The global real narrow and broad money measures followed here continue to grow solidly, suggesting a favourable economic outlook. Some commentators, however, argue that US dollar strength is tightening global liquidity conditions, warranting a bearish view of economies and markets. Are they right?
According to these commentators, capital reflux into the dollar is forcing central banks in emerging economies to intervene to support their currencies. This intervention amounts to “reverse QE” and has a contractionary influence on domestic monetary conditions. The negative impact can be measured by the rate of decline of US dollar reserve holdings of foreign central banks. The latter can be proxied on a timely basis by Federal Reserve custody holdings of US securities for foreign official and international accounts*. These dropped by $112 billion between the end of September and 11 March (weekly average data).
The theory underlying this argument is questionable. A flow of portfolio capital into the US will reduce the money supply in the rest of the world but raise it in the US, resulting in no change at the global level, other things being equal**. As noted in a recent post, US money growth has strengthened recently.
The current post, however, focuses on empirical evidence, asking whether US dollar reserves, as proxied by the Fed’s custody holdings, have been a better leading indicator of the economic cycle than the global narrow money (M1) measure tracked here. The main findings are:
- Real custody holdings led major swings in the global economy over the 1970s-1990s.
- However, global real narrow money was less volatile and more reliable in predicting recessions.
- The forecasting performance of real custody holdings has deteriorated since the early 2000s.
- Falls on the recent scale occurred in 2011 and 2014 without an ensuing recession.
The suggested conclusion is that the recent reserves decline would need to accelerate and be reflected in a slowdown in global real narrow money to warrant serious concern about economic prospects.
The first chart below shows the six-month rates of change of G7 industrial output and real narrow money over the last 50 years, with global recessions highlighted by shading. A consistent leading relationship is apparent, with real money contracting before each of the six recessions.
The second chart superimposes the rate of change of real Fed custody holdings***. This series also led the economic cycle reasonably well until the early 2000s but was more volatile than the change in real money and gave more false recession signals.
Unlike narrow money, real custody holdings did not contract before the 2008-09 recession. Fluctuations in recent years have been small by historical standards, with declines also occurring in 2011 and 2014.
A possible reason for the less impressive forecasting performance of real custody holdings since the early 2000s is that emerging economies have shifted away from exchange rate management as a key tool of monetary policy. This argues for monitoring monetary conditions directly via the money supply rather than by proxy using US dollar reserves.
*Foreign official holdings of US Treasury and agency securities totalled $4.6 trillion at end-December 2014, the latest available month, according to US Treasury data. The Fed’s custody holdings were $3.3 trillion as of the same date, implying about 70% coverage.
**A portfolio inflow will improve the US basic balance of payments (i.e. current account plus net direct / portfolio inflows), other things being equal. Overall balance implies an offsetting change in banking flows, i.e. a rise in banks’ net external assets. According to the monetary counterparts arithmetic, such a rise has an expansionary impact on the money supply.
***Fed custody holdings are available from 1972. The earlier series shown refers to foreign official holdings of US government securities, from the quarterly flow of funds accounts.
