# What drives the difference between M1 & M2 money supply (in the US)?

From what I understand the only entity that controls M1 in US is the Federal Reserve. Is it true that M2-M1( M2 minus M1; the part of M2 that is NOT in M1 like timed deposits) is controlled by the commercial banks?

Edit (by JS):

I think the question is of interest and of importance in current market conditions. I attach a picture that I have commonly encountered over the past 12 months: the market capitalization of the S&P 500 vs. M2 money supply.

These types of charts have gained popularity since the FED started massive balance sheet expansion in March 2020 to combat the market volatility induced by the Covid pandemic: in a way, such charts provide a "rational" explanation for the "irrationally high" stock market levels.

I would myself be interested in a thorough answer to the OP's question, ideally from someone with an Economics background who takes interest in central bank policy.

For example:

• why do so many analysts use market cap vs. M2 money supply, rather than market cap vs. M1 money supply?

• What exactly drives the difference between M2 and M1 money supply (I never studied macro properly, apologies if this is obvious)

Ps: it is not possible to upload a picture into a comment, so I chose to edit the OP's question instead.

• Hey Mike, I took the liberty to modify your question: pls let me know if that is not ok with you (I am happy to post it as a standalone question, but I think it's best to merge the two questions into one). – Jan Stuller Jan 21 at 9:10
• Thank you Jan. I think the two questions/remarks are indeed very related so they belong in the same place. However I'm sure that part of M1 is controlled by commercial banks( through loans that create more checkable deposits)) and not entirely by the Federal Reserve. I think the only money supply that is fully controlled by the Federal Reserve is the MB( monetary base). – Mike Cocos Jan 21 at 18:28

## 2 Answers

This is in response to the part of your question that asks about M1 versus M2, although it seems you've more or less answered parts of your own question. M1 is the simplest monetary aggregate and includes items most widely used as a medium of exchange (approximately 85% of household purchases are made using M1 balances); it is defined as follows:

\begin{align*} \mbox{Aggregate M1} &= \mbox{Currency held by the public} \\ &+ \mbox{Travelers cheques} \\ &+ \mbox{Demand deposits (checking accounts that pay no interest)} \\ &+ \mbox{Other checkable deposits (checking accounts that pay interest)} \end{align*}

M2 is a broader definition of money that adds to M1 other assets with check-writing features, such as money market deposit accounts, and other assets that can be turned into cash quickly with very little cost, such as savings deposits.

\begin{align*} \mbox{Aggregate M2} &= \mbox{M1} \\ &+ \mbox{Term deposits (deposits locked up for a period of time)} \\ &+ \mbox{Savings deposits} \\ &+ \mbox{Retail money funds (mutual funds investing in safe short-term assets)} \end{align*}

Looking at why central banks usually focus on M2 instead of M1 to monitor monetary policy gives us a quick sense for what drives M1 versus M2:

1. Interest rate changes. Higher rates entice people to switch balances in checking accounts, which pay little or no interest, into savings accounts, which pay more interest. This activity causes M1 to shrink but does not affect M2. Since people can relatively easily spend money from their savings account balances, it can be misleading to focus on trends in M1.

2. Financial innovations. The dividing line between checking and savings accounts has been steadily blurred with banks getting around the Fed prohibiting interest payments on checking accounts by, for example, creating savings accounts that earn interest but whose balances were automatically transferred into checking accounts when required. This actually led to the definition of M1 being expanded to include such accounts: "other checkable deposits."

3. Financial deregulation. Non-bank financial institutions such as mutual savings banks, credit unions, and savings-and-loans associations were at one time not allowed to have checking accounts, so their deposits were not included in M1. Current monetary aggregates include deposits at all financial institutions.

As you can infer from the above discussion, the main driver of growth in M1 versus M2 should be the interest rate offered on money substitutes, as long as the institutional structure of where firms and individuals hold their deposits doesn't undergo a significant change.

However, to therefore conclude that that the M2-M1 differential is primarily determined by commercial banks is probably too simplistic. Clearly, one effect of the Fed's massive bond purchases (QE) is to lower interest rates across the yield curve (and therefore more or less equalize the interest rate differential between checking and savings accounts).

Added Later

In fact, the following article shows how a prior episode of QE led to an increase in the growth rate of M1 versus M2: What's Driving up Money Growth?

• Great comment, thank you, really thorough and well explained. So basically it makes more sense to scale the market cap of (say) SPX with M2, rather than M1, because M2 also includes money markets, which are essentially the "checking accounts of professional institutions" , whilst M1 is mostly retail money: correct? – Jan Stuller Jan 22 at 7:09
• Good question but I don't know if I have a good answer. I have not yet come across a rigorous contemporary analysis of why M2 is superior to M1 in this context although I have seen a 1982 study by Sorensen suggesting that there is no difference. However, as per Mike's question and given the dramatic impact of QE on M1 in recent times, it appears as if M1 would be a better proxy. – Sharad Jan 22 at 17:57
• There is no switching cost from M2 to M1 and now all money is essentially M1 (by nature) and M2 (by definition). If and when interest rates move in the opposite direction, and no one is going to park any cash in checking accounts, all money will become M2 by nature and by definition. – Sergei Rodionov Jan 22 at 19:58

Broadly speaking, if something is in M2 and not M1, it's because there's some friction in spending that money, while M1 allows for mostly frictionless transactions.

M1 consists of currency in circulation, checkable/demand deposits, and travelers checks. All of these forms of money can be used to facilitate transactions immediately.

M2 further incorporates savings accounts, money market accounts/mutual funds, and low-value time deposits. These forms of money all require at least some amount of time or some sort of transaction cost and typically cannot be used for transactions directly and on demand. But, they can be converted into M1 relatively easily and then used for transactions. That's the primary difference.

• A follow-up question that comes to mind: QE is basically the FED buying US Government Bonds in open market transactions. What would then be the "path" of this printed money into the M2 "orbit"? I imagine the FED buys the bonds directly from an institution (be it a primary dealer of US govies, a corporate bank or a fund): this cash can then directly enter M2 via the money markets: correct? So (simplistically), each time the FED purchases some bonds via printed money, M2 grows by that amount, would you agree? (whilst most likely M1 stays constant, unless the money is "deposited") – Jan Stuller Jan 22 at 10:52
• This is interesting comment. If FED bought directly from bank and that bank then deposited the cash back to FED as excess reserves this would not be counted as M2. However, if indirectly the seller was a retail investor, or corporate, who deposited the proceeds back with a bank then that deposit would be regarded as M2, or?? – Attack68 Jan 22 at 15:41
• @JanStuller, money that enters money market accounts is not necessarily part of M2. M2 consists of low-value money market accounts (< \$100k) or those belonging specifically to individuals. Larger value accounts, like those of institutions, are part of M3. But that money will get to M2 eventually through salaries, dividends, etc., as the corporation transfers money to indviduals. – Amaan M Jan 22 at 17:35
• @Attack68, that sounds correct, but I'd also point out that since the bank's excess reserves have increased, they will lend out money until they're back at the reserve requirement. So, in that sense, whether the Fed buys from institutions or individuals, M2 will increase. – Amaan M Jan 22 at 17:39
• @JanStuller That makes sense, but they are highly correlated, so you'll likely have similar outcomes regardless of which one you use if you're looking at percent changes. – Amaan M Jan 22 at 18:38