By Mark Schaedel, CEO of DataBP

Copenhagen was wallpapered with billboards during the World Financial Information Conference last week. Pyth, the oracle network, had covered the town from the airport to the city centre with the slogan: “The Price of Everything. Everywhere. No License Fees.” Oscar Wilde defined a cynic as someone who knows the price of everything and the value of nothing. He could not have asked for a better illustration. 

The same week, at the Fixed Income Leaders Summit, the consolidated tape panel was asked whether the bond tapes have a free rider problem. They do, but it is not the one the panel discussed. The industry has started arguing about who pays for the tape and has still not asked who pays for the price on it. 

Start with what a price actually is. A displayed, executable quote is a public commitment to trade, exposed to adverse selection from the moment it appears. Somebody has to be willing to be wrong in public for a price to exist at all. Investors and market makers supply that commitment and carry that risk; venues provide the regulated plumbing that makes it visible and actionable. Consolidated tapes aggregate the result, and so do oracles, and it is worth being clear-eyed about what that means: neither of them creates a basis point of information. A tape is a mirror, and a mirror with a billboard budget is still a mirror. 

For most of the industry’s history, nobody needed to pay for the information separately, because there was nowhere else to take it. The quote and the trade arrived together at the central venue, and the transaction fee quietly compensated both. Decentralisation has cut that weld. The quote still carries all the risk it ever did, but the liquidity it attracts can now be executed somewhere else entirely, against a reference that cost the executing venue nothing. The transaction no longer pays for the information because the transaction no longer happens where the information was made. 

The numbers show where that leads. In 2025, more than half of US equity volume executed off exchange for the first time. In Europe, lit continuous trading has fallen from 54 per cent of turnover in 2022 to under 45 per cent in 2025 on ESMA’s own measure, and to barely 37 per cent on independent estimates. The NBBO and the fledgling EBBO now contain a shrinking minority of the market, yet best execution is measured against them, retail orders are priced off them, funds are valued by reference to them, and every derivative, index and tokenised product settles against them. When LIBOR stopped representing its market, regulators treated it as a systemic failure and rebuilt the benchmark. The equity benchmarks are drifting towards the same condition, and the regulatory response has been to approve more of the trading that moves activity away from them. 

Economists call this free riding: consuming a good that others paid to produce while contributing nothing to its production. In most settings it is a nuisance. In a market whose product is the price, it is a slow demolition, because the good being consumed is the very thing the consumption erodes. The internaliser executes against the public reference without displaying comparable interest of its own. The perpetual futures venue settles single-stock exposure against an oracle price. The oracle, for its part, mostly republishes the public order book, consumed under an exchange licence written for display screens, run through a model and resold. “No licence fees” turns out to mean somebody else’s licence, repackaged, and a business model whose entire margin is the fee it declined to pay. 

None of this is illegal, and I am not suggesting any of it should be banned. All of it, though, is built on public price formation, and that dependence appears nowhere in how data is licensed or how tape revenue is distributed. Policy is making three mistakes at once. 

The first is pricing data to its cost. ESMA’s reasonable commercial basis rules now make it unlawful to price data by the value it represents to the user; the UK insists fees track cost; the new tapes are tendered on price. Every one of those rules can see the plumbing and none can see the substance. Valuing a price by the cost of producing it is like valuing a Rembrandt by the cost of the paint. The value does not disappear when the rules refuse to recognise it. It transfers to the firms that depend on the price and contribute nothing to forming it. 

The second is paying every reported trade the same. A continuous execution against displayed interest, a midpoint match, a negotiated block and a benchmark-priced print convey very different information, and a tape that pays them identically is buying volume, not information. The US has already run this experiment twice under Regulation NMS, and the lesson for every new tape is the same: reward the behaviour the market needs, and expect anything cheap to manufacture to be manufactured by Tuesday. 

The third is treating the cost of consolidation as the value of the information. The UK bond tape tender bid its price cap from £1,209 to 68 pence and then landed in court. For roughly the cost of a chocolate bar, the UK bought itself a bond tape. What does that tell you about whether the data is complete, timely or sustainably produced? Nothing at all. 

Each mistake has a remedy, and the remedies fit together. Classify the uses of price data into two tiers and price them in opposite directions. The first tier is public access: retail investors, display users, researchers, the press. Price it low enough that intermediaries can offset the fee, with the explicit aim of reaching as many people as possible. The 15-minute delay, incidentally, belongs in a museum. 

The second tier is commercial dependence: the index provider with trillions tracking a weighted list of public closing prices, the internaliser, the systematic internaliser, the derivatives venue, the tokenisation platform and the oracle reselling to all of them. Could any of them exist without the price-forming market? No. Does any of them contribute a quote to it? Also no. That tier pays a licence fee scaled to its use: by assets benchmarked, by notional executed against the reference, by notional settled. Every one of those quantities is already measured for some other purpose, so the metering problem is a reporting problem, and reporting problems are the kind this industry has been solving since MiFID I. 

The proceeds fund the public tier and flow back to the contributors of price formation through tape allocations that can tell an execution that formed a price from a print that merely referenced it. The Market Model Typology, already mandated for post-trade flagging in the EU and UK, provides the vocabulary. The weights are the policy decision, and it is the one regulators are making right now, whether they realise it or not. The FCA’s CP26/31 rightly requires the UK equity tape to share income with contributors; the consultation this autumn on how that income is divided is where the contribution distinction has to be made explicit. 

Every other information industry settled its economics this way. Music did not survive digital distribution by charging everyone the same or by giving up on charging; it survived by metering use. Nobody put the royalty rate out to tender and congratulated themselves when it reached 68 pence. 

The data economy has arrived in capital markets, and the rules still treat price discovery as a utility to be run as cheaply as possible rather than as the asset on which everything else is built. Benchmarks do not fail overnight; they drift, and LIBOR drifted for years before anyone acted. The equity benchmarks are drifting now. Every consultation that closes without asking who pays for the price is a year the market will not get back.