Italy's green bond paid off, just not in basis points

Tuesday 22 September started badly for Italy's Treasury. That morning Istat confirmed that the 2025 budget deficit was 3.1% of GDP, a sliver above the EU's 3% ceiling, so Italy stays in the excessive-deficit procedure for another year. Economy minister Giancarlo Giorgetti said he took note of the figures "not without regret".

 

That afternoon the same Treasury sold €8bn of a new green bond maturing in October 2038, and the orders poured in. More than €110bn of them, from over 330 investors in more than 35 countries. Funds the Treasury classes as ESG investors ended up with around 80% of the bonds. Italian headlines called it record demand.

 

That was welcome news for the Treasury, but the order book tells us less than the headlines implied. The sale showed that investors will lend Italy large sums for 12 years at about 4.5%. It did not show that they value the green label enough to pay for it, and it says nothing yet about whether Italy's transition is on track. Italy borrowed €8bn. The other €100bn or so was never on offer.

 

What investors were buying

 

The bond pays a 4.40% coupon and was priced at 99.591, a yield of 4.492%. The 10-year BTP closed that day at 4.33%, nearly a full point higher than in January. The backdrop is the war that began when the US and Israel struck Iran on 28 February. Shipping through the Strait of Hormuz has been disrupted, and Brent crude peaked near $126 in April and was back near $100 this month. A 12-year eurozone government bond at close to 4.5% is a trade many fund managers would make whatever colour the label. France helped too. French bonds now trade more than 100 basis points over German Bunds, against 89 for Italy. For once, Rome was not the eurozone's main worry.

 

The order book itself overstates demand. Investors in syndicated deals routinely ask for far more bonds than they want, because they expect to be scaled back. In January the Treasury reopened its 2046 green bond for €5bn and collected more than €115bn of orders, 23 times the size. The ordinary seven-year bond sold alongside it drew about €150bn for €15bn. By Italian standards, September's cover of almost 14 times was routine, and January's green book was bigger.

 

No discount for going green

 

The question the order book cannot answer is whether the label made the loan any cheaper. The hoped-for saving has a name, the greenium: the few basis points of yield an investor gives up to own a green bond instead of an otherwise identical conventional one.

 

The new bond was listed on Borsa Italiana's MOT platform on 25 September, and that day it was quoted at 97.90, a yield of 4.68%. The conventional BTP maturing in September 2038 yielded 4.66%, and the March 2038 bond 4.64%. The green bond matures two months after the September bond. Adjust for that and the gap disappears. An independent curve model from Atlantis Data Solutions puts it just 1 basis point rich.

 

Italy's older green bonds tell the same story. The October 2037 green line yielded 4.59% on 25 September, 1 to 3 basis points below the conventional curve at that maturity, and the April 2035 line sat 3 to 4 points below. Neighbouring conventional BTPs often differ by as much. Since pricing, the new bond has simply moved with the market: its yield rose 19 basis points by Friday, while the 10-year rose 17.

 

None of this should surprise the Treasury. Economists at the Banque de France found an average greenium of 2.8 basis points across 21 euro-area sovereign green bonds between 2021 and early 2024, and 5.2 points on Italy's lines in 2023. In June 2025 the Treasury's own presentation to a World Bank and IMF debt forum in Paris noted "a slight decline in the greenium", which it put down to green bonds becoming mainstream.

 

The Treasury has not published the banks' price guidance, so public data cannot show whether the deal priced at the tight end. Earlier green deals tightened by 2 to 3 basis points during bookbuilding, and with a book this size this one probably did too. It hardly matters. A basis point on €8bn is worth €800,000 a year, in a year when Rome plans to sell between €350bn and €365bn of medium- and long-term debt.

 

Pricing the bond like any other BTP was the right decision. India shows the alternative. Its central bank cancelled a 10-year green auction in May 2024 when bids came in 1 to 7 basis points above the benchmark, and in June 2025 scrapped a 30-year sale despite Rs109.4bn of bids for Rs50bn of bonds. The prize India was holding out for, by the think-tank IEEFA's estimate, was a greenium of 2 to 3 basis points. It was not worth the unsold bonds.

 

Steadier buyers in a growing market

 

Issuers keep coming back even as the greenium shrinks. Green bond sales worldwide reached $653.5bn in 2025, the second-highest year on record, and the cumulative total has passed $4tn, according to the Climate Bonds Initiative. The money also seems to change behaviour. A BIS study published in March 2025 found that companies' emissions intensity fell by about 21% on average in the year after their first green bond, with the effect concentrated in carbon-intensive sectors. The market is still far short of the roughly $2tn a year the BIS puts on global climate investment needs, but it has become a real funding channel for the transition.

