Today’s Free Lunch brings the last part of our mini-series on financial sanctions against Russia. (If you missed them, here are the preview and parts one, two and three.) Thanks again to my colleagues Daria Mosolova and Claire Jones for their help in the past few months of looking into these issues, and all the experts we have talked to (you know who you are) in a journey that has taken my level of understanding from the obscure to the merely foggy.
Apologies for the mammoth length of today’s article, but this is an area where the more you look, the more you find to unravel. On Tuesday, I described the phenomenon of Russia’s “shadow reserves” — the large surplus energy earnings that were not placed under sanctions last year. As I explained, the motives for blocking access to official reserves, ie central bank assets, apply with equal force to the shadow reserves. I don’t know whether the western sanctioning coalition is tracking what has happened to the unsanctioned payments and is preparing to put restrictions on the resulting cash piles. But I do know that if it wants to, it should be able to have a good sense of where the money has ended up — certainly a much better sense than the limited detective work that can be done with public data.
As I pointed out last week, however, even in the case of central bank reserves, the sanctioning coalition has been slow to begin systematically mapping what is held where. So I fear that there has been even less monitoring of the unsanctioned payments, which may become even harder because of other sanctions decisions. “De-Swifting” Russian banks — kicking them out of the Swift interbank messaging network — has made it much more difficult (though not impossible) for targeted banks to transact across borders. But as Elina Ribakova pointed out to us, “the question is whether Swift actually helps following the money”. Swift is, after all, based in Belgium and also reports to the US. On the other hand, a person familiar with the sanctions decision-making in Washington told me Gazprombank had remained unsanctioned in part to encourage most transactions to go through a single channel so they would be easier to track.
We can be sure that Moscow is thinking about how to put its unsanctioned cash pile, however big it is, beyond the reach of western jurisdiction. The Putin regime’s operators understand that once Europe fully sheds its dependence on Russian energy, an important reason to limit sanctions will have gone. How to secure itself as far as possible against possible future sanctions would be the logical extension of Russia’s work to reduce its vulnerabilities to western measures since its first invasion of Ukraine in 2014. This “de-dollarisation” strategy is well described in a paper by Maria Shagina, and includes both shifting its reserves out of the dollar — most strikingly towards gold and Chinese renminbi — and building up alternative payment processes that don’t rely on Swift. (But as Alexandra Prokopenko has pointed out, Russia’s increasing reliance on China comes with risks of its own.)
How might Moscow have gone about protecting its shadow reserves? The starting point is that energy earnings not hit by sanctions will initially have been paid to state-controlled companies in dollars and euros. Take the case of European payments for gas. They would, in a first step, have been paid into a euro account whose ultimate beneficiary is Gazprom — “ultimate” because President Vladimir Putin’s demand to be paid for gas in roubles last year involved setting up Gazprombank accounts “on behalf of” buyers. So let us think about Bank GPB International, the Luxembourg subsidiary of the Gazprombank group (Gazprom’s bank).
What happens when the buyer pays euros for gas is that the buyer’s bank — either in the eurozone itself or through a correspondent eurozone bank, instructs (through Swift) Gazprombank to credit the account designated by Gazprom as the one to pay into. At the same time, it debits the buyer’s account and “pays” Gazprombank through the eurozone Target2 settlement system in the form of claims on the eurosystem of central banks. And when that payment crosses national borders, there will be a similar transfer of Target2 credits between the countries’ central banks. (Here is a good overview article about the international payment system, and here is the European Central Bank’s explainer of how Target2 works.)
So the first incarnation of a European payment for Russian gas is a claim of Bank GPB on the Banque centrale du Luxembourg, and the parallel BCL asset in the Target2 balances. In the normal course of events, these euros would then move on and partly be spent by Russia on imports, invested by Russian residents in various instruments abroad, or added to Russia’s central bank reserves (largely not held in Luxembourg). So the BCL balance sheet would only contain a stable “buffer” amount of flow-through funds related to Gazprom’s gas sales, which we would in any case not be able to distinguish from other monies in the aggregated public data.
