Britain's Bond Tape Counts 1.6 Million Licences, Not Users

Britain's new bond consolidated tape has acquired a striking headline number: 1.6 million subscribed licences. The Financial Conduct Authority repeated it in a 5 October speech, three and a half months after the tape launched.

The number shows that distributors can put the data in front of a large installed audience. It does not show that 1.6 million people use the tape, that 1.6 million terminals display it, or that the service has made bonds easier to trade.

The regulator's second set of numbers needs the same care. Much more bond trading is now reported without a long delay, but that change began with transparency rules in December 2025. The tape began consolidating those reports on 22 June 2026. Visibility, consolidation and liquidity are three different outcomes.

More than 50 customers, then a distributor multiplier

The operator supplied the missing shape of the licence count in its first-week update. ETS Connect UK said that more than 50 licences covered users and redistributors onboarded directly with it. More than 1.6 million licences came through redistributors, with the vast majority classed as individual licences.

That is not a criticism of the distribution model. The FCA designed it to extend the tape through the market-data platforms firms already use. Its licensing schedule says an individual licence is limited to one display or login session, an enterprise licence permits unlimited internal access, and a redistributor pays the operator for each licence passed to an end user.

The live fee table makes mass entitlement inexpensive. ETS Connect lists annual real-time and historical fees of one penny each for an individual. Enterprises with revenue below £50 million face the same data fees, while the largest enterprise tier pays £3,600 for real-time data and £900 for historical data. Retail historical access is free.

The 1.6 million figure therefore mixes licence types with very different relationships to use. Public disclosures do not reconcile it to unique people, unique firms, enabled terminals, paid subscriptions, logins, API calls or active viewing. A redistributed entitlement may be valuable access. It is still an entitlement, not an activity measure.

December changed publication; June changed aggregation

The FCA's launch release separates the sequence. New transparency rules took effect on 1 December 2025. The share of corporate-bond trades reported in real time then rose from under 5% to over 75%, while the government-bond share rose from around 30% to about 80%. The consolidated tape launched later and put those reports into one feed.

Those stated levels imply an increase of at least 70 percentage points for corporate bonds and roughly 50 points for government bonds. The October speech instead described a 70% increase for corporate bonds and 30% for government bonds observed in real time. The published pages do not explain how the latter figures were calculated or why the government-bond number differs.

They also do not state the denominator needed to reproduce the comparison: trade count or notional value, the observation dates, the definition of real time, the treatment of deferrals, and the precise instrument set. A June note from LSEG illustrates why that matters. It reported corporate-bond visibility at about 50% in December, 85% of euro credit notional visible within one trading day, and sovereign real-time publication at 16%. Those measures can all be true alongside the FCA's figures because they describe different windows and bases. They cannot be combined without the methodology.

The quality report tests received records, not missing trades

At launch, the FCA said the tape had 98% coverage of in-scope bond trading. Its stated perimeter is post-trade transparency data for bonds admitted to trading on UK venues, excluding exchange-traded notes and commodities. The release does not say whether 98% refers to contributors, trade reports, notional value or another base.

The operator's first data-quality report is encouraging but answers a different question. In July it received 1,069,757 reports and accepted 1,069,742, or 99.9986%. Fifteen were rejected. Among accepted reports, 26,174 were amendments and 21,097 were cancellations. Two records carried an outlier flag.

That shows strong validation and acceptance of records that reached the tape. It does not measure trades that should have arrived but did not, so it cannot independently reproduce 98% market coverage. Nor should accepted report count be read as unique trade count when amendments and cancellations are included.

The downstream path has also mattered. ETS Connect's operational notices record that some emerging-market credit notionals appeared about 100 times too large on Bloomberg in early July. The operator said its own data were correct and located the discrepancy outside its systems. Bloomberg temporarily suspended redistribution and later restored current data while historical records were repaired. Direct-feed accuracy did not guarantee correct display to every redistributed licence.

Liquidity still needs its own test

None of the published measures establishes that the tape narrowed bid-ask spreads, reduced execution costs or increased dealer competition. The licence count measures enabled reach. The quality report measures handling of received records. The real-time percentages measure publication timing under the earlier rules.

A market-quality audit would match bonds before and after 22 June, control for the December regime change and market conditions, and compare spreads, price dispersion, execution shortfall, dealer participation and volume. An access audit would separate direct users from redistributed entitlements, report active use by licence type, and show whether smaller firms and self-directed investors actually consult the data. A cost audit would include the price of the surrounding vendor product, not only the tape's one-penny licence.

The tape may yet improve price discovery and liquidity. Its early record already supports a narrower conclusion: Britain has built a cheap, broadly distributable and technically reliable view of reports it receives. The next proof must come from trading outcomes, not a licence counter.