The positive view here of US economic prospects for late 2017 / early 2018 is supported by additional monetary data in the Fed’s second-quarter financial (flow of funds) accounts released yesterday.

Previous posts have highlighted a strong pick-up in six-month growth of real narrow money, as measured by M1A divided by consumer prices, since early 2017*. The preferred narrow money measure here, however, is non-financial M1, i.e. covering holdings of households and non-financial businesses but excluding the financial sector. This can be derived from the financial accounts but the numbers are quarterly and lag the monthly M1 / M1A data by two months.

Figures through end-June show that real non-financial M1 has accelerated even more strongly than real M1A. Six-month growth, indeed, was the highest since early 2014, ahead of two quarters of 4% plus annualised GDP expansion – see first chart.

Broad money trends, meanwhile, remain stable and solid. The Fed stopped publishing M3 in 2006 but a close substitute can be reconstructed from the financial accounts. Total and non-financial M3 grew by 6.1% in the year to June – second chart.

The financial accounts also include data on sectoral net lending, i.e. saving minus investment. The household and domestic business sectors ran stable surpluses in the second quarter, of 1.5% and 0.6% of GDP respectively – third chart. The business surplus suggests favourable prospects for investment and hiring – no recession since 1950 (at least) began with the business sector in surplus.

The combined household / business surpluses reflect a general government deficit of 4.7% of GDP – worryingly large for an economy at or near full employment. There is, of course, little prospect of this being cut back any time soon and a significant risk of further widening if the Trump administration and Congressional Republican majorities coalesce around unfunded tax cuts.

*M1A = currency in circulation plus demand deposits. M1 also includes other checkable deposits. The two measures have behaved similarly in recent years. M1A is preferred for the historical reason that M1 was heavily distorted by sweep account programmes in the 1990s.