Why 400ms Settlement Changes What Money Can Do
Sub-second finality isn't a faster version of the same thing. It removes the reason most payment infrastructure exists, and makes new kinds of payment viable.
“Faster payments” sounds like an incremental pitch the same thing, sooner. It isn’t. A great deal of financial infrastructure exists purely to paper over the gap between when a payment appears to happen and when it actually does. Close that gap and the infrastructure has nothing to do.
Authorisation vs settlement: the gap nobody mentions
When a card is tapped, two separate things happen, and only one of them is fast.
Authorisation takes a second or two. The issuer confirms the card is valid and the funds are available, and places a hold. This is the part the customer sees, and it is the source of the belief that card payments are instant.
Settlement Money actually arriving in the merchant’s account takes one to three business days, sometimes longer. “Business days” is doing real work in that sentence: a Friday evening sale may not settle until Tuesday.
In between sits a period where the payment has visibly occurred, the goods are gone, and the money is nowhere. That interval is the origin of an enormous amount of financial machinery.
What the settlement gap costs a merchant
Float. Money in transit belongs to nobody in particular and earns interest for whoever holds it. At scale, across an economy, this is a substantial and largely invisible transfer from merchants to intermediaries.
Working capital. A business that pays suppliers today from sales that arrive Tuesday needs a buffer. That buffer is financed through overdrafts, credit lines, or merchant cash advances at punitive rates. An entire lending industry exists to fill a hole created by slow settlement.
Reconciliation. Because settlement is batched and netted, the deposit that lands is not the sum of the day’s sales. It is a batch, minus fees, minus refunds, minus chargebacks, possibly spanning several days. Reconciling that against the till is a real job that real people do, every day, in every business.
Reversibility risk. Because the transaction isn’t final, it can be undone. A chargeback can arrive months later. The merchant has shipped the goods and may lose both the goods and the money, plus a fee.
What ~400ms actually means
Solana targets a block time of roughly 400 milliseconds. In practice a transaction is confirmed in about a second under normal conditions comparable to a card authorisation, and fast enough that a customer at a till doesn’t perceive a wait.
The important part is what that confirmation is. It is not a promise that money will move later. The money has moved. There is no batch, no netting, no intermediary holding a balance in transit, and no clearing window.
The gap closes, and the machinery has nothing to do
Take the four costs above and apply instant final settlement.
Float has nowhere to accumulate there is no interval. Working-capital financing to bridge the delay is unnecessary, because there’s no delay. Reconciliation becomes trivial: each payment is a discrete, timestamped, verifiable event, and the merchant’s balance is simply the sum of them. Chargeback exposure disappears, because settlement is final.
That last one is genuinely double-edged and deserves saying plainly: chargebacks protect consumers. Removing them shifts risk from merchant to buyer, which is why card networks built them. A payments system without reversibility needs some other answer for “the goods never arrived” escrow, reputation, insurance. “It’s final” is a real advantage for the merchant and a real loss for the customer, and pretending otherwise is how crypto payments arguments lose credibility.
What this looks like for a merchantA flat $0.01 per transaction, settling in about 400ms directly to a wallet you own no, acquirer, no batch, no waiting for Tuesday. DYNK is pre-launch; this is what we’re building towards.
→The payments that couldn’t exist before
Speed changes existing payments. The combination of speed and a flat sub-cent fee creates categories that were previously arithmetically impossible.
Micropayments. Charging 5c to read an article has never worked, because processing it cost more than 5c. Every attempt therefore ended up as a subscription or as advertising not because those were better models, but because the payment rail imposed a minimum viable transaction size. Remove the floor and per-item pricing becomes possible again.
Streaming payments. If settlement is sub-second and sub-cent, money can move continuously rather than in lumps. Wages accruing per hour worked. Rent by the day. A cloud service billed by the second and paid by the second. Monthly billing cycles are an artefact of expensive settlement, not something anyone chose.
Machine-to-machine payments. Software agents transacting with each other need payments that clear without a human, without an account-opening process, and at a cost proportional to a very small unit of value. A percentage-fee, days-long, identity-gated rail cannot do this. A flat-fee, sub-second, permissionless one can and this is precisely the substrate that agentic software will need as it starts buying compute, data and services on its owner’s behalf.
Genuinely instant cross-border transfer. The domestic/international distinction is a property of correspondent banking, not of value. On a public chain there are no borders to cross, so a transfer to the next street and a transfer to another continent are the same operation at the same price.
What fast settlement doesn’t fix
It would be easy to overclaim here, so:
Fast settlement does nothing about price volatility arriving in a second is no comfort if the asset moves 5% before you convert it. It doesn’t solve on- and off-ramps, which remain the slow, licensed, expensive part of any real-world flow. It doesn’t address consumer protection, as discussed. It doesn’t by itself make self-custody usable for people who can’t manage keys. And it doesn’t remove regulatory obligation for anyone operating a business at scale.
Settlement speed is infrastructure, and infrastructure is necessary rather than sufficient. What it does is remove a constraint that has quietly shaped what payments can be for fifty years, long enough that the constraint stopped looking like a constraint and started looking like the nature of money.
It isn’t. It’s a design decision from an era of batch processing and correspondent banks, and it has an expiry date.
Frequently asked questions
How fast is a Solana transaction?
Solana targets a block time of roughly 400 milliseconds, and a transaction is typically confirmed within about a second in normal conditions. Full finality, the point at which a block is effectively irreversible takes longer, conventionally around 12 to 13 seconds, though most consumer payments are treated as settled well before that.
What is the difference between confirmation and finality?
Confirmation means a validator has included your transaction in a block. Finality means enough of the network has built on that block that reversing it is economically infeasible. Card networks blur this distinction by authorising instantly and settling days later; blockchains separate the two explicitly, which is clearer but often confuses newcomers.
How much does a Solana transaction cost?
A base Solana transaction fee is 5,000 lamports, which is 0.000005 SOL a fraction of a US cent at typical prices. Congested periods can add priority fees. The important structural property is that the fee is denominated per transaction rather than as a percentage of the amount transferred.
Is Solana fast enough for point-of-sale payments?
For practical purposes, yes. A sub-second confirmation is within the window a customer at a till will tolerate, and comparable to a card terminal's authorisation. The remaining obstacles to point-of-sale crypto payments are user experience, price stability and merchant tooling rather than settlement speed.