Sell and rebuy, or just move the coin?

The question itself is simple. You hold something on platform A and want it on platform B. Two ways to do it.

The two routes, side by side

On cost alone route one wins easily. But cost is not the only variable: what actually decides it is how the price moves during that gap, which you do not control, and whether the coin can still be withdrawn at all, which sometimes you also do not control.

Getting the arithmetic straight first

Route one's cost is knowable: one withdrawal fee, and a small one if you pick the chain well.

Route two has four components, two of which get overlooked:

1. The selling fee, at your spot rate;
2. The spread on selling — a market order eats the book, and thin pairs eat it noticeably;
3. The withdrawal fee on the stablecoin: usually cheaper than the original coin, because stables have more chains available;
4. The fee and spread on buying back, items one and two again.

Two and four are the important ones. Trading fees are proportional and predictable; spreads are neither. On a liquid major the spread rounds to nothing. On a pair with thin daily volume, one moderately sized market order can push the price several percent, and doing that twice is a real loss.

When moving the coin is clearly right

Three situations where route one wins more or less always:

You intend to hold this asset long term

Since you are keeping it either way, selling and rebuying only adds two trading fees and two spreads, plus a window where you might not get back in. A transfer should ideally leave your position unchanged.

The coin is not especially liquid

Spread costs swallow route two's theoretical advantage whole. A thin book in a violent market can be genuinely bad, and a market sell can cost several percent at the moment of execution.

You would rather not have to be right twice

Route two asks you to sell at a decent moment and buy at a decent moment. That converts a logistics task into two timing decisions, during a period when you are already tense, and tense timing decisions tend to go badly. My preference is explicit: decisions I can avoid making, I avoid.

When converting first is clearly right

Equally, several situations favour route two:

The coin is being delisted and you did not want it anyway

The archetypal case. A delisted asset may have no trading pair at the destination, so moving it there leaves you holding something you cannot do anything with. Selling while A still trades it and moving something universally accepted is a far smoother path. And there is a clock: trading usually stops well before withdrawals do, so past that line you can move it but not sell it.

The destination does not list the coin

A hard constraint. Check B's deposit page; if the coin is not there, route one does not exist.

There is no chain both ends support

A supports chain X only, B supports chain Y only — this happens. Options are a third platform as a waypoint, a bridge (more cost, more risk), or converting to a stablecoin. Converting is usually the least painful of the three. How to check for a common chain is in which chain to withdraw on.

You were going to trim the position anyway

Then do it here. A migration is a natural moment to look at holdings, and some of them you probably stopped wanting a while ago and never got round to dealing with.

Putting both routes side by side

The figures below are purely illustrative, not any platform's actual rates, and exist only to show the shape of the cost. Say you are moving ten thousand dollars of something.

Put differently: if the book is deep and the spread rounds to nothing, the gap between routes is basically two trading fees. If the book is thin, the spread alone can exceed everything else combined. Which means the question was never "which route has lower fees": it was "how easy is this thing to sell".

One precondition people forget: does the pair exist

Route two has a prerequisite that gets assumed: you need to buy the coin back at the destination, which requires the pair to exist there.

For majors on large platforms, fine. But for something relatively obscure you can end up in this position: the new platform does list it, but with a single pair and very little depth. Now the buy-back spread may be worse than the sell-side spread was at the old platform, and you have paid twice.

So before committing to route two, look at two things on the destination:

• Whether the pair you want exists at all, usually against a stablecoin;
• What the book looks like — distance between best bid and ask, and size at each level.

If the depth is not there, the arithmetic changes. In that situation the more sensible answer might be: sell at the old platform, move the stablecoin, and simply not buy back. If the asset trades poorly everywhere, that is itself information about whether to keep holding it.

A third option: split it

These are not mutually exclusive. What I actually do most often is mix:

• Being delisted, illiquid, or something I wanted to reduce anyway, sell into stables at A;
• Long-term holdings, liquid, still wanted, move the coin as-is;
• Dust — not worth handling individually; sweep it with the platform's small-balance conversion.

The advantage is that each position gets treated according to what it is, rather than one rule applied to everything. The cost is more operations and more withdrawal fees. Worth doing when you hold many assets; unnecessary with two or three.

If you genuinely cannot decide, halve it

There is an option that does not require you to be right: split the position and send half each way.

