UK real money contraction warned of 2022 economic stagnation and 2023 recession. Weakness has abated but real money measures have yet to resume expansion, casting doubt on hopes of a sustainable economic recovery.

The latest ONS numbers are consistent with a recession having started in Q2 2023. Among key features of the GDP release:

• Gross value added (GVA) at basic prices peaked in Q1 2023, falling by 0.03% in Q2, 0.16% in Q3 and 0.34% in Q4.

• The cumulative decline in GVA / GDP of 0.5% between Q2 and Q4 is inconsistent with a description of the economy as “flatlining”.

• Similarly, claims that the consumer has been holding up are no longer tenable given a 1.0% cumulative contraction in household consumption between Q2 and Q4.

• GDP / GVA fell by 0.2% and 0.3% respectively in the year to Q4, meeting a stronger recession definition than the two-quarter rule (in contrast to Japanese GDP also released today).

• Nominal as well as real GDP fell in Q4, with the GDP deflator rising at a 2.0% annualised pace between Q2 and Q4.

The suggestion of cyclical peak in Q1 2023 is supported by the LFS employment measure, which reached a high in the three-month period centred on March. (The LFS aggregate is 10% larger than the PAYE employment series, reflecting coverage of self-employment.) Aggregate hours worked also peaked then, falling 1.5% through November.

Real money measures began to contract in H2 2021. GDP stagnated from Q2 2022, consistent with the usual lag. The six-month rates of decline of real narrow and broad money reached a peak in March 2023, warning of H2 economic contraction – see chart 1.

Chart 1

Six-month real money momentum has recovered significantly but has yet to turn positive. Slowing inflation has been a key driver, while nominal broad money is no longer contracting. Economic weakness may abate in H1 2024 but current monetary trends appear inconsistent with a meaningful recovery. Early rate cuts are urgently required to limit still-significant downside risk and head off an extended inflation undershoot.