M2 update: Ok for now, but the Fed should ease

This blog has the distinction of periodically tracking and analyzing the M2 measure of money supply for almost 2 decades. Unlike most other analysts, I focus not just on the supply of money but also on the demand for money. For background, see this post from October 2020 in which I noted rapid growth in M2 but was not alarmed given the huge increase in money demand at the time.

The key to understanding the relationship between money and inflation is to not lose sight of the demand for money. Milton Friedman famously taught us that inflation happens only when the supply of money exceeds the demand for it. When the M2 money supply accelerated in 2009 and 2020-21, it did not immediately ignite inflation because the demand for money also accelerated—with Covid shutdowns it was quite difficult to spend money. But when the demand for money began to decline in early 2021, the money that had been stored in bank deposits and under mattresses began to be released into the economy, and this was the fuel for rising inflation in 2021 and into early 2022. By mid-2022 the Fed had responded to rising inflation by raising interest rates and slowing the growth of M2, thus bringing money supply and demand slowly back into balance. This gave me the confidence in the summer of 2022 to predict that the peak of inflation was now past and it would begin to decline. Since then I have continued to believe that inflation would remain low, given that although the demand for money was declining to pre-Covid levels, M2 growth remained very slow.

The charts that follow are updated with the latest data. Fortunately, there has been no significant change in monetary conditions, and that is why I continue to see inflation remaining relatively low and stable.

Chart #1

Chart #1 shows the level of the M2 money supply. From 1995 to late 2019, it grew by about 6% per year. During that same period, inflation registered about 2% per year. M2 growth then exploded in early 2020, fueled by $6 as $6 trillion of transfer payments were sent to the public to compensate for Covid-related shutdowns—money that the Fed essentially "printed" and helicopter-dropped into the economy. M2 began to slow in mid-2022 as the Fed began to raise interest rates. Today M2 is only $1.3 trillion above where it might have been had we not suffered from the government's disastrous Covid policies and the Fed's failure to react to huge swings in money demand in a timely fashion.

Chart #2

Chart #2 shows the 6-mo. annualized rate of change in the M2 money supply. Growth has been below 6% since April '22, almost 4 years. Over the past year, M2 has grown only 4.9%. Milton Friedman would be pleased.

Chart #3

Chart #3 shows the the ratio of M2 to nominal GDP, which is arguably a good proxy for money demand. Think of it as the percentage of our annual income that we collectively prefer to hold in the form of readily spendable money (currency, checking accounts, time deposits, CDs, and money market funds). Money demand appears to have returned just about to where it was before the Covid era. Both M2 growth and money demand have now returned to "normal" levels, and they are roughly in balance. This is very good.

Chart #4

But, you object, inflation is still above the Fed's 2% target. Yes, the Core Personal Consumption Deflator is up 3.1% in the year ending January '26, and the overall Personal Consumption Deflator is up 2.7%. But the CPI is up only 2.4% as of February, and the CPI less shelter is up only 2.2%. We're not talking about a meaningful overshoot. Of course, I would prefer to see 0% inflation, but what we have today is the next best thing.

Chart #4 shows the 3 major components of the PCE deflator. Note that durable goods prices today are lower than they were in late 2022, and non-durable goods prices have only increased 2% since mid-2022 (which works out to an annualized rate of only 0.6%). The major source of inflation in recent years comes from service sector prices, which are heavily influenced by shelter costs and wages. As I've shown numerous times in recent years, the government has been systematically over-estimating shelter costs. The PCE deflator doesn't have an ex-shelter version like the CPI does; if it did it would likely show inflation being much closer to 2% than to 3%.

••••••••••••••••

So what about the Iran war and soaring fuel prices? Yesterday I paid $7.09 per gallon to fill my tank. Yikes! Surely this will upset the inflation apple cart, no?

Most likely we'll see an upward bump in the official inflation numbers in the months ahead, given the recent increase in energy prices. But that's not how inflation works. With the economy on a strict diet of 5-6% money growth, there is no reason to expect that the entire price level will jump higher. Energy prices will go up, but other prices will come down. Already we're seeing non-energy commodity prices softening. And with the recent uptick in 30-yr mortgage rates and the surge in new home sales for sale, real estate prices are more likely to decline than to rise. 

Most important, however, is this: wartime uncertainties are very likely to spark an increase in money demand. Without a corresponding increase in money supply—and especially since the Fed shows no willingness to lower interest rates—the economy will soon feel starved for money. That's deflationary. We won't see deflation for several months, but don't be surprised if it happens later this year.

