These commentaries contain theories and models that I come up with myself and should therefore be considered “work in progress”. Since I don’t intend to pre-test my ideas “till death do us part” you should expect them to be revised and modified as my own understanding grows. Last week I wrote about a new way to calculate the money supply and during this week I persuaded myself that I had made an error.

I have come to believe that money within the banking system should not be counted as part of the money supply. In last week’s letter the Actual Money Supply did include some money within the banking system (bank borrowings and net amounts due to related foreign offices of banks) and therefore I am herewith retracting last week’s letter. I recalculated the Actual Money Supply (AMS) and the Actual Inflation Rate (AIR) and changed last week’s letter on my website. The new version will be sent to you in a separate email right after this one. Please read the updated version and discard the incorrect, prior letter.

The reconstituted Actual Money Supply also changes the gold price model I sent you two weeks ago, so I am retracting that letter as well. The gold model has undergone another major change this week that I hope will have increased its accuracy; however, the results point to a lower theoretical gold price which I understand is in conflict with the current rise in the gold price and perhaps counter intuitive to many gold investors. That being said, I use this model for my own investing and trading and, in light of the new results, have placed tight stop loss orders in my gold futures trading account. I will update the gold model and resend the June 27th letter to you as soon as possible. In the meantime, please do not rely on the theoretical gold price discussed in that letter.

Before I explain why I decided to exclude money within the banking system from AMS I would like to say that creating a better model of the Actual Money Supply should shed significant light on the inflation versus deflation debate and allow us to carefully monitor changes in the money supply on a weekly basis and I hope to incorporate weekly inflation (AIR) updates from now on in every commentary. There are numerous applications of this data that I will explore during the rest of the year and that I hope will be of value to you.

Now back to money within the banking system.

All the money in the financial system is either deposited in bank accounts or held as notes and coins. Bank account deposits, whether they are demand deposits such as checking accounts, or term deposits such as savings accounts, constitute the bulk of the money supply. If we add to these deposits the currency in circulation (notes and coins held by the non-bank public) we have the Actual Money Supply (AMS). Deposit accounts are the liabilities of banks and the money supply therefore increases when banks’ liabilities increase. Today’s question is whether we should also include other liabilities of banks, or not. To answer that question let’s first look at what happens when non-bank debt is created.

If a corporation borrows money from the public by issuing bonds then money moves from the deposit accounts belonging to the public into a bank deposit account belonging to the corporation. The total amount of money therefore did not change since the total size of all deposit accounts has remained the same. The total amount of bank liabilities does not change when a non-bank entity issues debt and therefore, neither does the money supply.

From an accounting point of view the public has less money but gained an asset -- the corporation’s debt. The corporation has more money (an asset) but it also has an offsetting liability – the bonds it issued. Non-bank debt, including government debt, does not change the money supply; it merely rearranges who owns the money.

The situation is very different when a bank borrows money from the public. The bank’s own cash is held in internal deposit accounts. These internal deposit accounts are not reported together with customer deposit accounts since one is an asset of the bank and the other is a liability. Customer deposit accounts are liabilities of the banks whereas its own cash in an internal deposit account is an asset. When the bank borrows money from the public, the public’s deposit account balance declines by the amount of the loan -- instead of money, the public now own the bank’s bonds, and bonds are not money, they are financial assets.

Last week I took the position that money banks borrowed must be included in the money supply. This money is no longer included in deposit account balances and therefore not included in the money supply, so I included banks’ borrowings in the money supply to account for it.

Most bank borrowing is not from the public, but from the Federal Reserve Bank, the Federal Home Loan Bank or the US Treasury. The accounting of bank borrowings from these institutions as opposed to the public is somewhat different but the results are the same in that these borrowings represent money that is held in the banks’ internal deposit accounts and would only be counted in the money supply if we included banks’ borrowings.

However, this week I convinced myself that we should not include such borrowings.

The bulk of the money that banks own (regardless of the source) is part of their reserves and cannot be spent -- therefore the money is “sterilized”. When banks borrow money they typically do so in order to adjust their reserve requirements and therefore bank borrowings do not directly affect the amount of money available to the public. Any money that does become available, either through the creation of new bank deposits, or by direct expenditures of the bank, will immediately show up as an increase in deposit account balances or currency and then be included in the money supply.

Since the whole purpose of constructing money supply data is to figure out how much money is available to be exchanged for goods and services, it would make sense to disregard the sterilized money in the banking system. Such money cannot affect prices and therefore becomes immaterial to the money supply. The small amount of discretionary money that banks may have should not materially affect the money supply while including the sterilized money in the banking system would materially overstate the money supply and could materially affect the Actual Inflation Rate.

The data used to construct the Actual Money Supply therefore now includes only deposit account balances (both demand and term deposits) plus currency held by the non-bank public. This simple aggregate is, however, not computed by any regulatory agency that I am aware of, nor any private organization – yet it more accurately describes the money supply available to be converted into goods and services, or to be used to discharge debts, and will therefore tell us the Actual Inflation Rate of US dollars.

As I mentioned, I will re-write last week’s article that deals with the Actual Money Supply in more detail and send it to you as soon as possible and then also update the gold price model.