Look at the number they want you to see. €0.99 per order. Less than a tram ticket in Frankfurt. Less than the espresso I bought this morning at the Hauptbahnhof while waiting for a train that never came.
Now look for the number they did not print. The spread.
I have spent nineteen years reading crypto price lists, and the rule has never changed: the number in bold is marketing, the number in the footnote is the business. Deka's price list for the Sparkassen crypto rollout contains exactly one fee and zero spread disclosures. That is not an oversight. That is the product design, written in the one place most customers never look.
The €0.99 order fee is an anchor, not a price. The real cost of trading crypto through a German savings bank is buried in a principal-trading spread the institution has explicitly reserved the right to adjust — unilaterally — whenever volume rises or volatility spikes.
That is the whole story. The rest is distribution.
Let me give you the lineage, because the timeline matters and most coverage is skipping it.
The DSGV — the umbrella body for Germany's Sparkassen — spent years as one of the loudest crypto skeptics in European retail finance. In 2022 the group's committees voted down crypto trading proposals outright. Speculative assets, they argued, did not belong inside a savings-bank mandate. The customer base — roughly fifty million people, deeply conservative, savings-account loyalists — was not the audience for this.
Then the ground shifted. MiCA came into force and handed custodians a legal rail. Börse Stuttgart Digital Custody picked up the first German MiCA license in January 2025. And by 2025 the DSGV reversed course and handed the build to Deka, its own securities house. By the autumn of 2026 — the price list I have been reading is dated September — five savings banks went live with crypto trading inside their existing banking app, under the dual supervision of MiCA and BaFin.
The assets: BTC, ETH, XRP, SOL, POL. The mechanics: request-for-quote execution, omnibus custody across four licensed providers, free custody, free deposits, and "free" withdrawals restricted to those same four providers. No minimum order. A €250,000 ceiling.
None of this is technically novel. That is precisely the point. This is not a blockchain story. It is a distribution story with a compliance wrapper, and the wrapper is the most expensive part.
Here is where the forensics start. I have audited enough custody and execution architecture to know the interesting decisions always live in the plumbing, never in the press release.
First: the execution model. Deka runs a request-for-quote system. You place an order, Deka returns a binding quote inside a limited time window, you accept or reject. That is it. This is not an exchange. There is no order book, no continuous price discovery, no visible counterparty. Deka is not matching your trade against another customer. Deka is the counterparty. When you buy, Deka sells. When you sell, Deka buys. This is principal trading — the same structure a prime broker uses — dressed up in a consumer app with a friendly green button.
And here is the thing about principal trading: your cost is whatever the principal decides it is. The quote bundles a "market reference price" with Deka's own markup. You see the total. You cannot decompose it. So you cannot calculate your own transaction cost. Compare that to any public exchange where the spread is printed on screen in real time, and the transparency gap is not marginal. It is categorical. Code does not lie. People do. And a price list that omits the variable determining ninety percent of the outcome is lying by omission.
Second: custody. Customers do not get an on-chain address. They get a contractual claim against Deka, which holds assets in an omnibus wallet — one shared pool — across four custodians: Crypto Finance, a Deutsche Börse company; Tangany; Bullish Europe; and Börse Stuttgart Digital Custody. Read the withdrawal terms carefully, because this is where the marketing performs its best trick. "Free withdrawals" sounds like a win. It is not. You can withdraw to four approved providers. You cannot withdraw to your own Ledger. You cannot withdraw to a self-custodied address, another exchange, or a DeFi protocol. The word "withdrawal" has been quietly redefined to mean "internal transfer."
That makes the product, in technical terms, a crypto-asset exposure certificate. Not ownership. Not keys. A claim on a balance sheet, sitting inside a walled garden with the gate welded shut. The capital in this system will never touch a chain. It lives on the CeFi ledger layer, and it dies there. For anyone counting on this news as an on-chain liquidity event, the honest answer is that its contribution to DeFi TVL is approximately zero.
