history tells you that losing tokens look bad now — which proves nothing, because they look bad because they lost. As of 12 September we record these numbers at the moment we open, so in a few weeks we will be able to tell you where to set them from evidence rather than from folklore.
One thing we did measure: a score floor of 30 blocks almost nothing, because every live token we trade scores between 47 and 94. If you set one, set it meaningfully higher or it is decoration.
Speed, so you are not entering a move that has already happened
Whale Tail streams every whale transaction as it lands rather than checking on a timer, so a typical copy is about 1.4 seconds behind the whale. That matters most with whales who move quickly: you get in near the price they got, instead of into the move they already made.
Things that sound sensible and did not survive the data
Two pieces of advice you will see elsewhere that did not hold up when we checked them against our own positions. We mention them because acting on either would cost you something.
“Avoid whales that lots of people already copy”
The worry is that copiers trade against each other and the last one in gets the worst price. We checked it properly — same whale, same pool, same minute, first mover against everyone behind them — and found no penalty for arriving later. Our most-copied member actually does +0.49% better than whoever leads the batch.
Honest caveat: our busiest whale has around ten followers. This may simply be a problem that starts further up than we are. We will keep testing it as we grow, and we will tell you if it changes.
“Close positions after a set time”
A time limit interrupts about as much working profit as it saves. It cuts the good positions alongside the bad ones. A loss floor does not have that problem, which is why we recommend one and not the other.
Why your wallet and Meteora can disagree
Sooner or later you will see a position marked green on Meteora or a portfolio tracker while your wallet went down. Nothing is broken. They are measuring different things, and it is worth understanding which is which.
| What it measures | |
|---|---|
| Meteora and most trackers | The position only — from the moment liquidity went in to the moment it came out. |
| Your wallet | Everything, including buying the token to get in and selling it to get out. |
A liquidity position usually ends holding the token rather than SOL, so getting out means a swap — and a swap costs. We measured the cost of exiting every open position on the platform: it runs from 0% to 5.5% depending on how thin the pool is. That cost is entirely real, entirely yours, and entirely invisible to a tracker that stops counting when the liquidity is withdrawn.
The position figure is not wrong. It answers “how did this liquidity perform while it was deployed?” Your wallet answers “am I up?” When they disagree, your wallet is the one to believe, and Solscan will always settle it.
Two small things that quietly cost you
- Dead token dust. Positions sometimes leave behind a scrap of a token nobody will buy. It is worth almost nothing, but the account holding it is not.
- Empty position accounts. Every Meteora position locks about 0.052 SOL of rent, returned when the position closes properly. A position that closed badly can leave that rent stranded.
Neither is dramatic on one position. Across a few hundred it is real money sitting still, and it is recoverable.
If you only do three things
- Leave the 15% floor on. It is the only setting on this page with three independent pieces of evidence behind it.
- Size so that one bad position is boring. If a single loss would bother you, it is too big.
- Judge a whale on what it returned to members, over enough closes to mean something — never on what it made for itself.
Everything here is measured on our own closed positions and will change as we learn more. When it does, we will say so and tell you what changed.