Since Operation Epic Fury was launched, bond yields have risen. There are a lot of arguments swirling around as to why. Maybe they're higher because oil prices surged, pushing up inflation expectations.

Maybe they're up because investors are questioning the credibility of the Fed and/or the soundness of US debt. Or maybe — as Stephen Miran wrote last week — it's because the invasion of Iran marked the moment that investors began ratcheting up their long-run economic growth expectations. What we can say is that higher bond yields mean that governments now have to cough up more when they sell bonds than they did before the shooting began.

How much more? Around $50bn this year, Alphaville reckons. But it's an annoyingly fiddly thing to estimate. Because Alphaville was stupid enough to stick up our hands and volunteer to crunch the numbers when kind Main FT colleagues came calling, we thought it would be only fair to make you suffer the workings.

It's not that hard, right? Just take total bonds issued and multiply by the amount that yields have risen? Sadly not. To work out how much more governments have locked in as additional annual interest costs, we need to run a counterfactual analysis, comparing the interest cost each new bond issued locked in against the interest rate they could've locked in, if yields had not risen (like they have). And given that some bonds are issued with a term of two years, some three years, some 40 years, we're going to have to match their coulda-woulda-should issuance costs against the appropriate point for each sovereign yield curve.

Actually, for most G7 countries, we need two yield curves: a nominal one and an inflation-linked one. So the calculations are not at all complicated, but there's quite a lot of data to collect. We first downloaded details of every bond auction, syndication and other meanderings into primary markets we could find from each of the G7.

Reader, this was a lot more complicated than you might assume. For sensible countries like the UK, the US, Germany, Canada, France and Japan, this is straightforward enough, if a bit time-consuming. Sure, different debt management offices have their esoteric rinky-dinks.

The UK's DMO has separate downloads for outright auctions, syndications and tenders. The French separate syndications, while the Canadians separate green bond issuance. And they all have their own formatting conventions.

But they each look like they're at least trying to make their data accessible. Then there's Italy. You'd think that this Banca d'Italia link to download results to current-year auctions might be just the ticket.

But it omits not only syndications, but also their chunky retail issuance programme. So we went pleading to friend-of-Alphaville and Italian bond guru Gus Baratta, who pinged us some more comprehensive data. Next, we compared actual yields at which governments sold their debt to secondary market levels at the end of February.

Here are some of the intermediate bits of data for the US, UK and Italy (just their non-inflation-linked bits though): All we then needed to do was multiply the gap between each bubble and each country's prewar yield curve as of end-February by the size of bond issued, and tot up the totals. Ta da: $16.25bn. Interestingly, at least for us, the country with the proportionate biggest hit to issuance costs so far has been the UK — where issuance-weighted rates of interest locked in so far are a full 71 bps higher than prewar levels.

This compares to a lowly 21 bps bump in issuance-weighted rates of interest so far locked in for Japan. And, to be clear, we've excluded bills from our calculations, and focused entirely on fixed and inflation-linked coupon bonds. We thought that this might be because the UK skewed its issuance longer out the curve than Japan, but the average maturity of non-bill issuance is almost identical for the two sovereigns.

Moreover, the rise in 10-year Japanese government bond yields is not a mil...