Eurozone February monetary data were extraordinarily negative, suggesting that interest rates were already at a restrictive level before the 50 bp rate hikes in February / March. 

Economic sentiment has lifted in early 2023 in response to a collapsing gas price and China’s reopening but the impact of monetary restriction has yet to kick in. 

The headline M3 broad money measure was down again in February and has fallen in four of the last five months. The six-month rate of change turned negative and is the weakest since 2010 in the aftermath of the GFC – see chart 1. 

Chart 1

Bank deposits are contracting at a faster pace than then because of a portfolio switch into money market funds and short-term bank bonds. This switch has been motivated by relative yields but the banking crisis could give a further boost to money fund inflows. 

Corporate money trends are particularly alarming. Bank deposits of non-financial corporations contracted at a 4.0% annualised pace in the latest three months, with the overnight (M1) component down by 16.6% - chart 2. Household deposits fell in February and are barely up over three months, with a shift out of overnight accounts suggesting weak spending intentions. 

Chart 2

Talk of households still sitting on substantial spendable “excess” savings is suspect. Allowing for inflation erosion, household M3 deposits are below their pre-pandemic trend – chart 3. 

Chart 3

Country deposit data suggest that a core / periphery divergence is opening up, with Spain following Italy into year-on-year contraction – chart 4. 

Chart 4

Monetary weakness partly reflects a collapse in credit growth: three-month loan momentum was running at an annualised 7.6% as recently as September but turned negative in February – chart 5. 

Chart 5

Corporations have been repaying short-term loans in size since November, consistent with a downswing in stockbuilding, which reached a record share of GDP in Q4. Numbers could bounce near term as firms draw down credit lines while they still can.

Current monetary stagnation implies that policy-makers’ worries about a sustained inflation overshoot are as misplaced as their deflation panic in 2020 when money growth was surging.

UK annual broad money growth peaked in February 2021. It should be no surprise that annual inflation was still riding high in February 2023, based on the “monetarist” understanding of a roughly two year lead. 

Annual money growth, however, collapsed after February 2021. Non-financial M4 rose by 2.4% in the year to January and by only 0.9% annualised in the latest three months. 

A consensus concern is that a coming inflation decline will fail to return it to target – one informed commentator expects stickiness at about 4%. No explanation is offered of how such a scenario is compatible with barely growing broad money. Is velocity expected to pick up, against its long-term downtrend? Or is 4% inflation projected to coexist with economic contraction of 3% pa – the implication if money growth runs at 1% pa and velocity is stable? 

The collapse in annual money growth closely resembles a decline over 1990-93, following which annual core RPI inflation fell below 2% in H2 1994, consistent with a core CPI rate (unavailable then) of about 1% – see chart 1. 

Chart 1

The push-back to a similar scenario now is that the unemployment rate is much lower than at the start of the 1990-92 recession. Average earnings growth, however, was significantly higher then – the annual increase in total pay was above 10% (three-month moving average) when the recession started versus below 6% now. Private pay momentum is already slowing despite limited labour market cooling to date – chart 2. 

Chart 2

The 1991-1994 inflation plunge, moreover, occurred despite upward pressure on import prices from a 12% drop in the effective exchange rate between 1990 and 1993 (calendar year averages). There is no such currency headwind to an inflation decline now. 

Annual core CPI inflation rose in February but three-month momentum remains well down from its May 2022 peak – chart 3. Commodity prices signal a coming slowdown in food inflation – chart 4 – while energy prices will soon be falling year-on-year. The February inflation result is irrelevant for assessing 2024-25 prospects and the MPC should ignore it. 

Chart 3

Chart 4

Lending by the Fed to depository institutions jumped from $15 billion to $318 billion between 8 and 15 March - see chart 1 (red line). The emergency loans – mostly via the discount window and via the FDIC rather than under the new Bank Term Funding Program – were the main driver of a $441 billion surge in banks’ reserves at the Fed. 

Chart 1 

These developments do not represent an easing of monetary conditions, except relative to a much tighter baseline that would have resulted from the Fed failing to accommodate increased demand for monetary base due to the banking crisis. 

• Unlike QE, Fed lending to the banking system has no direct impact on money stock measures (i.e. money held by households and non-bank firms). (QE has an impact to the extent that securities are purchased from non-banks.) 

• Unlike QE, the reserves rise is temporary and will reverse if the crisis abates and lending is repaid. 

• The emergency / temporary nature of the lending / reserves rise implies no incentive for banks currently experiencing inflows to expand assets. (QE can have secondary monetary effects by encouraging lending / securities purchases.) 

Resolution of the crisis requires the authorities to arrest broad money contraction. A run-down of the Treasury’s cash balance at the Fed won’t be sufficient; QT needs to be suspended / reversed to offset a cutback in lending by troubled banks. Consideration should also be given to limiting the drain of deposits to money funds, e.g. by capping their access to Fed’s overnight reverse repo facility.

