What’s Money? What’s Not?

Analysts who subscribe to the theory that modern-day boom-bust cycles are caused by large swings in the money supply would do well to ensure that they are using the proper metric for measuring money to begin with. Given that the extreme growth of the money supply sets in motion the widespread misallocation of capital and thus sows the seeds of the inevitable bust, there are few economic statistics where accurate measurement becomes so crucial to the analysis of general liquidity conditions throughout the broad economy.

This is not merely an exercise in splitting hairs, for there have been many cases in the past where differing money supply metrics have diverged substantially, leading to diametrically opposed conclusions about the future direction of the global economy. It is therefore a critical step in any analysis of liquidity conditions to properly flesh out an accurate definition of money, which was briefly elaborated on in a previous article discussing alternative methods to more properly measure the money supply (see Follow the Money). Briefly, the article explained that commonly reported metrics reported by official government sources (in the case of the referenced article, using M1 and M2 as examples) are insufficient measurements of money since they contain components that do not in practical terms behave like money under real-world conditions. Analysts who follow these metrics run the risk on being substantially off-base in their conclusions during important turning points in the economy.

The article referenced above focused on money supply metrics in the context of the United States, specifically outlining how the widely followed M2 metric has historically given off false and misleading signals when used as a predictive measure of modern-day boom-bust cycles. The problem, however, is not limited to the US, as we have recently seen certain Canadian analysts similarly use official Bank of Canada metrics which are not reflective of the actual money supply to make their case that the country is undergoing tightening liquidity conditions that will trigger an impending economic bust. Most recently, the Canadian real estate site Better Dwelling posted an article analyzing the official government M2++ money supply figure to make the case that liquidity is contracting throughout Canada. Now it so happens that liquidity is in fact contracting and the thrust of the article is therefore correct, but this is certainly not because of what is being pointed to by the official M2++ figure.

The problem with using M2++ is that it contains many items that aren’t actually money, but rather assets or simple loans. By way of example, the M2++ figure contains items which are neither cash nor perfectly secure, immediately convertible, par-value claims on standard money. In essence, they contain items which are not money-substitutes and therefore should not be counted as part of the money supply. These include items such as money market funds, fixed time-deposits (such as GIC’s), debt instruments (such as Canada Savings Bonds), but also, and most egregiously, non-money market mutual funds. In the case of non-money market mutual funds which invest in a wide assortment of financial assets such as stocks and bonds, it is obvious that due to the constantly changing market value, often-extreme volatility, and the fact that there are typically multi-day redemption periods, these instruments fail the classic definition of money-substitutes and should be excluded from measures of the money supply.

In the case of money market funds it is perhaps less obvious that these instruments should be excluded from the money supply. But we need only look at a simple example to see why money market funds are not actually money. By way of example, if Lawrence adds $10,000 to his money market fund, and the fund uses Lawrence’s money to buy short term interest bearing T-bills from Aziz, then money is ultimately transferred from Lawrence’s bank account to Aziz’s bank account. Money has simply moved from one account to the other, but since Lawrence’s money market account has increased by $10,000, M2++ will erroneously show an increase in the money supply of $10,000 even though no new money has actually been created.

Similarly for a time deposit such as a GIC, if Aziz lends $10,000 to the bank to purchase a GIC (to be repaid in a year’s time with interest), money has moved from Aziz’s account to the bank’s, but no new money has actually been created in the transaction. But once again, by including the GIC as part of the money supply, M2++ gives the impression that the money supply has again expanded by $10,000, but this is not the case at all. Aziz can no sooner make a purchase with a GIC than he can with a money market fund or a stock certificate that he owns through his broker. M2++ may well be measuring securities with differing degrees of liquidity, but securities do not serve as money in the real world, no matter what the Bank of Canada is reporting. Anyone who believes otherwise should try buying groceries at their local supermarket with a GIC or government bond.

But the criticism of the money supply metrics used at Better Dwelling do not stop at M2++, for in other articles, a different money metric called M1+ was used to measure the money supply to essentially make the same argument that liquidity was contracting throughout the Canadian economy. As far as accuracy goes, M1+ is leaps and bounds ahead of M2++, but still not entirely reflective of the true state of money since it excludes “notice” deposits, which are effectively savings accounts. While it might appear that it is correct to exclude savings deposits based on the strict definition of money due to the theoretical waiting period that banks may impose on depositors prior to withdrawals, in practice this waiting period is never enforced in either Canada or the United States. This is different than in Europe, where notice periods on savings accounts are enforced, which is why money supply metrics in North America should include savings accounts, but money supply metrics in Europe should not.

What then is the most accurate and truest money measurement that exists in Canada today? It would seem to be M1+ plus savings deposits. The closest metric that comes close to matching this in Canada is the statistic M1++, which is surprisingly one that the analysts at Better Dwelling never actually use. Although M1+ and M1++ tend for the most part to move in the same direction, the magnitude of the changes are often vastly different. Note that while it may be the case that all of M1+, M1++, and M2++ are today pointing to the same conclusion regarding overall liquidity, this is not guaranteed to always be the case, which is why it is important for analysts to be clear on what constitutes actual money and what does not.