[Salon] Confiscation of Russian Reserves

https://ep.ft.com/permalink/emails/eyJlbWFpbCI6ImFlMDFmZWI1YjhjOWM5NTFmOGI1YjNiY2E5ZDkwM2Y2YzFjZiIsInRyYW5zYWN0aW9uSWQiOiJhMDJhYWYzZS1hY2M0LTRlMmYtYjQ4Ni1jZjUzZjNlNWEyOGMifQ%3D%3D

Martin Sandbu

European Economics Commentator

‌

PREMIUM

January 4 2024

Happy new year! I hope you all had a restful break; thanks to my colleague Valentina for holding the fort over Christmas. 2024 brings a new FT newsletter — Claer Barrett’s six-part guide to sorting out your financial life. Sign up here!

2024 also looks important for one topic that Free Lunch follows with interest: it seems ever more likely that the western sanctioning coalition may finally take title of the Russian central bank reserves it blocked in the early days of Vladimir Putin’s full-scale assault on Ukraine. The idea is to transfer the money to Ukraine, enforcing the obligation Moscow has to compensate for the damage it has inflicted. Below, I guide you to the latest reporting on what is happening, before doing the one thing I can contribute at the moment: try to make sense of what public knowledge there is of how much has been blockedand where, given the blocking jurisdictions’ unforgivable unwillingness to comprehensively publish their information.

I have dug into spreadsheets so you don’t have to

Until now the coalition has shied away from confiscation. But the FT has recently reported that US diplomats are circulating legal arguments for confiscation and have proposed that the G7 make preparations for a political decision on the two-year anniversary of the full-scale invasion.

Two factors support this change in attitude. One is that both the EU and the US are encountering political obstacles to quickly mobilising funds for Ukraine. Two weeks ago, IMF head Kristalina Georgieva warned that this was putting Ukraine’s impressive economic recovery at risk. At the same time, the legal and economic debates are largely concluded — all the arguments are now well understood, and all that remains is to make a political decision one way or the other.

For those who need a quick overview of where the arguments stand, the Kyiv-based Centre for Economic Strategy has published two short reports on the legal and economic aspects of confiscating Russian state assets. (Before Christmas, I took part in the centre’s online launch as well as in an Atlantic Council webinar on asset confiscation in the context of the dollar’s role.) The definitive statement of the legal case for confiscation as a “collective countermeasure” under international law against Russia’s illegal actions is a recent briefing paperby a global group of lawyers convened by Philip Zelikow. And in this week’s FT, Simon Hinrichsen sets out the many precedents for valid compensation claims for war damage and confiscation of assets to satisfy them.

For what it’s worth, my sense is that the politics is turning decidedly towards confiscation. (I am told David Cameron’s return to government as UK foreign secretary has emboldened Washington in spurring on the rest of the G7.) But I remain worried about the unexplained unwillingness of governments to make public the exact size, composition and geography of the Central Bank of Russia’s reserves. (Indeed, I remain worried about why it took western countries so long even to decide to collect and centralise this information, even internally to their governments.)

There are basically three public sources on the CBR’s assets. One is the CBR itself, in particular its 2021 annual report that gives figures for the end of 2021 (I first dipped into CBR data last year, but that was not based on the data closest to the full-scale invasion that is now available). The second is western governments, a few of which have produced statements, sorely lacking in detail, of CBR assets in their jurisdictions. The third is Euroclear, the Belgian securities depository that holds securities in custody for the CBR and where cash from those securities has accumulated because Moscow’s access to the holdings has been blocked. (I explained the Euroclear arrangements here.) The company provides useful, if limited, quarterly reports, of which the latest is here.

The good news is that we can compare the information from the different sources and be confident that when they agree, they probably reflect the truth. In addition, it bears noting that the CBR was highly respected by the global central banking community before the full-scale invasion, and its published numbers from that period are probably as reliable as those of other central banks.