But we can ask what we would expect to see in a situation where gas earnings balloon and many of the normal routes for that money to move on are closed down. We would expect a sudden increase in both the BCL’s Target2 assets and its liabilities to Luxembourg-resident banks. And then, if Russia managed to find new ways to spirit the money away, we should see a fall in both.
In fact, here is what we see:
Now, I have no way of telling if this data in fact reflects that Moscow had Gazprom pile up record euro earnings in Luxembourg through GPB until mid-2022, then abruptly managed to move them somewhere else. But this is what it would look like if that had happened. And we would be seeing something similar inside the dollar system for oil sales and any non-euro gas sales (such as liquefied natural gas contracts). As it happens, US banks saw a significant rise in liabilities to Russian counterparts in the first half of 2022 and a steep drop in the third quarter.
Luxembourg authorities are, of course, perfectly able to verify whether this is a red herring. One hopes that the sanctioning coalition has been receiving a very detailed breakdown of information from the BCL and Luxembourg bank supervisors. One is even allowed to hope that it will start to tell us something about what they find.
Supposing something like this is indeed what has happened, what would have been the escape routes for this money once it started moving out? Here are three possibilities.
One is simply that Gazprom and Gazprombank turned in their euros, exchanging them for roubles with unrelated parties in order to pay taxes or dividends at home. In that case, the new holders of the euros would be those who have been trying to get money out of Russia. Matthew Klein attributes a lot of the Russian asset accumulation in 2022 to households transferring hard currency-denominated deposits abroad or taking out foreign exchange cash; he points out that Russian data showed this happened in large quantities. In this scenario, Gazprom/Gazprombank and their ilk would have sold their accumulated euros — probably on the Moscow Exchange, on which more in a moment — and those euros would have been bought by households trying to get their money (and often themselves) out of Russia.
But I am unpersuaded, for a simple reason: the international banking data I analysed on Tuesday that shows large increases in Russian claims on western banks, shows no change in the claims of households. Mostly the changes are in western liabilities to Russian banks; or conversely Russian banks’ deposits in western ones. In addition, of course, there are capital controls on taking money out of Russia. Above all, surely Moscow would have wanted to build up its shadow reserves, not accommodate all capital outflows.
Here is a second possibility. This also involves Russian exporting companies exchanging many of their euros and dollars for roubles (or making their buyers do so), because Russian law has required it. And here we return to the Moscow Exchange. A smart study by Bank of Italy economist Michele Savini Zangrandi points out that the Russian decree demanding that gas buyers convert their payments into roubles also specified that the conversion must take place on the Moscow Exchange through its National Clearing Centre division, the central counterparty for currency trading. Zangrandi suggests that linking the NCC to energy payments would protect it from sanctions, much like Gazprombank has been. This sounds plausible: although the US has imposed sanctions on its chair, the NCC itself remains connected to Swift and keeps unrestricted correspondent accounts in euros and dollars with JPMorgan in Frankfurt and JPMorgan Chase in New York, respectively. (You can look up the account numbers if you are looking to buy large quantities of roubles.)
But what exactly does it mean to exchange euros (or dollars) through the NCC? Apart from outright selling the euros for others — unrelated to the Kremlin — to buy, could it for example involve simply committing euros as collateral for a future trade or a current rouble loan? Could it mean placing euros with the NCC to hold in its correspondent account on the ultimate owners’ behalf? Could it mean taking a derivative position for which only the limited margin calls have to be honoured up front? These various options will determine on whose balance sheet the euros will sit, and in particular who takes the currency risk. It seems conceivable that Gazprom could pay its rouble taxes to the Russian government by borrowing roubles against euro collateral (sitting in Luxembourg) or that the NCC borrows roubles to buy Gazprom/Gazprombank’s euros from it. Either way, there could be an enormous currency mismatch on some entity’s balance sheet.
But from both the Kremlin’s and western policymakers’ points of view, this shouldn’t matter too much. What matters instead is that these kinds of manoeuvres would retain hard currency at Putin’s disposal: shadow reserves. As far as euro holdings are concerned, they would presumably sit in the NCC’s account in Frankfurt, if they have not been sold off to other state-connected companies. Either way, that is easy for German authorities to know. (Ditto for the Federal Reserve and NCC dollars in the New York account.) One hopes that they have found out and shared their findings with the rest of the sanctioning coalition.
