US Recession – The Lead-Lag Problem


The big question that the article from MacroStrategy wants to answer is whether the US is heading for a recession. Furthermore, should there be a recession, would the landing be soft or hard? Unfortunately, a data-dependent Fed will, as always, only realise that it has conquered inflation, with its inherent built-in 12-month lag, when it is already too late to engineer a soft landing.  

Some lagging indicators (wages, employment, job openings and inflation) point to solid growth. However, other reliable leading indicators either indicate a recession (LEIs, 12-month rate rise, yield curve inversion and house purchase mortgage approvals), or that there will very soon be a recession (the real Fed funds rate and exhaustion of excess savings relative to NGDP).


Best brace for a hard landing.

[/vc_column_text][vc_video link=”https://www.youtube.com/watch?v=eGcx5nYj6mM” el_width=”80″ align=”center”][vc_column_text]

The relationship between broad money supply and nominal GDP

Personal consumption expenditure account for 68 % of the US GDP, which would naturally mean that US consumer would seem to have a substantial impact on the economy’s direction. The US’s unemployment rate is at 3.4 %, the lowest since 1953. Furthermore, the roughly 5.7 million unemployed citizens have 11 million job openings to choose from should they want them.

Private sector wages are currently at $ 33 per hour, which is 16.2 % higher than what they were three years ago at pre-pandemic levels, and it is growing at an annual rate of 6 %. Given for the above data a standard work week of 38 hours, the wage number now translates to a yearly salary of 65 thousand dollars.

It certainly shows that with the inflation level still around 6.4 % and the US household sector deposit accounts holding $ 4.5 trillion more than they did pre-pandemic, talk of recession might still be premature.

One of the key effects of COVID was that through QE-funded fiscal disbursement, it sent household savings through the roof. Note this is the first time since World War Two that we have genuinely seen “excess savings”. However, employment, wages, and inflation are all lagging indicators which says a lot about what has happened already and very little about what is to come. Figure 1 below indicates that despite the private sector’s more than 1 trillion dollars in excess savings (blue line), these are now converging and will soon be overtaken by nominal GDP (red line) which is rising at 1.8 trillion dollars per year. The crossover is on track for around August 2023.

Figure 1: US Private, non-financial, money supply and nominal GDP ($bn)

Should the lines cross, the switch into deflation will likely be sudden. The immense money printing over Covid (QE to fund the fiscal largesse) enabled and fuelled inflation. Logically, therefore, once the excess savings have been absorbed into a higher level of nominal GDP, the inflationary impulse will naturally have become exhausted.

Declining money supply

The nominal broad money supply has now declined by about $750 billion since April, about 10% annualised in real terms. The US economy has not experienced a double-digit decline in the real stock of money since 1946/7, which caused CPI to fall from +20% in Q2 1947 to -3% by Q3 1949, triggering the 1948-9 recession. Without the present stock of excess savings to draw upon, the economy would most likely already be in recession, which points to a potentially sudden and swift shock once savings and nominal GDP come back into balance, most likely starting sometime in or around Q3 2023.

Yet the Fed does not see this threat. The 2023 policy course that NY Fed President John Williams laid out a couple of weeks ago explicitly outlined a further +25-75bps of hikes in the Fed funds rate and implicitly incorporates a further money supply drain of up to $1.1 trillion via Quantitative Tightening. It is hard to square this policy approach with the stated desire for a soft landing. Over the last 70 years, 9-out-of-10 rate hikes of 200 basis points in a year have proved sufficient to trigger a recession. Yet the current Fed has already done a 450-basis point increase (red ellipse).

Figure 2: Fed funds rate, twelve-month nominal change (%)

This could be significant because a real Fed funds rate of 2.5 % is usually sufficient to trigger a recession (see red circles and dotted blue arrow in Figure 3 below).

Figure 3: US Real Fed funds rate (%) and recession

Consumer Price Index

An indicator that makes the market feel complacent is the inflation number because of the current 6.4 % print. However, the Fed changed the weightings of the elements within the Index. The weighting for used vehicles was lowered, giving the US a “better” inflation print. The better print comes from the lowered weighting, which took away the fact that used vehicle prices are currently falling drastically. It is estimated that inflation is heading lower, and we are looking at a range between 2.5 % to 3.5 % around August 2023.

Added to the structural downshift in inflationary stimulus stemming from the renormalisation of excess savings in nominal GDP terms, the second quarter and third quarter of 2022 base effects from higher than trend inflation will also come into play over the next six months. In simple terms, once the high inflation numbers of the past year wash out, the number will look a lot lower.

The current trajectory of CPI, on a straight-line extrapolation, will have decelerated to the 2% Fed target by around end-2023. However, it is possible that headline CPI could print as low as 2.5% as early as June 2023 due to base effects (see red circles in Figure 4 below).

Figure 4: US Consumer Price Index

A low inflation print usually comes in sometime after the recession. If the inflation index starts to flatline, it takes twelve months before the ‘end’ of inflation is revealed in the year-on-year data. In 12 of the last 13 recessions, this resulted in inflation bottoming after the conclusion, not the beginning, of the recession (see red circles in Figure 5 below). By the time the Fed realises it has been successful against inflation, in other words, it is invariably too late to engineer a soft landing.

Figure 5: US CPI inflation (%) and recession since 1934

The market is taking its lead from solid inflation, wages, employment and job openings data, but these are all lagging indicators. The annual percentage change in the Conference Board Leading Economic Indicators (LEI) Index has a 100% track record, at least since 1960, and turned negative in July last year. It is now down a definite 6.5 % year-on-year (see red ellipse in Figure 6 below). The Index is made up of 10 different leading indicators to show the health of the US economy. Some of the indicators in the Index are average weekly hours manufacturing, building permits, leading credit index etc.

