The short version
- 1
The liquidity premium is typically one to three percentage points for a 30 to 90-day commitment. We think that is thin compensation.
- 2
A locked deposit carries identical counterparty risk to a flexible one — you simply lose the ability to act on warning signs.
- 3
On some platforms a flexible rate matches another platform's locked rate. A flat 4% flexible beats a locked 3.5% with no argument required.
- 4
Locked genuinely makes sense for money with a known future date that you would not have touched anyway — a tax bill, a planned purchase.
What the premium actually is
Every serious platform offers both. Flexible pays less and lets you leave; locked pays more and does not. Across the venues we track, the gap on stablecoins runs from about one to three percentage points for terms of 30 to 90 days.
Bitget reaches around 11% on fixed-term stablecoins while its flexible products pay low single digits — one of the widest spreads available. OKX pays roughly 5% to 8% on fixed terms against about 2.6% flexible. Nexo offers fixed terms up to twelve months at a premium over its flexible rate.
The standard framing is that this is a liquidity premium, and that framing is correct as far as it goes. You are being paid to give up access. The question this page exists to answer is whether you are being paid enough.
- 1–3 pts
- Typical premium
- ~0.5%
- What 2 points is worth
- Identical
- Counterparty risk
- Nov 2022
- When it mattered
For a 30 to 90-day commitment
On a 90-day term, in absolute terms
Locked and flexible carry the same
$940m frozen with no warning
Side by side
| Property | Flexible | Locked |
|---|---|---|
| Withdraw at any time | Yes | No |
| Rate fixed for the period | No | Yes |
| Typical stablecoin rate | 2.6–8.5% | 5–11% |
| Interest paid | Daily | At term end |
| Early exit penalty | none | Interest forfeited |
| Counterparty risk | Yes | Yes |
| Can act on bad news | Yes | No |
| Rate carries to renewal | No | No |
| Income is forecastable | No | Yes |
Pricing the option you are selling
Here is the way we think about it, and it reframes the decision entirely.
When you lock a deposit for ninety days, you are selling the platform an option. Specifically, you are giving up the right to withdraw at any point during that window, and you are being paid a premium for it. The question is whether the premium matches what the option is worth.
Two percentage points annualised, over a ninety-day term, is about half a percentage point in absolute terms. On a $10,000 deposit, that is $50. So the question becomes: is $50 adequate compensation for being unable to remove $10,000 from a centralised crypto platform for three months, no matter what happens?
Put that way, most people answer no. And the reason is that the scenario in which you would most want to withdraw — the platform showing warning signs, withdrawals slowing, rates being cut abruptly — is exactly the scenario in which the lock binds.
What the premium buys, in absolute terms
- $10,000, 90-day term, +2 points
- $50
- $10,000, 90-day term, +3 points
- $75
- $10,000, 12-month term, +2 points
- $200
- $50,000, 90-day term, +2 points
- $250
- What you give up
- The ability to leave
Annualised percentage points look larger than the absolute sums they produce over a short term. Convert before deciding.
Lock-ups you did not choose
A distinction worth making, because the two get conflated and they are not the same thing.
A lock-up is a platform's commercial restriction on a savings product. It exists because the platform wants stable funding and is willing to pay for it. It is negotiable in principle and varies between competitors.
An unbonding period is a blockchain protocol's own rule on staked assets. Cosmos imposes 21 days, Polkadot 28. No platform can remove these at the protocol level, because they exist to prevent validators misbehaving and immediately exiting.
What platforms can do is absorb the mismatch. A venue that maintains a liquidity buffer can let you withdraw staked DOT immediately while it waits out the 28 days itself. That costs the business money, which is why such venues typically publish lower staking rates — 6% on DOT with instant withdrawal against roughly 10% to 13% nominating directly with the full unbonding period.
That is a locked-versus-flexible trade too, and the same arithmetic applies. Our Polkadot page works it through in detail.
| Restriction | Set by | Removable | Example |
|---|---|---|---|
| Savings lock-up | Platform | Yes | 30–90 day term |
| Unbonding period | Protocol | No | Cosmos 21 days |
| Exit queue | Protocol | Partial | Ethereum, variable |
| Withdrawal suspension | Platform | No | In every set of terms |
When locked genuinely makes sense
We are not arguing that locked products are always wrong. There is a clear case where they are the right answer.
Money with a known future date. If you are setting aside funds for a tax bill due in April, a planned purchase, or any obligation with a fixed timing, and you genuinely would not have touched that money regardless — then the lock costs you nothing you were going to use. You are being paid a premium for a constraint you had already imposed on yourself. Take it.
When the premium is genuinely large. A three-point spread on a 30-day term is a different proposition from a one-point spread on twelve months. Short terms with wide spreads are the good end of this market.
As part of a ladder. If you are going to use locked products for a meaningful balance, splitting across several maturities means a portion becomes available at regular intervals rather than everything at one date. This is ordinary fixed-income practice and it applies directly.
What does not make sense is locking your primary liquid reserve, or locking for twelve months to capture one or two extra points at a platform you have not thoroughly evaluated.
Where we would start
Sometimes the flexible rate is simply the better rate
A flat 4% flexible on USDC and USDT with daily accrual beats several platforms' fixed-term offerings outright — no lock, no premium to evaluate, and the whole balance available at any moment.
- 4% flexible, no term commitment
- Interest accrues daily from day one
- No minimum transfer
- FCA registered, Gibraltar FSC DLT licence FSC0686FSA
Our position
Default to flexible. The premium on offer across this market is usually smaller than the value of being able to leave, particularly at centralised platforms where the catastrophic scenario is a withdrawal suspension you cannot see coming.
Use locked for money that already has a date attached, and prefer short terms with wide spreads over long terms with narrow ones. Ladder if the balance is meaningful.
And before accepting any locked product, run the arithmetic in absolute terms rather than percentage points. Two points annualised on ninety days is fifty dollars per ten thousand. Once you see the number rather than the rate, the decision usually makes itself.
One final check worth doing: compare the locked rate you are being offered against the best flexible rate available elsewhere. Several platforms' fixed-term products pay less than another venue's flexible account, which makes the whole liquidity question moot.
Flexible versus locked: common questions
How much more do locked crypto savings pay?
Can I withdraw early from a locked product?
Is a locked savings account safer than flexible?
What is the difference between a lock-up and an unbonding period?
Should I ladder locked deposits?
Do fixed rates stay fixed after the term ends?
Sources and further reading
Rates, terms and regulatory details on this page were checked against the following sources on . Variable figures move constantly — always confirm with the provider before depositing.
- 01 Bitget — Earn products — fixed-term rates and caps
- 02 OKX — Earn — flexible versus fixed-term structure
- 03 Nexo — Earn Crypto — fixed terms up to twelve months
- 04 SEC — Genesis and Gemini Earn charges — the withdrawal suspension that froze $940m