The PNRR and the naive friend
(… I should be making slides for Vicenza, but the train is shaking too much—and it arrived early in Bologna, by the way, which can't be said because it contradicts the terrible narrative I'll be discussing in Arona. So I'll spend some time with you, waiting to arrive somewhere stable …)
As you may have noticed, lately I've been devoting less attention to the black factory of discomfort (X, also known as the sewer), while I often express myself on the blue factory of discomfort (FB), that social network that everyone thinks is dead, but which in the provinces (and beyond) has its traction. I don't feel like contaminating Instagram, but I'll restart TikTok soon if two or three of the people belonging to the "you're our heroh11!!1! How can we help youh11!1!" legion show me the distinguished courtesy of doing what I tell them—or even if they don't (you'll wonder why the first thing I do when we meet is tell you to fuck off: if you spent not just a day, but an hour with me, you'd wonder why I don't do worse).
I get the feeling that Facebook is less infiltrated by bots, that is, it allows me to escape the quicksand of self-referentiality that non-existent blogs inevitably fall into (because to get there, you need to know it exists!), without excessively polluting or impoverishing the debate. In fact, typically, about 75% of the feedback I receive, which is often in the six-figure range (but the goal is seven figures), comes from people who don't follow me and for some mysterious reason find my content in their feed. Consequently, I try to use more esoteric (aka "idiot-proof") language, abandoning the esoteric vocabulary familiar to our community . Moreover, the very nature of the platform dictates a certain healthy conciseness, self-contained writing (because if you insert links, for example to cite a source, the algorithm penalizes you), and a targeted and incisive use of iconography (because if you use more than one image, the algorithm penalizes you). We're therefore dealing with a slightly different literary genre, but one that offers its own rewards, such as fooling the simpletons by presenting them with proof of how foolish they were in believing Giuseppe G. Conte (where G. stands for "ghrhadhuidhamendhe"). As an example, for those who aren't too bold with "I don't go on social media because I don't want to be tracked" (but then use their cell phones and pay by card), I offer this knockout one- two punch. Pieces of simpletons everywhere, but you, who are the spearhead of the debate, deserve more, and so I'll give you something more here, in the house of Debate.
Let me start by saying that I'm speaking to people who know how the PNRR works, and therefore understand that even the "non-repayable grant" must actually be repaid with interest ( explanation and sources here ). Our discussion of the PNRR's financial burdens will therefore primarily concern those related to the so-called "loans" (or loans, or debts). This is because only in this case do the financial burdens, being paid directly by the Italian government to the Commission, enter the public accounts under a specific heading called "interest." The "non-repayable grant" expenses are repaid through the EU budget, and therefore, in our public accounts, they pass through the form of a contribution to that budget, without it being particularly easy to distinguish between what goes into structural funds, what goes into NGEU capital repayment, and what goes into NGEU interest repayment. The overall cost (the total flow of interest we will have to pay between now and 2056), therefore, is underestimated, and the €60 billion mentioned by the excellent Liturri in today's La Verità is, by his own admission, an educated guess . It could be less, but also more, and my educated guess is that since Europe's reckless energy policies condemn us to a future of inflationary pressures, it's very likely that the final bill will be higher. Here, however, I want to talk to you about something else, one that's relatively more documentable (and in fact Giuseppe documents it today in La Verità): what is, at least approximately, the interest rate on bonds issued by the EU to finance the "recovery"? What, in other words, are the interest rates on EU bonds, and are they really that favorable compared to those on BTPs? I'll first give you the summary answer, taken from Giuseppe's article, and then the details:
The bottom line, as you can see, is that at a certain point, around 2023, the PNRR loans actually had a significant advantage, in the order of 60 bps (BTPs at 4%, EU bonds at 3.4%), but this advantage concerned only two tranches (the so-called "installments"), for a total of €22.9 billion (a teardrop in the ocean of public debt), while for all the other tranches the advantage was smaller (15 bps initially, 7 bps at the end), and note well: when the plan was launched , there was no way of knowing that it would be advantageous for a certain period , and this is because essentially no one except us had considered that COVID had been a gigantic supply-side shock that would lead to supply-side inflation, with rising rates and therefore widening the spread between central and peripheral countries. The benefit, ex ante (eggzandi, as the imbeciles say), was neither foreseen nor taken into account, and in fact the project was not advertised by saying that it would be convenient, but that it would be ghradhuido.
But… how can we reconstruct this data?
Understanding the cost of "recovery" has been a concern of ours since the beginning. This is evidenced by this parliamentary question from a friend of ours , which received this response at the time (substantially in line with the data reconstructed ex post by Giuseppe):
But where do these numbers come from?
Naturally, since it has to sell its EU bonds on the market, the Commission provides investors with all the information they need to decide whether or not to purchase them. This information is, of course, on a page I've already mentioned, the " Investor Relations " page. As you can easily imagine, even though the repayment of the installments will be over thirty years (with pre-amortization, etc.), the fundraising is not entirely based on thirty-year bonds: bonds of varying maturities have been placed on the market, and therefore the rate corresponding to each installment can only be calculated conjecturally by referring to the average of the issues placed during the disbursement period of the installment. Thus, for example, when we approach the placements made in 2023, the picture looks like this:
with nice interest rates above 3%, which the original drafters of the plan never dreamed of reaching, and which, as Giuseppe's summary shows us, were on average 60 points below the BTP in that year , but they were not zero and they were not half of the BTP, as some naive idiot friend has been braying for days in the blue factory of hardship, arguing that since the loan was for 200 billion and an interest installment of 2.8 billion, then the interest rate was 1.4% (like saying that if I pay a 200 euro installment to buy a 12,000 euro car the APR is 1.6%. Easy like that, right?).
And here there would be a discussion to be had.
“On financial mathematics,” you, my 20 (thousand) readers, will say?
No.
On democracy.
But it's a delicate and painful discussion, and we'll have to talk about it another time.
This is a machine translation of a post (in Italian) written by Alberto Bagnai and published on Goofynomics at the URL https://goofynomics.blogspot.com/2026/02/il-pnrr-e-lamico-ingenuo.html on Fri, 27 Feb 2026 15:40:00 +0000. Some rights reserved under CC BY-NC-ND 3.0 license.