This is the introduction to a series of posts.
the old view
During the first decade of the 21st century and the early years of the following decade, a fairly unanimous narrative developed about foreign capital inflows into U.S. bond markets. This narrative highlights three types of flows which approached or crossed the $1 tn mark and, due their size, not only affected financial markets narrowly but created real economic and social-political spillovers which shaped – and still shape – the world today.
In decreasing order of importance:
1. Foreign central banks started in the late 1990s, animated by recent FX volatility in a number of emerging markets and as part of a renewed adoption of mercantilist, export-led growth policies, to intervene in FX markets and accumulate foreign currencies, primarily U.S. dollars. These funds were then used to acquire high quality U.S. debt securities with non- or negligible credit risk. Such purchases were not the result of a specific duration view or return-enhancing mindset but were rather forced; where else to park several trillion USD. At its height, U.S. debt owned by foreign official accounts reached around $5 tn.
2. Starting at a similar time, structured products began their march into portfolios of institutional investors, riding on a wave of supposedly precise credit ratings and above average yields. While inflows where insignificant during the early 2000s, they seriously accelerated from 2004 when European banks and parts of Asia began to arbitrage versatile risk weightings and started to purchase structures backed most frequently by U.S. real estate, but more or less anything that provided sufficiently reliable cash flows to be sliced into tranches. Holding structures used by foreigners varied, making precise numeric assessment of this trend difficult. The peak holding level stemming from this type of flow seems to have been somewhere around $1 tn.
3. Finally, animated by diversification benefits, foreign private investors also acquired standard U.S. fixed income products: Treasury and Agency bonds as well as non-securitized corporate bonds. Admittedly, underlying incentives for these flows were limited as
- U.S. interest rates were low compared to other nations and reserve flows pushed down term premia, lowering duration compensation even further;
- the dollar declined starting in 2002, disincentivizing currency-unhedged investments;
- U.S. fixed income markets, while the deepest globally, are also the most competitive and credit risk premia was priced very competitively from 2004 onwards until the crisis, while other parts of the world, like EMs or structured products, were offering more appealing opportunities.
For these reasons, growth was limited to ~$1.5 tn during the 2000s; a notable amount but rather insignificant relative to the size of the underlying markets.
a changing landscape
The growth of FX reserves slowed markedly by around 2012 and began to decline in 2014, leading to sales of U.S. Treasuries, once more volatile or easily accessible assets (equity and cash respectively) were liquidated. This was primarily driven by (1) China, where lower growth compared to the previous decades along with a more assertive political regime led to capital outflows and FX debt repayments to align balance sheets to this altered outlook and (2) Saudi Arabia, which is still in the process of adjusting to a world of lower oil prices – and partially does so by drawing down FX reserves. Both of these developments have been widely covered and joined with a generally lower level of appetite for EM assets explain the flat to slightly downward sloping trajectory of FX reserves.
Ownership of private-label U.S. ABS by foreign private institutions today hovers, after peaking at close to ~ $1 tn in 2007, at around $300 bn with maturities and losses accounting for the difference. While a large portion of tranche securities repaid at par, stricter regulatory statutes, a changed perception of credit ratings, lack of new issuance and the resultant lack of depth in markets by deleveraging investment banks impeded a quick reinvigoration of this market segment.
It is the last, previously more or less inconspicuous, section of private foreign demand for regular U.S. debt securities which has been driving inflows during the past years, with growth exceeding $2.6 tn since the start of this cycle in early 2009.
Compared with the older (and possibly completed) flows by foreigners as part of FX reserve accumulation and into structured products, the current trend of inflows into regular bonds by the foreign private sector is perhaps somewhat less well understood. To a degree this is natural; as a chronologically newer, ongoing and thus fluid development, the analytical energy directed at it is lower. However, precisely the continuous character of such inflows and their effects on markets call for a deeper examination – a task this series of posts seeks to address from a number of angels.
The analysis is divided into three posts.
The first section “A different, $2.6+ tn foreign private buyer of U.S. debt” sets out to establish an answer to the simple question of who exactly has been buying U.S. debt securities over recent years and lays the foundation for this series. This process starts at the country level, before proceeding to sectors within countries and, in a number of cases, single institutions or funds large enough to consider in greater detail. In addition to assigning dollar amounts of purchased bonds to a large number of players, the focus is also set on trying to apprehend the underlying institutional realities these investors face and how they conduct and view their acquisitions in light of their overall asset allocation.
“Effects, balance sheets and new linkages” seeks to connect the balance sheets of the bond issuer on one side and the ultimate owner on the other and map the causal links connecting the two. Effects on interest rates, credit and FX crosses are discussed as well as how institutions structure their currency hedging programs and the secondary effects thereof. The section concludes with a functional categorization of buyers based, not as previously on their nationality, but on their funding structure and hedging behavior.
The final section, “TCJA, open FX positions and $1+ tn negatively carrying bond investments”, looks ahead by taking up the categorization established in part two and considers potential drivers for the three most important types of buyers and why their behavior seems set to change significantly in the years to come. Special attention is given to the accounting underlying currency-hedged investments in a low risk premium world, which is discussed at greater length in an extra post, “FX-hedged yields, misunderstood term premia and $1 tn of negative carry investments”.
This series is based on a dozen or so memos initially released during early & mid 2017. Given its production date, the first two posts conclude with year-end 2016 data. Developments in 2017 will be briefly touched on later but are more a less an extension of prior trends, with no new notable actors outside those covered in this series through 2016.