Safe Assets and Reserve Management

Abstract

Managers of official dollar reserves are bound to pay attention to the debate over safe assets: their investment portfolios operationally define such assets. This chapter argues that reserve managers need not worry about a shortage of safe assets. The debate turns first on whether demand for dollar safe assets is likely to grow as rapidly as emerging market economies. Second, it turns on whether the supply of dollar safe assets can only grow with US fiscal deficits. Neither holds. On the demand side, emerging market economies’ growth does not require dollar reserves to grow at the rate observed in the early 2000s. In retrospect, rapid dollar reserve growth reflected emerging market economies’ response to dollar depreciation. When the dollar cycle turned to appreciation, foreign exchange reserves stopped growing. On the supply side, law and policy extend state backing to various IOUs and thereby make safe assets. US government support for the housing agencies Fannie Mae and Freddie Mac makes their debt into safe assets, albeit with wobbles. US government support for banks, including Federal Reserve liquidity, Federal Deposit Insurance Corporation insurance, and, in 2008, Treasury equity can make US bank liabilities safe. In the rest of the world, government support of non-US banks allows ones from well-rated countries to compete with US banks in selling safe dollar deposits. Moreover, international and interregional organizations, non-US sovereigns and agencies all compete with the US Treasury in selling safe dollar bonds. In allocating their dollar foreign exchange reserves, central bank reserve managers make room for all such competitors. In particular, they invest more than a third of their dollar reserves outside of US Treasury securities.

The author thanks Jacob Bjorheim, Michael Bordo, Claudio Borio, Mark Carlson, Jean-Francois Rigaudy, Jochen Schanz, and José María Serena Garralda for discussion, Marius Cara, Jürgen Köstner, Maximilian Plattner, Ramona Raschke, Siegfried Ruhl, and Irene Sánchez Aizpurúa for help with Table 3 and Bilyana Bogdanova and José Maria Vidal Pastor for research assistance. An earlier version of this paper is available as Bank for International Settlements Working Papers no 769 https://www.bis.org/publ/work769.htm.

Annex: Sources for the Instruments Held in Dollar Reserves

This chapter draws on four sources to assemble the official portfolio of US dollar investments as of June 2017. Three are official and straightforward; one depends on a combination of data compiled by a market source and informed estimation.

The Treasury International Capital (TIC) annual survey of portfolio investments in the USA (US Treasury et al. (2018)) provides the bulk of dollar investments in the USA. In addition, TIC data on the own and custody liabilities of banks in the USA to foreign official institutions form the basis of the Bureau of Economic Analysis compilation of bank-related liabilities.

Similar liabilities of banks outside the USA are reported by the Bank for International Settlements (BIS). These data cover both cross-border liabilities to the official sector and local foreign currency liabilities. The latter would capture, for instance, a dollar deposit of the Bank of England in a (foreign or domestic) bank in the UK.

While the official sector provides these three sources, the final one is market-based. ICE Bank of America Merrill Lynch compiles an index of dollar bonds issued by various official obligors outside the USA of which central banks hold about half. The larger index, the ICE Bank of America Merrill Lynch Foreign Government and Supranational index, contains all such bonds with more than one year of remaining maturity, a fixed coupon and more than $250 million outstanding. These numbered 716 in early December 2018, with an aggregate value of $1.176 trillion. Its short-term counterpart contains such bonds with a year or less to maturity, with $150 million or more outstanding. These numbered 128 in early December 2018, with an aggregate value then of $216 billion. These indices are marketed separately to allow portfolio mandates to exclude the shortest-term bonds.

These sources come with various limitations. On the one hand, the TIC data include sovereign wealth funds like the Norwegian Government Pension Fund in their definition of official investors. As a result, the decomposition of official investments by instrument should add up to more than the estimated global dollar reserves from the IMF, which exclude the holdings of sovereign wealth funds. On the other hand, the US Treasury et al. (2018) survey has difficulty in pinning down the ultimate beneficial owner of US securities that are held by custodians, particularly those outside the USA. This difficulty could lead to an undercounting of officially held US securities.

The estimate of officially held bonds issued by obligors outside the USA is new to this chapter and comes with several caveats. First, the sovereign, supranational, and other official issuers in the index certainly do not exhaust the universe of bonds issued by non-US obligors that are held by central banks. For instance, judging from the TIC data, central banks must hold dollar corporate bonds issued by firms incorporated outside the USA. Second, the proportion of the bonds in the index that are held by central banks is estimated with no great precision based on reports from underwriters in the primary market placement of such bonds, which are provided to the issuers themselves. The author asked the treasuries of some of the most prominent supranational and agency issues for summaries of such reports. Such data were generally publicized in investor presentations.