YieldMeasurement · 2026-09-26
What safety actually costs
We keep a daily rate history across 43 venues. That lets us publish something a rate table cannot: not what a number is today, but what it is compensating you for. Every figure below is recomputed from our own data on each build.
The price of a weaker counterparty
Measured by YieldScope on 2026-09-26: across the venues we track, the 6 passing four or more of five public safety checks pay an average 3.38 % on withdraw-anytime stablecoins, while the 13 passing two or fewer pay 7.51 %.
| Checks passed | Venues | Mean | Median |
|---|---|---|---|
| 4-5 | 6 | 3.38 % | 4.0 % |
| 3 | 7 | 3.5 % | 3.0 % |
| 0-2 | 13 | 7.51 % | 6.0 % |
Medians are shown beside the means on purpose. The mean of the weaker group is pulled up by two venues paying double digits; the median gap is much smaller. Quoting only the mean would be the same arithmetic we criticise elsewhere. Sample sizes are small — treat this as a direction, not a constant.
What a lock-up buys
Measured by YieldScope on 2026-09-26: across 9 cases where one venue offers both a locked and a withdraw-anytime product for the same coin, the locked rate is a median 3x higher.
| Venue | Coin | Withdraw anytime | Locked | |
|---|---|---|---|---|
| WhiteBIT | BTC | 1.38 % | 10.4 % | x7.53 |
| WhiteBIT | ETH | 1.48 % | 10.4 % | x7.02 |
| WhiteBIT | USDT | 1.94 % | 11.1 % | x5.73 |
| WhiteBIT | USDC | 2.06 % | 10.4 % | x5.04 |
| Nebeus | USDC | 4.0 % | 12 % | x3 |
| Lune.fi | USDC | 18 % | 29.5 % | x1.64 |
| Lune.fi | USDT | 18 % | 29.5 % | x1.64 |
| Lune.fi | BTC | 14 % | 21.5 % | x1.54 |
This is the single most common way a rate table misleads: the big number belongs to a product that holds your money, the small one to the product you can leave. Both sit in the same column on most comparison sites.
What a bank advertises against what you get
Measured by YieldScope on 2026-09-26: across 44 savings products in 16 countries where the bank publishes both a headline rate and conditions an ordinary customer can actually meet, the advertised figure is a median 1.27x the achievable one — and at the extreme, Standard Chartered Hong Kong advertises 3.5% on an account whose base rate is 0.001%.
| Bank | Country | Achievable | Advertised | |
|---|---|---|---|---|
| Standard Chartered Hong Kong | Hong Kong | 0.0 % | 3.5 % | x3500 |
| ING Australia | australia | 0.01 % | 5.5 % | x550 |
| Trust Bank | Singapore | 0.05 % | 2.4 % | x48 |
| ZA Bank | Hong Kong | 0.5 % | 18 % | x36 |
| N26 | Germany | 0.25 % | 1.5 % | x6 |
| ING Deutschland | Germany | 0.75 % | 3.75 % | x5 |
| Alior Bank | poland | 2.0 % | 6.5 % | x3.25 |
| ПСБ (Промсвязьбанк) | Russia | 5.0 % | 16 % | x3.2 |
The median matters more than the extreme here. Half of these products advertise less than 1.27x what you get, which is honest by the standards of this industry. The tail is what does the damage: four banks stretch past 6x, and the widest is a factor of 3,500. Records we could not verify on the bank’s own product page are excluded entirely — an unconfirmed number cannot be measured.
How this is measured
Base, withdraw-anytime rates only, taken from each venue’s own page and refreshed daily since 2026-05-25. Safety checks are regulation, proof of reserves, insurance fund, withdrawal terms and track record — five public, binary criteria, sources linked on each venue page. Nothing on this page is typed by hand; it recomputes on every deploy, so it cannot drift from the data. Rates are variable snapshots, not promises. Not financial advice.