FS International Says Europe Can Find the EUR 345bn for High-Speed Rail. A EUR 10m Certification Bill in Every Country Is What Stops It

The one thing I keep asking rail sponsors is what their signalling specification will look like in year seven. Very few have an answer. That question now sits on top of a EUR 345bn European build programme.

Railway Gazette International reported the exchange on 28 September 2026. Luigi Corradi, chief executive of FS International, put it plainly at InnoTrans. We can find the money, he said, the problem is stability. He then gave the number behind the complaint. Certifying a train in a single country can cost more than EUR 10m.

Money was never the binding constraint

The European Commission launched its high-speed rail plan on 5 November 2025. It targets an integrated network by 2040. Completing the planned TEN-T network carries a EUR 345bn infrastructure bill. Tripling the network at 250 km/h and above is costed at EUR 546bn. The Commission's working range to 2040 now runs from EUR 354bn to EUR 540bn.

Officials on the panel could not say whether those projections still hold. That admission matters more than the size of the number. A cost estimate nobody will stand behind is not a financing case.

Certification behaves like a variation with no cap

Laurent Fourtune of Kevin Speed described approval timelines of anywhere between eight months and eight years. That is not a schedule. It is a range wide enough to swallow a financing structure.

Corradi described trains carrying four signalling systems at once. He pointed to different versions of ERTMS between France and Italy. For a contractor that is a design problem. For an investor it is worse.

A 25 or 30 year availability concession prices a fixed technical scope against fixed payments. A specification that moves mid-concession turns scope into an open variation. Nobody writes the cap, because nobody knows the next version.

The trackside baseline is still moving too

The European ERTMS Coordinator's Third Work Plan put the position on record. About 10,600 km of the TEN-T core network is equipped, or 17 per cent. About 8,730 vehicles carry onboard equipment, or 19 per cent of the fleet. Railway News reported on 15 March 2026 that the 2030 core target will be missed by roughly half.

The Work Plan is direct about the cause. Industrial capacity is not the bottleneck. The gap is the absence of binding, synchronised and funded national deployment plans. More than 20 legacy Class B systems remain in service across the network.

What a frozen specification is actually worth

The European Union Agency for Railways offered the remedy at the same event. Executive director Oana Gherghinescu said the agency would freeze technical specifications. The stated purpose is breathing space for industry.

Read that as a commercial instrument rather than technical housekeeping. A frozen baseline lets a lessor underwrite residual value. It lets a bidder price 25 years of maintenance. It lets a lender size debt against revenue that a revision will not strand.

Spain offers the upside case. ADIF chairman Pedro Marco de la Peña pointed to more than 3,000 route-km running ERTMS, the largest deployment in Europe. The technology works at scale where one owner committed to one baseline and held it.

The read-across for CEE programmes

Central and Eastern European sponsors are buying signalling right now. Rail Baltica, CPK in Poland, the Czech high-speed programme and Romania's Rhine-Danube corridor sections are all mid-procurement. Each is committing to a version of ERTMS inside contracts that will outlive it.

The exposure is concrete. A cross-border corridor is only as usable as its least harmonised section. A Baltic or Carpathian line delivered to a superseded baseline still needs a retrofit before through-running works. That retrofit sits in very few business cases.

There is a bid-market effect as well. Where a specification is unstable, the field thins. The firms that stay charge for the uncertainty. The buyer pays that inside the tender price rather than in a later claim.

What sponsors should do with 2027

The Commission has set 2027 as the date for cross-border bottleneck timelines. An ERTMS deployment plan is due during 2026 and a rolling stock proposal in 2027. Those three documents will decide whether private capital treats European high-speed rail as financeable at all.

Sponsors chasing competitive bid fields should stop selling the pipeline and start selling the baseline. Contractors and lessors can price almost any risk that holds still. The money has already said it can be found. The specification is what has to stop moving.