Friday , December 13 2024
Home / Mike Norman Economics / Steve Roth — Actually, Only Banks Print Money

Steve Roth — Actually, Only Banks Print Money

Summary:
Actually, not really. Only governments have "printing presses" and other attempts to "print money" are called counterfeiting. Steve Roth knows this, of course, so he must being using "printing money" in another sense. Banks don't have printing presses in the basement, literally or figuratively. Governments that issue their currencies as government liabilities do. Banks increase M1 money supply by adding assets as receivables to their balance sheets as receivables  and booking corresponding liabilities as payables. The reverse shows up on the borrower's balance sheet. Both sets of books are in balance, with net zero on the respective balance sheet. Credit and debit entires match. Now the bank has an asset on its balance sheet and the customer has a credit in a deposit account. So

Topics:
Mike Norman considers the following as important: , , , , ,

This could be interesting, too:

Mike Norman writes Jared Bernstein, total idiot. You have to see this to believe it.

Steve Roth writes MMT and the Wealth of Nations, Revisited

Matias Vernengo writes On central bank independence, and Brazilian monetary policy

Michael Hudson writes International Trade and MMT with Keen, Hudson


Actually, not really. Only governments have "printing presses" and other attempts to "print money" are called counterfeiting. Steve Roth knows this, of course, so he must being using "printing money" in another sense. Banks don't have printing presses in the basement, literally or figuratively. Governments that issue their currencies as government liabilities do.

Banks increase M1 money supply by adding assets as receivables to their balance sheets as receivables  and booking corresponding liabilities as payables. The reverse shows up on the borrower's balance sheet. Both sets of books are in balance, with net zero on the respective balance sheet. Credit and debit entires match. Now the bank has an asset on its balance sheet and the customer has a credit in a deposit account. So even though the net is zero in the financial system, purchasing power increases in nongovernment owing to the increase in M1 money supply as a result of the deposit in the customer's deposit account.

Why is this not a form of "printing money" then, as Steve Roth claims?

When a central bank issues the government's currency it creates liabilities on its balance sheet and assets on the balance sheets of banks as deposits at the central bank. Assets increase in nongovernment as a whole without a corresponding liability in nongovernment. There is a change in aggregate nongovernment financial assets.

When banks create a deposit by extending a loan, there is no change in aggregate net financial assets in nongovernment. The net is zero. When central banks deficit spend, there is. The net is the amount of the deficit.

A currency is the unit of account that which the government sets and which it accepts as payment of obligations it imposes on nongovernment, chiefly taxes but also tariffs, fees and fines. When banks settle accounts with the government as proxies for customers, they have to settle in the government's liabilities, either cash or bank reserves in the central bank payments system. This is what "printing money" implies. 

All currency users including banks have to obtain the currency as government liabilities, and the only source is government issuance of the currency through spending or lending, since only government can create government liabilities. In modern monetary production economies, governments delegate this power to their central bank as the government's financial agent. 

The currency sovereign's ability to create its own liabilities is unlimited. The constraint is the availability of real resources for sale in the currency, since over-issuance risks inflation. Correspondingly, under-issuance risks contraction and ultimately deflation.

For example, when a taxpayer pays taxes to the government that issues the currency, one of two things happen. Either the taxpayer pays in cash at a government office, or tenders payment through a bank, e.g., a paper check or electronic check. If cash, that cash is deducted from currency in circulation, reducing M1. 

If by check, the check has to clear in the government's payment system. In this case, the government charges the bank's account in the payments system and the bank corresponding charges the customer's account at the bank. If there are insufficient funds in the customer's account, then the bank bounces the check and the funds are not deducted from the banks' account at the central bank. This illustrates the huge difference in kind between the "money" banks generate through credit issuance in M1, and the "money" that the government issues through its financial agent, the central bank.

What happens if the bank doesn't have enough funds in its account at the central banks to cover its obligations? Either the bank borrows in the interbank market at the policy rate set the central banks sets, or the central bank just lends the funds to the bank so the payment system to clear. But the bank is charged a penalty rate set above the policy rate to discourage this, and a bank that abuses the "discount window" risks its standing and ultimately its position in the payments system.

Again, a pseudo-problem arises from using "money" without a proper technical definition, or else not paying attention to the existing definitions. Similar problems arise with other key financial and economic terms, such "the interest rate," "capital," and "saving."

SteveRoth is correct that loans creating deposits adds to the M1 money supply, but this is not currency issuance in the customary meaning of the phrase. Government's issuance of liabilities in the government's unit of account is not the same as banks creation of its own liabilities in the government's unit of account, which the bank cannot issue, being currency users and not issuers. 

If banks could "print money" in the sense of issue currency as the government creates its own liabilities by marking up accounts on its spreadsheet, banks' risking would not risk insolvency owing to default on loans they extend. This is not the case. Conversely, government bonds are default-risk free since the government can always issue currency to cover its obligations to security holders.

However, because bank credit increases M1, which adds to purchasing power in nongovernment, bank lending can affect the price level and spur inflation. Similarly, contraction in bank lending that is not offset is deflationary.

Steve Roth also thinks that issuance of government securities to offset deficits, which is mandatory in the US, sterilizes those funds so that they don’t add to money in the sense of M1. This is not the case. The funds that government injects increase M1 since currency is issued by crediting deposit account, which adds to M1. 

Taxes subtract from both M1and also aggregate nongovernment financial assets. Payment for purchase of newly issued government securities decreases M1, but those government liabilities are simply switched from deposit accounts in the payments system to time deposits at the central bank. The amount of nongovernment net financial assets remain the same in aggregate.

Moreover, government securities are the most liquid form of financial asset after cash, short term bills are essentially cash equivalents, and government securities are the best form of collateral, spending is not affected by draining the monetary base as deposit accounts at the central bank to time deposits at the central bank. 

Issuance of government securities does not sterilize deficit spending. There are other reasons to issue government securities, but this is not one of them. Chiefly, government securities issuance drains the monetary base, facilitating the central bank hitting its target when not paying interest on excess reserves and not choosing to set the policy rate to zero. Government securities also provide interest-bearing default-free risk instruments for the financial community and nongovernment savers. Issuance of government securities is not necessary operationally. It is a policy choice, hence political.

Ok, this looks like semantics and logic-chopping. But it is not. It’s just as important to create conceptual models that are correct as it is formal models. This means weeding out the weasel words, clarifying concepts and getting description right. Some formalism in needed in this regard, but it is not econometric but double-entry accounting. This requires thinking in terms of T-accounts, and that needs to be check by writing it out.



Asymptosis
Actually, Only Banks Print Money
Steve Roth
Mike Norman
Mike Norman is an economist and veteran trader whose career has spanned over 30 years on Wall Street. He is a former member and trader on the CME, NYMEX, COMEX and NYFE and he managed money for one of the largest hedge funds and ran a prop trading desk for Credit Suisse.

Leave a Reply

Your email address will not be published. Required fields are marked *