 

For Italy, the clearest gain is in who holds its debt. ESG investors took about 80% of the new bond, against roughly 55% of January's 2046 tap. Central banks and other official institutions took 20.8%, and pension funds and insurers another 10%. The Treasury says green issues attract mainly long-term, buy-and-hold investors. Not all ESG money is patient, and ABN AMRO has found that most ESG funds trade actively. But the holders a green label brings in do seem to stay put when markets turn. Studying the Covid sell-off of early 2020, economists Serena Fatica and Roberto Panzica found that green bonds suffered lower net sales than equivalent conventional bonds, and that among bonds from the same issuer, mutual funds' holdings of the green ones ran about 7% higher. That stability is worth more than a basis point while the European Central Bank is shrinking its BTP holdings by about €72bn this year, after €73bn in 2025.

 

The claim that green bonds are simply less risky needs more care. Credit risk is identical, because a BTP Green and an ordinary BTP are obligations of the same state. The argument rests on price behaviour, and there the evidence is mixed. ABN AMRO found that yields on green government bonds were less volatile than on conventional ones, measured over one year and over four. A study by Ewa Dziwok and co-authors found the opposite during the 2022 energy shock, when Germany's green twin bonds transmitted volatility while ordinary Bunds absorbed it. The evidence on holders is firmer than the evidence on prices.

 

Where the money actually goes

 

The other gain is disclosure. An ordinary BTP funds the budget in general, and nobody can say what a particular bond paid for. A green BTP comes with a public record: each year, most recently in June, the Treasury publishes an allocation and impact report, reviewed by an outside assessor, that matches green issuance to named categories of spending. The money itself goes into the Treasury's general account at the Bank of Italy, like the proceeds of any other bond, and the matched spending can date from up to three years before the sale. So the bond does not have to build anything new. It makes the spending visible, and that also opens the spending to challenge.

 

The June report covers the €13.6bn raised from green bonds in 2025, matched against spending from 2022 to 2025. The largest share, €6.2bn or 45.2%, went to energy efficiency in buildings, and €6.09bn of that was the Superbonus. The rest, about €80m, went to public housing. Transport took €4.7bn, or 34.7%, covering rail maintenance and electrification, new high-speed lines, electric buses and incentives to replace old vehicles. Renewable energy got €150m. That is 1.1%. Anyone who pictures Italian green bonds paying for solar farms is mostly paying for railway work and the Superbonus.

 

The Superbonus and the taxonomy gap

 

The rail allocation is relatively straightforward. The Superbonus raises harder questions. Introduced in 2020, it gave homeowners tax credits worth 110% of the cost of energy-saving renovations, and the energy agency ENEA puts the bill to the state at €131.5bn as of 30 June. Meloni's government has blamed the scheme for the swollen deficits that put Italy into the EU's excessive-deficit procedure in 2024. So on the day Italy learned it would stay in that procedure, it sold a bond from a programme whose single biggest use of money last year was the Superbonus.

 

The Treasury estimates that the 2025 allocations will avoid about 38 million tonnes of CO2 and add around €20bn to GDP over five or six years. Both figures are model outputs. The same report also tests the allocations against the EU taxonomy. By that measure, 82% of the money went to activities that make a substantial contribution to an environmental objective, and about 69% also passed the "do no significant harm" test. That is an improvement on the year before, when the figure was about 60%. The Treasury has put the gap down to missing data: most public bodies receiving the money are not required to report on sustainability. That may well be true. It still means that for close to a third of the money, the Treasury cannot show it passes the EU's own test.

 

Four questions for the next report

 

Italy has built one of Europe's largest sovereign green programmes, with more than €70bn issued since its debut in March 2021. The steadier buyers and the disclosure are real gains, and holding on to them depends on how the next report answers four questions.

 

• What exactly was financed? Named programmes and projects with amounts attached, rather than budget chapters. The €4.7bn transport figure should be split into rail upkeep, new lines, electric buses and vehicle subsidies.
• When was the money spent? Each report should say how much of the allocation matched spending the state had already made before the bond was sold, and how much was new.
• What changed on the ground? Measured results, such as energy saved in renovated buildings or kilometres of line electrified, published next to the modelled tonnes of CO2.
• Would it have happened anyway? This is the hardest question, and sovereign issuers rarely attempt it. A tax credit that pays for renovations homeowners had already planned does little for the climate, whatever the bond behind it is called.

 

The December 2025 green bond framework promises closer alignment with the EU taxonomy. The Treasury should turn that promise into a single number, the taxonomy-aligned share of each year's issuance, and put it on the first page of every report with the projects behind it named.

 

The Treasury knows where the difficulties are, because it listed them in Paris: coordinating ministries, tracking spending over time, getting data on small projects, and avoiding double counting with EU recovery funds. It also warned that insufficiently transparent reporting "can weaken investor confidence, particularly among ESG-oriented investors". Those are the investors who took 80% of this bond.