Third, the number that actually matters: the cost math. Take a €1,000 order. The €0.99 fee is 0.10%. Fine. Now assume a principal spread in the 0.5%–1.5% range — where retail brokerage spreads have historically clustered, and Deka's own price list gives me no reason to assume better. Total cost lands between 0.6% and 1.6%. The spread is 85% to 94% of what you actually pay. The advertised fee is a rounding error bolted onto the real bill. Yield is a tax on ignorance. Here the tax is the spread, levied on the customer who believes the headline fee is the fee.
Fourth: the leakage nobody is discussing. SOL and POL are proof-of-stake networks. Native staking yields on those assets run roughly 3%–7% annualized. The price list mentions no staking service and no yield pass-through. If Deka stakes the underlying inventory and keeps the reward, that is an invisible cost that dwarfs €0.99 — a silent skim on the customer's position that never appears on a statement. Free custody is not free either. Custody has real costs: people, compliance, insurance, infrastructure. "Free" means cross-subsidized, which means the spread has to carry it, which means the spread is probably wider than the peer group's. Someone is paying. It is always the person who cannot see the line item.
Fifth: the asset list itself. A conservative institution that voted against crypto in 2022 goes live with XRP, SOL, and POL alongside BTC and ETH. Three high-volatility assets, one with a history of regulatory litigation, one whose market-cap standing has visibly eroded. That does not look like a risk committee's prudent shortlist. It looks like a liquidity provider's inventory. Think about who actually decides what is tradable. If you are Deka, you need assets your market-maker can quote reliably and hedge cheaply. The customer-facing "careful curation" narrative and the supply-side "what can we source" reality are two different documents. Only one of them is shown to the saver in Stuttgart.
Sixth, the tax trap. Under §23 EStG, German investors enjoy a one-year holding period after which gains are tax-free. But in this structure the customer must rely entirely on Deka's records to prove the holding period, because there are no keys and no independent on-chain proof of possession. Change custodians, or let Deka's systems migrate, and the evidentiary chain can break. The customer is trusting an intermediary's database to defend a tax position. That is a different risk class from self-custody, and nobody is putting it in the brochure.
Now the counterintuitive part, because the consensus read on this story is wrong. The consensus says European banks are finally embracing crypto, institutional adoption is here, this is bullish. I do not buy it. Look at the marketing posture. Deka is not advising. Deka is not promoting. Customers receive a full loss-warning. There is no minimum order but there is a €250,000 cap. That is not an offensive product. That is a defensive one. Read it as customer retention. For a decade, Sparkassen depositors have been quietly migrating to Trade Republic, Scalable, Bison, and Bitpanda to buy Bitcoin. The savings bank watched the assets walk out the door. A €0.99 in-app trading rail is not an attempt to win the crypto market — it is an attempt to stop the deposit base from leaving. The €0.99 is a retention cost, not a revenue line.
And here is the second blind spot: the split. Deka captures the spread — a percentage of transaction size. The local Sparkasse branch, if it gets anything, gets a slice of a €0.99 fee. On a per-euro basis, Deka's economics scale with volume; the branch's do not. That is a governance fight inside the network waiting to happen, not a unified product. So when the bull-market chorus tells you fifty million Germans are about to buy crypto, apply the check I always apply. Check the supply schedule. Always. There is no supply schedule here — no token, no unlock table, no inflation curve. So the tell is the spread schedule. And the spread schedule is exactly the document Deka has decided you do not get to see.
Strip it to the atomic truth: this is a fee story sold as an access story. The access is real but narrow — five assets, four custodians, one shared wallet, no self-custody, no chain. The fee is small but irrelevant — the spread is the bill, adjustable at the seller's discretion, precisely when volatility is highest and you most want to exit. The question I am sitting with is not whether Sparkassen succeeds at crypto. It is whether a closed-loop, principal-priced, self-custody-in-name-only product survives contact with customers who eventually learn the difference between a withdrawal and a transfer — and whether the next generation of German savers forgives the €0.99 once it discovers what it was hiding.