May TIC Data: Still Buying U.S. Assets, but Just the Liquid Ones

These days, the TIC data released monthly by the U.S. Treasury, detailing capital flows to and from the U.S., often seems anti-climactic given sharp moves in the FX and treasuries markets. Yet despite the lag, the data released yesterday, which details May purchases, tells a few stories.

Most importantly, it illustrates the fact that in the face of capital inflows to overheating emerging market economies, the central banks of these countries kept buying U.S. dollar assets in May. Q2 was the first quarter of significant reserve accumulation of the last year. Preliminary estimates from RGE Monitor suggest that reserve accumulation was around $180 billion in the quarter (adjusted for valuation); the first significant increase since mid 2008. As in 2008, China accounts for the bulk of the accumulation (around $140 billion).

Despite supra-national reserve currency rhetoric, and given Beijing’s reluctance to have its currency appreciate, there was little choice but to buy dollars. China added $38 billion in U.S. short and long-term treasuries for a net increase of $26 billion in U.S. short and long-term assets. The discrepancy can be explained by China’s reduction in its U.S. dollar deposits and its continued reduction in agency bond holdings.

Foreign investors, however, shunned long-term U.S. assets. The major foreign buyers of U.S. assets went back to the short-end of the curve, buying T-bills and adding other short-term claims. Total purchases of T-bills by foreign official investors amounted to $53.1 billion.

This shift could help explain why long-term treasury yields rose in May. With concerns looming about the U.S. fiscal position and the dollar’s value, perhaps the move to the short-end of the curve is of little surprise.  It also suggests that the U.S. government is again becoming more reliant on bills for financing, as it did toward the end of 2008. This may not be sustainable in the longer-term.

While a decrease in the U.S. current account deficit means that the U.S. might be less reliant on foreign finance as a whole in 2009, the U.S. has become even more reliant on Chinese financing. China has been the largest reported holder of U.S. treasuries for some months now, and as of May China accounts for 20% of total outstanding foreign holdings of U.S. short and long-term U.S. treasuries, nearly equaling the combined holdings of Russia and Japan.

Since last fall, China dramatically scaled up its purchases of the shortest-term, most liquid U.S. assets. Between July 2008 and May 2009, the country purchased $196 billion of treasuries with maturities of less than one year. In part this might reflect a shift during last fall within China’s U.S. dollar portfolio (it also vastly decreased its holdings of US agency bonds, while slightly adding long-term treasuries). But as the chart below shows, China’s purchases of long-term U.S. assets fell sharply over the last year, and continue to fall:

12 month rolling sums of Chinese purchases of U.S. assets


So why short-term assets? Investing in the most liquid assets could keep funds freer for other purchases, including dollar denominated loans to resource countries. In theory, with shorter maturities, China could allow these assets to expire and not re-purchase them. However, in an environment where Chinese growth is re-accelerating, Q2 is unlikely to be the last in which China receives hot money inflows. As a result, expect further dollar purchases. It is little wonder why Chinese officials were worried about dollar holdings this spring, given how many U.S. assets they were buying.

Beyond China

Like China, Brazil also added short-term claims in May, with $12 billion in short-term claims offsetting net sales of $9 billion in treasury bonds. Short-term treasury holdings rose by almost $10 billion. Brazil has also wanted to diversify its reserve holdings.

Gulf oil exporters likewise added to short-term holdings in May, probably because their local liquidity needs exceeded their concerns about the dollar’s valuation. Given lower oil prices, the region’s sovereign wealth funds have fewer new funds at their disposal. That may be changing slightly, but increases in domestic spending and reductions in oil output limit new available funds. Meanwhile, with a shift from the dollar peg off the table as a policy option, reserve diversification is limited, as the need for dollar liquidity and dollar financing remains high.

Based on the reported data, the GCC has a reported dollar portfolio of about $400 billion–including $140 billion in U.S. equities, which hasn’t budged much in the last two years. Between June 2008 and May 2009, GCC holdings of long-term treasuries increased by about $30 billion, to almost $200 billion total, despite a slight decrease in May. Though, holdings of Agency bonds fell by about $10 billion,  in the last year.

The GCC total dollar portfolio is likely to be significantly bigger–over half of the estimated $2 trillion managed by public and private sector GCC investors. The discrepancy can be explained by the GCC’s tendency to buy through intermediaries. It seems likely, however, that the use of local intermediaries has increased, as the flows of U.S. dollar assets from the GCC have been higher in the last year. But again, the currency pegs may constrain the GCC to dollar purchases.

Japan, Russia and Canada, had notable net sales of U.S. assets in May.  Japan’s shrinking current account surplus could reduce its purchases of U.S. assets.

Canada’s sales may reflect a shift away from government bonds to equities outside of the U.S., in the midst of the rally.

Russia’s net sales, mostly of short-term assets, are a bit more puzzling. Russia has been reducing its U.S. dollar assets for some time, but given the inflows Russia received, one would have expected dollar purchases. In fact Russia’s central bank data on its FX interventions suggests that it bought $18 billion in U.S. dollars in the month of May. One possible explanation is that Russia could be adding to offshore dollar deposits–which would not be captured in U.S. data.