My advice to the Fed would be to relax monetary policy soon in anticipation of a meaningful increase in money demand. But sadly, the Fed is usually reactive and rarely—if ever—proactive. 

Jobs and War: the Fed needs to ease

The February jobs report was a lot weaker than expected (-92K vs an expected +130K). But viewed from a broader perspective, it's just more of the same slower growth that we have seen over the past year. Jobs numbers are notoriously volatile to begin with, and on top of that, in February most of the country was hit by a huge winter storm while a nursing strike resulted in a loss of 28K health care jobs.

Chart #1

Chart #1 shows the monthly change in private sector jobs (-86K in February). These are the jobs that really count. Notice how volatile they have been in recent years. Almost every move up or down in the monthly numbers has been reversed in the subsequent month, and after-the-fact revisions are frequent and can be huge. I've been arguing for many years that you can't draw conclusions from one month's jobs numbers—you have to look at the underlying trend in the numbers. 

Chart #2

Chart #2 is a more the sensible way to look at these data: how do they change on a year over year and a 6-mo. annualized basis? The 6-mo. annualized change in private sector jobs has been consistently low (between 0.2% and 0.4%) since last June. Viewed from this perspective, today's number was just more of the same slow growth that we've been seeing since last summer. Closing the border has resulted in a big slowdown in jobs growth, but not a crash landing. The economy is most likely experiencing a soft landing which will be followed by a pickup in growth once the uncertainties of the Iran War are resolved, and the world learns to love AI-fueled productivity.

As we await further developments on these fronts, the following charts are "green shoots" that augur better times ahead.

Chart #3

Chart #3 shows an index of housing affordability (a function of interest rates, home prices, and incomes). For most of the past four years housing has been very expensive for almost everyone, thanks to high interest rates, soaring home prices, and modest real income growth. Fortunately, the most recent datapoint suggests that things are beginning to turn for the better, albeit slowly.

Chart #4

Chart #4 compares the level of 30-yr fixed mortgage rates to an index of new applications for mortgages (i.e., excluding refinancings). Here it is easy to see how high mortgage rates depress the demand for new mortgages, and how lower mortgage rates in the past six months have resulted in a modest uptick in new mortgage applications. Things are improving on the margin, but it's going to take a long time before we see a boom in the housing market. Too many people are still locked into 3% mortgages and are thus reluctant to give them up; at the same time, many millions of new buyers are locked out because mortgages are still quite expensive. 

One potential source of optimism: Trump is said to be considering the elimination of the capital gains tax on real estate. This would likely result in more homes for sale and for lower prices, since sellers could lower their asking price knowing that they won't have to pay a capital gains tax. 

For that matter, Trump ought to go the full nine yards and reduce or eliminate the capital gains tax on all asset sales. At the very least he should allow people to index their cost basis for inflation. Taxing inflation gains is morally unjust. Ah, you say, but wouldn't lower capital gains taxes result in a huge increase in budget deficits? No, on the contrary! It's important to remember that the capital gains tax is the only tax that one can legally avoid—forever—simply by not selling an appreciated asset. Cutting the capital gains tax would likely result in a surge in capital gains tax collections. I for one would celebrate this by selling some highly appreciated stock in order to redeploy the funds elsewhere. 

Cutting or eliminating the capital gains tax would provide a powerful boost to economic growth by making homes more affordable and by freeing up capital everywhere that is locked into appreciated assets. 

Chart #5

Chart #5 compares the rate on 30-yr fixed mortgages to the level of the 10-yr Treasury yield, which traditionally has been the main determinant of mortgage rates. Two factors are behind the lower mortgage rates of recent months: lower Treasury yields and lower spreads. Since the Covid crisis the spread between the two has been high and volatile, but it has returned to relatively "normal" levels over the past year. Thus, to get a meaningful decline in mortgage rates going forward, Treasury yields are going to have to do the heavy lifting. The case for lower Treasury yields depends on lower inflation expectations and an easier policy stance from the Federal Reserve.

Consider this: the Iran War has caused a huge increase in uncertainty in the world, not least by interrupting the flow of oil. Any increase in uncertainty tends to increase the demand for money, because people on the margin try to reduce their risk in exchange for cash or cash equivalents. If the Fed does nothing to change the supply of money (such as by lowering interest rates), an increase in money demand will result in a tightening of monetary policy. That in turn will create deflationary pressures and likely disrupt or slow economic growth.