That sounds like avoiding the decision, and it has a real justification:

• You no longer need a view on short-term direction, because you hold both sides of it;
• If the price rises during the gap, half of it was held; if it falls, half was in stables;
• It is psychologically easier to execute, because there is no "what if I chose wrong" to sit with.

The cost lands between the two routes and there are more operations. But for most people a second-best plan they actually finish beats an optimal one they stall on until the deadline.

A detail about ordering

Sell first, move second. If part of the position is going to be sold, sell it while A is still trading normally rather than thinking "move what I can now and deal with the rest later". Later frequently means after trading has stopped, at which point you are holding something you can move but not sell, and your options have halved.

If the destination is not settled yet, opening the account takes time too, verification can stall for days. We walked that flow and noted where it sticks in opening a Binance account, step by step.

How to think about the gap

Route two has an unavoidable feature: between selling and buying back, you do not hold the asset. That might be twenty minutes or several hours.

Nobody knows what the price does in that window. So I have only two not-very-clever but practical positions on it:

1. Do not treat the gap as a timing opportunity. "Sell now, buy back after it drops" sounds reasonable and turns a migration into a short-term trade. If you genuinely can call short-term direction, you do not need a migration as the vehicle for it.
2. Compress the gap. Have B's account open, verification cleared and the deposit page loaded before you sell; sell, move, buy back immediately. The shorter the window, the less of this uncertainty you are carrying.

Which stablecoin, if you go that way

A minor decision, with two pieces of experience worth passing on:

• Pick the one with the most pairs at the destination. You will use it to buy the original asset back; if that pair does not exist you have to convert again and pay for another layer.
• Pick the one with the most chains available. Part of the value of a stablecoin is that the roads are good — the same asset withdrawable over five networks means you can always find a cheap one.

On the risks of stablecoins themselves I am not going to expand here, because it is a separate subject. One reminder only: a stablecoin is not cash, it is a liability of its issuer, and there have been episodes of visible short-term deviation from par. Holding one for a few hours mid-migration makes that risk negligible; planning to sit in one long-term is a different decision and deserves its own thinking.

A checklist you can just follow

Compressing all of the above into questions, answered in order. When you run out of questions you have your answer:

1. Does the destination list this coin?

   No, route two. Hard constraint, and the remaining questions do not apply.

2. Is there a chain both ends support?

   No, route two, or pick a different destination. Do not go looking for a bridge.

3. Can it still be traded?

   Trading already stopped, route one only. The window for selling has closed.

4. Do you still want to hold it?

   Yes, route one. No, route two, and take the opportunity.

5. Is the book deep?

   Thin, lean route one. The spread will eat route two's entire theoretical advantage.

6. Still unsure?

   Split it in half, one each way. Not avoidance: the least stressful handling when you genuinely cannot call it.

The first three are facts you can look up; only the last three involve your own judgement. Answer the factual ones first — very often it is settled by question one or two and the agonising never becomes necessary.

Do not rebuild the portfolio immediately

Whichever route you take, once the assets land there is a temptation: since everything is being reassembled anyway, why not adjust the allocation while you are here.

The instinct is fine; the timing is wrong. Migrating puts you in a "I have just been through something annoying" state, and allocation decisions made in that state tend to look extreme in hindsight: either over-cautious, selling things that should have been kept, or over-compensating on the grounds that since you went to the trouble you may as well add.

What I do is land it as-is, leave it a week or two, and revisit once the episode has receded. The safety of the assets and the shape of the portfolio are two problems; solve one at a time. The safety problem was solved the moment the transfer confirmed. The allocation problem is not urgent.

One more layer: the records

Route two produces two more trade records than route one. In some jurisdictions the act of selling can create a reporting obligation where a plain transfer does not. Rules vary enormously and I cannot give you a specific answer, only a flag: if you are somewhere with crypto reporting requirements, the difference between these two routes is substantive and worth confirming before you act.

Either way, keep the trade records, the withdrawal references and the transaction hashes. What to keep and for how long is in which transfer records are worth keeping.

Specifically to these two routes, what needs keeping differs:

• Route one produces a single withdrawal record; the hash and the reference are enough;
• Route two produces two trade records plus a withdrawal record, and all three matter. The price and timestamp on the trades are the important part, if you later need to establish the cost basis of this holding, that is what does it.

An operational note: export functions usually have a date-range limit, with some platforms allowing only three months or a year at a time. Export the relevant period right after the move rather than going back months later — particularly when the reason you moved was that the platform looked unsteady, which makes future access itself uncertain.