The public sources allow us to make some “triangulations” between them. I have done this based on a spreadsheet (email me if you would like a copy). These are the comparisons I find most instructive:

The CBR data means $350bn worth of reserves have been immobilised in sanctioning jurisdictions. What has been made public about Euroclear is consistent with this total and the CBR’s reported allocation between instruments.
The CBR’s breakdown of its assets by currency denomination aligns closely with that by the jurisdiction of the counterparty or security issuer. In particular, its reported euro-denominated reserves are exactly equal to the total claims on counterparties in France, Germany, Austria, and “others” ($207.8bn, or €182.5bn at the exchange rate of December 31 2021). So “others” very probably comprise the rest of the eurozone plus Switzerland.
(Switzerland has declared SFr7.4bn (about €8bn) of immobilised CBR reserves there, but the CBR reports negligible SFr-denominated holdings. The most likely explanation is euro-denominated claims on Swiss banks.)
Also, the amount of Canadian dollar-denominated claims (US$19.6bn) is only slightly more than the claims on Canadian counterparties (US$16.5bn).
In contrast, the CBR reports US$-denominated claims of about $28bn more than the amount of claims on US-based counterparts ($67bn vs $39bn), and sterling-denominated claims of about £5bn more than the claims on UK-based counterparts. It reports Japanese yen-denominated claims of about $20bn-worth less than the amount of claims on Japanese residents. This implies similar magnitudes of non-yen-denominated claims on Japanese counterparts, which could account for most of the USD and sterling gaps.
Three public sources — the growth in Euroclear’s balance sheet, its currency breakdowns of Russia-related assets, and Belgium’s announcement that a total of €191bn-worth of CBR reserves is immobilised in the securities depository — together imply the following: all the sterling and Canadian dollar-denominated claims are accounted for through Euroclear (as well as the tiny Australian- and Singaporean-dollar claims), but none of the $36bn yen-denominated ones, only a quarter of the $66bn US$-denominated ones, and only two-thirds of the €182bn euro-denominated ones.
The reported €191 total of assets held at Euroclear amounts to about four-fifths of the entire CBR’s reported holding of securities ($236bn worth of government securities and $28bn worth of non-government ones) as opposed to other assets (deposits and a small amount of claims on international organisations). Depending on the exchange rate, the CBR must hold $50bn-$60bn worth of securities outside of Euroclear according to its reporting. We cannot tell if this is in sanctioning jurisdictions or in China. But if the CBR’s allocation of reserves between securities and deposits is about the same in both, nearly all of the remaining securities are in China, so nearly all the reserves immobilised outside of Euroclear must be in the form of deposits in the sanctioning countries’ central banks and commercial banks.

These conclusions reveal some inconsistencies that raise important questions:

The best estimate of the amount of reserves that have been blocked by the sanctioning coalition is about $350bn (at end-February 2022 exchange rates). That is significantly more than the $300bn approximation that is often thrown around in policy discussions. It also means that more than one-third of the assets available for confiscation are not in custody at Euroclear. But my excellent Brussels colleagues report that internal documents in the European Commission operate with a number of only €260bn, or about $280bn at today’s exchange rate. Why is this number so much lower? What and where is the missing $70bn?
The CBR reports $40bn in claims on US counterparties. At most $16bn seems to be through Euroclear (and probably less, since some US$ claims will be on non-US counterparties), leaving at least $25bn unaccounted for. Yet my colleagues report that G7 officials operate with a much lower $5bn immobilised in the US. Why the difference?
Where, and in what form, is the $36bn-worth of yen-denominated assets? Without further information, we should assume deposits in the Bank of Japan and Japanese commercial banks.
Where, and in what form, are the claims on eurozone (plus Switzerland) counterparts not held in Euroclear? They amount to at least €60bn, or somewhat more if any of the Euroclear-held securities are euro-denominated claims on non-eurozone issuers (most likely Japanese, given the gaps mentioned above), and are in all likelihood mostly deposits. The French finance minister has previously said €22bn was immobilised in France, which his ministry confirmed to me last October. But in December, my colleagues reported €19bn and that national and commission authorities operate with negligible numbers in other eurozone jurisdictions. That leaves more than €30bn unaccounted for.
CBR data implies that Russia accumulated virtually no additional central bank reserves in the first two months of 2022, even though energy prices were already high then.

I finish by noting three points. First, the role of Euroclear makes it a focal point in the discussions — but this has allowed governments (including non-eurozone ones) to brush aside the remaining questions about assets in their jurisdictions.

Second, the maturing of securities blocked at Euroclear has produced an accumulation of cash there — but since this cash is kept in the currencies of the securities that produced it, the non-euro portion must be held by Euroclear through subsidiaries or correspondent banks resident in the US, Canada, and the UK (and Australia). That money is not beyond those governments’ legal reach despite Euroclear’s custodial role: they could seize it.

Third, as I describe above, just the discrepancies in the available information are much larger than the funding packages that the EU and the US are struggling to pass. That alone should show the imperative of finally putting all the information into the public domain.