The Fed should not worry that higher oil prices will be inflationary; they should worry that Iran War uncertainty that is not offset by easier monetary policy will be contractionary. 

Chart #6

Chart #6 shows the latest ISM survey of service sector purchasing managers. This is arguably one of the most bullish indicators of late. With 3 stronger readings in the past 3 months, it's a good bet that the economy's largest sector is improving.

It pays to remain optimistic.

UPDATE (3/13/26): Rising uncertainty and a general increase in volatility are working to increase yields across the board. Spreads between 30-yr fixed mortgages and 10-yr Treasuries are already so compressed that 30-yr fixed mortgage rates have nowhere to go but up in this environment: in fact, they recently ticked higher, from a multi-year low of 5.98% to 6.11%. The Iran War is now working to undermine the US housing market. 

All things considered, the outlook is getting brighter

Here's a selection of charts that I'm paying special attention to these days. If there's a theme emerging from these snapshots of the economy and financial markets, it's a generally upbeat one, featuring continued low inflation, moderate and improving growth, and a gradual improvement in the fiscal outlook. 

Chart #1

Chart #1 shows the 6- and 12-month growth rate of private sector jobs—the ones that really count. From a long-term perspective, it looks like the economy is growing at an unusually slow pace—barely a crawl. But considering the massive deportations of illegals and the virtual closure of our borders, it's not surprising. Some will argue that this shows an economy hovering on the brink of "stall speed," but I think that's an analogy that doesn't really apply to an economy. Economies have a strong tendency to grow if they are not burdened by abrupt and unforeseen changes in monetary and fiscal policy. Meanwhile, it is comforting to see that new claims for unemployment show absolutely no sign of any fundamental deterioration in the outlook for business profits. Firings are low and stable. The worst that can be said about jobs is that there aren't many new jobs being created. But that could be improving.

Chart #2

Chart #2 shows the monthly change in private sector jobs, which appears to have picked up in the past month. It's too early to claim victory after just one month of improvement, but some of the following charts also show recent improvement.

Chart #3

Chart #3 shows the results of a monthly survey of purchasing managers. It has shown pretty lackluster levels for the past several years, followed by an exceptionally strong January report. This could be the first sign of a long-expected surge in the outlook for business manufacturing activity. Historically, readings below 47 have frequently coincided with recessionary conditions. Readings north of 50 almost always occur during periods of decent economic growth. If the recent report is not reversed, then we could be on the verge of a very welcome growth spurt.

Chart #4

Chart #4 shows the level of the federal government workforce. The past year has been dominated by an astonishing 11% decline in the federal workforce—the by-product of Trump's aggressive attack on the Deep State, in particular the Dept. of Education. As the chart also shows, the federal workforce today is as small as it has been since 1966! There are now 324K fewer regulators of the economy. A small government workforce equates to a significant reduction in regulatory burdens, which are typically a cost that most businesses would be happy to avoid. This frees up resources for productive activity, and this is an unalloyed Good Thing. NO ONE came close to predicting that the federal government workforce would ever decline by as much as it has in the past year.

Note that the spikes in the data that occur every 10 years correspond to the temporary hiring of people needed to conduct the Census. 

Chart #5

Chart #5 shows the year over year change in the Consumer Price Index, compared to a version of the CPI that excludes shelter costs (which make up about ⅓ of the CPI). The gap between the two that shows up in the past several years is the direct result of an over-estimation by the BLS of shelter costs. That problem is now behind us, and we are left with inflation that is only moderately above 2%. 

Chart #6

Chart #6 focuses on shelter costs (which are called Owner's Equivalent Rent). This is what you would be paying if you had to rent the house you live in, and it is something that the BLS manufactures inside its computers, since nowhere does there exist such a thing in the real world. Here we see that in the past month and past 3 months, the annualized change in this measure of shelter costs has fallen back—finally—to what it has tended to average in normal times. That explains why the gap between the two lines in Chart #5 has disappeared. 

Chart #7

Chart #7 has been featured regularly on this blog for the past several years. It is designed to show that the BLS's calculation of shelter costs (OER) is highly correlated to the year over year change in housing prices (blue line) from 18 months prior. It took longer for OER to fall (given the prior decline in housing price inflation) than I thought, but the two lines are now back in sync at a level that is consistent with 2% overall inflation. I note further that national home prices rose only 1.2% last year, and, adjusted for inflation, they are 2.2% below their mid-2022 peak. Housing prices are thus quite likely to exert downward pressure on inflation in the months to come. 