Markets are moving towards the view that the Fed will be forced to suspend or reverse interest rate hikes in response to the SVB crisis but a cessation of QT is a more important requirement for restoring banking system stability. 

QT also caused the 2019 repo rate crisis, which ended only after the Fed restarted securities purchases (Treasury bills) – portrayed, of course, as a “purely technical” measure rather than a return to QE. 

The US weekly broad money proxy calculated here has contracted since April 2022. Weakness initially reflected the US Treasury “overfunding” the federal deficit to rebuild its cash balance at the Fed. QT has been the driver more recently – see chart 1. 

Chart 1

The drain of deposits from commercial banks has been magnified by competition from money market funds, which are able to place overnight funds with the Fed at an interest rate (currently 4.55%) within the Fed’s target range for the Fed funds rate (4.5-4.75%). 

Balances in retail and institutional money funds grew by 5.5% (11.3% at an annualised rate) in the latest 26 weeks, while commercial bank deposits contracted by 2.2% (4.4% annualised) – chart 2. 

Chart 2

In combination, Treasury overfunding, QT and outflows to money funds have resulted in a 30% decline in banks’ reserve balances at the Fed from a peak in December 2021 – chart 3. 

Chart 3

The deposits / reserves drain has caused banks to sell securities and, more recently, restrict loan supply – chart 4. 

Chart 4

The new Bank Term Funding Program will allow banks to avoid selling securities at a loss but fails to address the system-wide loss of deposits due to QT. The Fed facility, moreover, is more expensive than the deposits it may replace. 

Fortuitously, downward pressure on broad money has recently been relieved by a run-down of the Treasury’s cash balance at the Fed, reflecting the debt ceiling impasse. The decline, indeed, may have been accelerated to inject liquidity into the banking system – the balance fell from $345 bn to $247 bn between Wednesday and Friday last week. 

Such relief, however, is temporary. The authorities’ actions to date may be sufficient to avert another bank collapse but the banking system will remain under pressure, with negative economic implications via rising deposit / lending rates, until QT-driven monetary contraction ends.

A Eurozone recession can now be ruled out, according to the ECB and a PMI-hugging consensus. 

Someone forgot to tell the monetary data. 

The favoured Eurozone narrow money measure here – non-financial M1 – fell for a fifth consecutive month in January, while broad money – non-financial M3 – was unchanged following marginal gains in November / December. 

With inflation data remaining hot, six-month contraction of real narrow money (i.e. deflated by consumer prices) reached a new record, of 5.4% or 10.5% annualised – see chart 1.

Chart 1

A post last month noted that UK sectoral money trends were displaying a recessionary pattern: corporate broad and narrow money holdings were falling in nominal terms, suggesting a cash flow squeeze, while households were moving large sums out of sight deposits into time deposits, consistent with a shift in consumer behaviour from spending to saving. 

The same trends are now on show in the Eurozone: corporate M2 and M1 deposits fell in the three months to January, as did household M1 deposits – chart 2. 

Chart 2

The no-recession bandwagon gained momentum following Eurostat’s flash estimate that Eurozone GDP grew by 0.1% in Q4. Recently released national details paint an uglier picture. 

The Q4 fall in German GDP was revised from 0.2% to 0.4%, which will feed into an updated Eurozone number next week. 

More significantly, expenditure breakdowns show that domestic final demand weakened sharply in Q4 in France, Germany and Spain – at annualised rates of 1.6%, 3.7% and 5.4% respectively. The GDP impact was cushioned by a rise in net exports driven by import weakness and a further increase in stockbuilding – charts 3-5. 

Chart 3

Chart 4

Chart 5

Eurozone stockbuilding, therefore, appears to have risen further from its record (in data since the mid 1990s) share of GDP in Q3 – chart 6. A violent reversal from lower peaks in 2007 and 2011 was a key driver of the 2008-09 and 2011-12 recessions. 

Chart 6

The no-recession narrative was bolstered by February PMI results showing a pick-up in Eurozone services activity and new business. Manufacturing new orders, however, remained contractionary and are a better guide to the cyclical trend (since the key economic cycles – stockbuilding, business investment and housing – involve goods demand; there is no independent services cycle). 

The move off the lows in manufacturing PMI results has been mirrored in the German Ifo manufacturing survey. Business expectations, however, remain weak by historical standards and an indicator of demand inflow has risen by less, stalling between December and February – chart 7. 

Chart 7

Residential construction expectations, meanwhile, plumbed another record low in February. Survey weakness has been reflected in hard data: housing construction new orders in Q4 were down 35% from Q1 and the lowest since 2014. Dwellings investment was a drag on GDP during H2 2022 but the orders plunge suggests a further big negative impact to come – chart 8.

Chart 8