The Fed has Provided the Bulk of Money Supply Growth since 2008
We have discussed the topic of money supply growth extensively in these pages over time. Below is a brief recap of how the system works in the US. Note that although fractional reserve banking and central bank-directed and backstopped banking cartels are in place all over the world, there are several “technical” differences between them. So the workings of the US system cannot be transposed 1:1 to e.g. Japan’s system or the euro system.
There are two possibilities of growing the fiat money supply: In “normal” times, commercial banks will extend loans which are partially “backed” by fractional reserves. These loans create new deposit money, which once again can serve as the basis of further credit creation, which again creates new deposit money, and so forth. It can be shown mathematically that based on a hypothetical fractional reserve requirement of 10%, extant deposit money in the system can be grown 10-fold (for a detailed discussion of the “money multiplier”, see here).
In actual practice, reserves have not represented a constraint for credit and money supply growth by commercial banks for quite some time. In the US banks can e.g. “sweep” money from demand deposits into so-called MMDAs (money market deposit accounts) overnight, letting these funds “masquerade” as savings deposits, which allows them to circumvent reserve requirements. Moreover, if credit demand is so strong that interbank lending rates (i.e., the Federal Funds rate) threaten to rise above the target rate set by the Federal Reserve, the central bank will supply additional reserves to the extent necessary to keep the rate on target. Thus the required fractional reserves will be supplied even if commercial banks don’t have sufficient excess reserves to lend to banks short of reserves.
None of this has been of importance since the 2008 crisis however, as “QE” has created such an overhang of excess reserves that interbank lending rates have continually wallowed close to the lower end of the 0.00%-0.25% Federal Funds target corridor. Moreover, up until late 2013/early 2014, commercial bank credit growth had slowed to a crawl anyway. So barely any money supply growth has come from the banking sector after the crisis. Enter the Fed, and “QE”.
In theory, if the central bank buys securities directly from banks, it would only issue bank reserves in payment (the selling bank receives a check drawn on the Fed, and upon depositing it, its reserves account at the Fed is credited). In actual practice however, QE in the US system concurrently also creates new deposit money at close to a 1:1 ratio. Most of the broker-dealers the Fed uses as counterparties in its open market operations belong to banks, but they are legally distinct entities (i.e., they are legally non-banks). Hence, when their accounts are credited, not only bank reserves are created, but new deposit money as well.
Our friend Ronald Stoeferle, one of the managers of the Incrementum fund in Liechtenstein, has mailed us an interesting chart that compares the growth rates of the official US monetary aggregates and total debt in system since 2008. It shows how the Fed really had to put the pedal to the metal to create money supply growth. System-wide debt growth meanwhile remained subdued (the government has grown its debt enormously and corporations have also expanded their debt load, households however have deleveraged):
Growth rates of US monetary base, M2, M3 and total credit market debt owed since 2008 – click to enlarge.
Note that the Fed no longer calculates M3, but several people have reconstructed it using alternative data sources (e.g. here is an article explaining how our friend Bart at Nowandfutures is calculating M3 these days. John Williams of shadowstats has also reconstructed the series).
Even though “QE” translates directly into deposit money (which is counted as part of the money supply; by contrast, bank reserves are not part of the money supply, as they remain outside of the economy), the smaller base from which the monetary base started out in 2008 meant that base money had to be expanded to a far greater extent in percentage terms to achieve the growth in the broad monetary aggregates depicted above.
It should be noted that all deposit money created as part of “QE” operations represents so-called “covered money substitutes”, as the bank reserves covering it have been created concurrently. By contrast, if new deposit money comes into being in a commercial bank credit operation based on fractional reserves, the bulk of the money substitutes (i.e., the deposit money) created in the process consists of uncovered money substitutes. If more than a certain percentage of depositors were to attempt to withdraw their demand deposits at the same time, they would find out that the money is actually not there. Due to QE, nowadays a far larger percentage of the deposit money in the system is actually of the covered variety than was previously the case (approx. 29% vs. about 5% in the pre-crisis era).
What Should be Counted as Money?
As readers know, we prefer the “Austrian” measure of the money supply, money TMS (which stands for “true money supply”) over the official broad money supply aggregates. As Murray Rothbard noted in his essay “Austrian Definitions of the Supply of Money”:
[…] money is the general medium of exchange, the thing that all other goods and services are traded for, the final payment for such goods and services on the market.”