Chart #8

Chart #8 shows the 6-mo. annualized growth rate of the all-important M2 measure of the money supply. M2 rose on average by about 6% per year from 1995 through 2019, a period during which the CPI rose by 2% or less per year. More recently, M2 growth in the year ending January '26 was 4.3%, but as the chart shows, growth on the margin has slowed to a mere 3.6% (annualized) over the past six months. In normal, pre-Covid times, most economists would have predicted that 3.6% M2 growth would lead to a slowdown in economic growth and a decline in inflation. Yet today the chatter is all about whether inflation is going to rise. This could be one of those times when the market is caught looking in the wrong direction.

Chart #9

Chart #9 shows what I consider to be a good measure of money demand: M2 divided by nominal GDP. This effectively measures how much of our annual income we like to hold in the form of readily-spendable money (currency, checking accounts, retail money market funds, etc). Money demand today is pretty much the same as it was prior to COVID. The Covid years were characterized initially by a gigantic increase in the demand for money, which was then followed by a return or "normal" levels. My interpretation of all this is that monetary conditions are just about right: money supply is matched by money demand, and that equates to the absence of monetary imbalances that might fuel higher inflation. 

Chart #10

Chart #10 compares the strength or weakness of the dollar (a rising blue line equates to a weaker dollar, while a declining blue line equates to a stronger dollar) with the inflation-adjusted level of a basket of basic commodity prices. Note the very strong correlation of these two variables over time—except for the period following the Covid crisis, when supply chains were disrupted but the consumers wanted to ramp up their spending. Commodity prices back then were much stronger than the strength of the dollar would have predicted, and they helped fuel rising inflation. Recently, however, the two variables have come back into line with each other. Commodity prices are no longer "too strong," while the dollar is still reasonably strong itself. Conclusion: commodity prices are not a source of inflation these days. Need I add that gasoline prices today are very close to their 20-year average ($2.88/gallon), and that they have fallen almost 40% from their 2022 high? Moreover, crude oil prices today are actually almost 10% lower than their 20-year average. 

Chart #11

Chart #11 shows the level of real GDP growth as it compares to two different trend lines (it's important to note the use of a log scale on the y-axis, which makes it easy to see growth trends). GDP rose by about 3.1% per year from 1966 through 2007, but since 2010 it has managed to grow by only a little more than 2.3% per year. Growth may be picking up of late, as noted in the comments above, but even if it does, the economy is still a lot less dynamic than it has been during most of our lifetimes. I think the "disappointing" growth of GDP since 2010 is due to several factors: 1) the $16 trillion the world spent on futile attempts to prevent global warming, 2) a huge increase in US transfer payments, and 3) a significant increase in regulatory burdens. All of which, I'm pleased to note, are in the process of reversing!

Chart #12

Chart #12 shows the level of federal government spending and revenues over the past 36 years. The difference between the two lines is, of course, the federal deficit, which is currently running at about $1.7 trillion per year. Note also the huge surge in federal spending triggered by the Covid crisis—that consisted of approximately $6 trillion in transfer payments which were effectively monetized by the Fed. It was all a nightmare, but it's fading rapidly. Note that spending has not reached new highs since 2022, and it has been flat for more than a year. The fiscal outlook is definitely improving.

Chart #13

Our national debt is just shy of 100% of GDP ($31 trillion), but we are not on the brink of a fiscal abyss. On the contrary, it is not unreasonable to think that Congress can manage some degree of spending control, and it is not the case that the economy faces a crushing burden of debt in any event, as Chart #13 shows. The true burden of debt is not the size of our national debt, but the cost of servicing that debt as a percent of our national income. Today that burden is significantly less than it was during the 1980s, mainly because interest rates are far lower than they were back then. If Congress exercises even modest restraint and the Fed doesn't have to raise interest rates (which they won't have to if inflation remains under control), then we can gradually reduce our deficits and the burden of our debt

All things considered, things don't look so bad at all!

UPDATE (2/27/26)

The Producer Price data for January '26 were released today. There are a lot of different versions of this data, (core, ex-energy, final demand, etc.) but I typically just focus on the aggregate index, which is shown in the chart below. From mid-2022, when inflation pressures generally peaked, the PPI is up at a 0.5% annualized rate through January '26. In other words, price pressures at the producer level have virtually ceased.

Chart #14