If this is indeed what has happened, then there should have been a shift of Target2 claims from the BCL to the Bundesbank. Now, German Target2 assets are so big it is hard to make much of moves of “only” a few tens of billions, and some of their movements reflect pandemic monetary policy action. But for what it’s worth, I have charted the changes in the two Target2 balances below. Most interesting is how they diverged from mid-2022.
Neither of these first two possibilities seems like they would satisfy Moscow, however. If euros have been sold off to help capital flight, then they are no longer within the Kremlin’s reach. If they remain in the NCC’s correspondent account, or can be traced to other state-connected companies’ euro accounts, then they remain within reach of sanctioning governments.
So we should expect huge efforts at the third possibility: to hold on to the hard currency but move it to friendlier jurisdictions. And I don’t mean exchanging dollars or euros for renminbi or gold, which has limited use and which the Russian government has a lot of already. The challenge is to find someone who can hold hard currency on your behalf beyond the reach of sanctions.
Here is one way. Gazprombank could issue a euro or dollar-denominated loan to a friendly company in a friendly third country. The recipient would see a hard-currency credit in their bank account, for which their local bank will have a matching claim on its central bank. That friendly third country’s central bank will, in turn, have a matching claim on the originating bank’s central bank — say an increased reserve deposit with the eurosystem central banks (paid by reducing Gazprombank’s claim), or in the case of a dollar loan, this will have shifted to a deposit with the Fed. All above board, and untouchable if the friendly third country doesn’t join the sanctions and won’t itself be hit with sanctions.
If I were the Kremlin, I would be looking at places such as Turkey, the United Arab Emirates and India (China is trickier because of its tight capital controls) — just the countries US and EU sanctions officials have paid many visits to this past year. So it’s interesting that Gazprombank made a big loan to Rosatom’s Turkish subsidiary last year, ostensibly to finance a big nuclear plant, but seemingly issued up front rather than drawn down as and when needed. If this is what happened, there should now be about $15bn sitting in a Turkish bank account, matched by the increase of about that amount that could be observed in Turkey’s foreign exchange reserves last summer — no doubt a welcome capital inflow for a currency under pressure. In time, maybe when sanctions are lifted, the loan could be paid back or spent as the Kremlin sees fit.
It would not be without risk: Turkey could conceivably use up its official reserves in defending the lira, which would cause problems once the Rosatom loan was to be either spent or paid back. But it seems preferable from the Kremlin’s perspective to either see the hard currency disappear altogether or to leave it within the sanctioning jurisdictions. Again, these numbers do not prove that Putin used this transaction to move shadow reserves beyond the reach of sanctions. But if he did, it would show up in numbers just like these. And if so, there are surely more manoeuvres of this sort going on.
Almost everything I have described above as plausible is information that the western sanctioning countries have or can get: movements in and out of GPB in Luxembourg, of the NCC’s Frankfurt and New York correspondent banks, and changes in a third country’s reserves with the western central banks. My hope is they are already scrutinising it intensely; my wish is that they went public with it. In the meantime, I would love to hear what funny Russian financial manoeuvres Free Lunch readers may have noticed.
Other readables
-
What do shipping containers and artillery shells have in common, asks Paul Krugman.
-
Volkswagen has decided to build an electric vehicle plant in South Carolina, and is planning a battery plant in the US for which it thinks it will receive up to $10bn in subsidies. My FT colleagues report that the car company is putting on hold a battery factory project in Europe, to “[wait and] see how the EU would respond to Washington’s incentives before pressing ahead”. My take is that the whiff of blackmail is hard to avoid, given that it makes sense to build in Europe for the undoubtedly expanding European EV market whatever the US does.
-
Martin Wolf is pessimistic about Europe’s ability to sustain its post-national aspirations in a hardening world.
Numbers news
-
El Salvador has built a prison that is supposed to become the world’s biggest by population — and the most overcrowded by design, according to my FT colleagues’ data analysis. The Central American country has the world’s highest incarceration rate.
-
Swati Dhingra, a Bank of England interest rate-setter, does not want rates to go higher, judging that the bulk of UK inflation is trade-related and not caused by domestic pressures.