This seems straightforward enough and surely everyone would agree with this definition. In a fiat money system, we can differentiate between “standard money” – i.e., banknotes – and deposit money. Both are equally serviceable for effecting final payment for goods and services and hence form part of the money supply in the broad sense.
Why was it thought necessary to create the Austrian money measure TMS and what makes it different from the official monetary aggregates? It all comes down to the definition of money cited by Rothbard above. The official measures such as M2 contain components that are actually not money according to this definition, while excluding some that are.
The most important of the non-money components are money market funds. Since money market funds buy short term debt securities and issue share units to investors, they are merely a credit intermediary: The money they use in order to buy e.g. commercial paper shows up in the form of deposit money on the accounts of borrowers. Investors holding mutual fund shares aren’t holding money; they cannot use their mutual fund shares for payment. These shares must first be sold, and only thereafter the money received for them can be used in payment. Counting these money market funds as part of the money supply therefore results in double-counting.
For more details on which components of the money supply aggregates are not part of money TMS and which components that are not part of the “Ms” are included in it, readers should check out Michael Pollaro’s excellent and extensive article on the topic. The article also contains a list of references to essays by various other “Austrian” economists on the topic.
In terms of money TMS – this is to say actual money – the Fed has been a bit more effective in blowing up the money supply than is indicated by the growth of M2 and M3. At the beginning of 2008, the broad money supply measure TMS-2 stood at $5.3 trillion; as at the end of December 2014, it stood at approx. $10.703 trillion, in short, it has more than doubled.
We have recently show the chart of TMS-2 in a different context, but here it is again:
US money TMS-2 (broad true money supply) – click to enlarge.
The difference between the growth rate of TMS-2 and M2 is largely due to the latter’s money market funds component – M2 started from a much higher base in 2008, and due to the stock and bond market rally since 2009, money market fund investments have been drawn down in favor of investment in “risk assets”.
Note that while changes in money market fund holdings may occur on account of people replacing them with investment in stocks and bonds, this decision has no influence whatsoever on the amount of money in the system, as every purchase of securities is matched by a sale. All that happens is that the ownership of securities and the money used to pay for them changes.
This is also why the “money on the sidelines” argument often cited by stock market bulls really makes no sense. Whenever a trade takes place, there is as much “money on the sidelines” after it as there was before it. Only a change in ownership occurs. The only sensible thing that can be said in this context is that the overall supply of money has more than doubled since 2008 courtesy of the Fed’s electronic “printing press” – in that sense, there is indeed more “money on the sidelines”.
Below is a chart showing the annualized growth rate in commercial and industrial loans in the US. The annualized rate of growth has recently accelerated to about 13.8%, which means that commercial banks have so to speak taken the baton from the Fed in terms of creating money supply growth:
Annualized growth rate of US commercial and industrial loans – click to enlarge.
These are actually typical boom time credit growth figures. They are counterbalanced a bit by a much slower growth rate in consumer credit. This is the main reason why the contribution of bank lending growth to money supply growth hasn’t been strong enough to achieve much more than keeping money supply growth roughly steady since the end of “QE”.
It remains to be seen whether the recent collapse in the oil price will affect these credit growth rates. A lot of credit has been pumped into the oil patch in recent years, and this activity seems now likely to grind to a halt. It seems therefore possible that the slowdown in the broad money supply growth rate in evidence since its 2010 and 2011 peaks will soon resume. Currently (i.e., as of year-end 2014), the year-on-year growth rate stands at 7.97%, which is down from the 16.7% and 15.67% peak growth rates in 2010 and 2011 respectively, but roughly still in the same range that has prevailed since late 2013 when “QE” was discontinued.
TMS-2, year-on-year growth rate – click to enlarge.
As this chart also indicates, asset price bubbles tend to peak with a lag to peaks in money supply growth rates, usually after a certain (unknowable) threshold in the annual growth rate is undercut. The threshold just prior to the 2008 crisis was very low (less than 2%), but it was e.g. at about 5% in 2000 before the Nasdaq bubble broke. What level of money supply growth will be decisive this time around is something we will once again only be able to ascertain in hindsight, but the fact remains that such a threshold exists.
The Fed has been responsible for the bulk of money supply growth since 2008, but this has recently changed. For the moment, the commercial banks are “back in the game” and have replaced the effect “QE” had on money supply growth by ramping up their inflationary lending. Traditional bank credit growth has ergo once again become an important measure to watch. The sideways move in broad money supply growth that could be observed over the past year could still continue for a while, but we suspect that there will eventually be a further slowdown. If so, it will be bad news for the asset price bubble.