Figure 6: US Conference Board Leading Economic Indicators index (% change)

If the LEI does not proclaim a recession this time, the false flag will be unprecedented. This, combined with the 90% reliability of the twelve-month change in the real Fed funds rate, the economic predictions are not good. Other historically reliable leading indicators are also flashing red.

Yield curve inversion

Yield curve inversion may pre-date a recession by 12-18 months, but it’s rarely wrong (see red ellipses in Figure 7 below). You must go back to 1966 to find the 10-year/3-month yield curve sending out a false alarm.

Figure 7: US yield curve (10-year UST less 3-month T-Bill, %)

Yield curve inversion does not just happen by chance. It is a display of the Fed’s stated policy intent. An inverted yield curve is the bond market’s reflection, or accurate anticipation, of not only a Fed tightening of more than 200 basis points in a single year (see red line in Figure 8 below) but, consequently, of an imminent recession too. The necessity for tighter monetary policy is brought on by an economy that is running too hot. The sole purpose of higher Fed funds rates are thus effectively demand destruction, aka recession. Whilst a ‘data-dependent’ Fed always hopes for a soft landing, far too much of the data the Fed watches is backwards looking.

Figure 8: US yield curve and 12-month change in Fed funds (%, inverted)


Yet another reliable recession leading indicator is the housing market. When housing affordability deteriorates, recession assuredly follows (see purple circles in Figure 9 below). The record deterioration in affordability (red line) witnessed this cycle (green ellipse) has led to a 41% decline in mortgage applications for house purchases (blue line). From the graph, we can see that it is a bigger drop than the pre-global financial crisis housing bust. Although existing homeowners, especially those who fixed their rates over the last decade, will remain well insulated, without new purchases, the economy loses all the attendant spending on fixtures and fittings that come with any new home.

Figure 9: US housing affordability and mortgage approvals (%)

Employment Rate

The US unemployment rate is currently 3.4 % and at a record low, as indicated by Figure 10 below. The red circles show that the lowest level of unemployment almost always predates the recession, which is indicated by the grey bars. When a recession hits, unemployment tends to rise quite significantly. Thus, it is safe to say that recession causes unemployment. However, it is also important to note that low levels of unemployment do not mean that the economy is robust enough to steer clear of a recession. Unemployment is always at its very lowest just before recession strikes (see red circles in Figure 10 below).

Figure 10: US unemployment rate (%) and recession since WWII

One more argument that can be made in favour of a cyclically robust consumer is the 11 million unfilled job openings, which is nearly two for every one of the roughly 6 million currently unemployed. Some might think it unfeasible to believe that, even if a recession did kick off, that those rendered unemployed could not easily find alternative work with so many jobs to fill. However, as we have already discussed, the excess savings that those who have chosen to stay out of the workforce since COVID have presumably been living off will be used up sometime around August. It thus seems reasonable to expect that in the absence of a recession, the employment ratio (see Figure 11 below) will continue to normalise (see the dotted arrow), back up towards the middle of the post-War trend. A renormalisation of the employment ratio to 62.6%, the mid-point of the trend, would represent about 8.7 million more workers who could be forced back into the jobs market as their savings run out over the next six months.

Figure 11: US employment/population ratio (%)

In short, things in the job market can and do change quickly during a recession; However, employment conditions are always a lagging indicator. The problem which then arises is that at precisely the point when these people need to find work again, it might be exactly when the employment opportunities dry up.

It possibly goes without saying that a recession is typified by a contraction in business deposits (see green circles in Figure 12 below), but what is not so intuitive is that this is often accompanied by an increase in households’ deposits for precautionary reasons (blue line).

Figure 12: Annual change in deposits, households, and businesses (%)

Even before the household sector’s excess savings buffer is used up therefore, which will probably occur later this year, the US personal savings rate (see blue line on Figure 13 below) will most likely start to normalise from the current record low rate of 2.9 % back up towards the 10-year average of 8.5 % (red line), which would place as much as a 1 trillion-dollar squeeze on consumption.

Figure 13: US Personal savings rate (%)

An earnings recession

A normalisation of the savings rate would also make it much harder for firms to raise prices sufficiently to fully absorb increased materials, wages, and finance costs, without losing sales. This matters when US profit margins are at their highest since the 1950s (see green ellipse in Figure 14 below). The 2020 ‘recession’ was an exception that saw profit margins widen, but margins usually implode by between 400 to 800 basis points. Consequently, the last seven recessions witnessed an average real earnings decline of 27 %. It would be more than a little strange if the next recession did not feature something similar. There is just no room for complacency.

Figure 14: US Pre-tax profit margins (%)

It is interesting to note that whilst four lagging indicators (wages, employment, job openings and inflation) may still be pointing to solid growth, at least half a dozen reliable leading indicators are all either already indicating a recession (LEIs, 12-month rate rise, yield curve inversion and house purchase mortgage approvals), or at least there will very soon be a recession (real Fed funds rate and exhaustion of excess savings relative to NGDP).

In conclusion, it follows that the Fed has placed a large amount of emphasis on lagging data, which will slowly start to unwind and catch up with what the leading indicators are showing us. The Fed has been too aggressive in its monetary policy, and signs of cracks in the US economy are starting to show. Even if they stopped right now, they still would not be able to revert a recession. The groundwork for triggering a recession has already been done.


Best brace for a hard landing.