How Things WorkMost Popular PostsMyth BustingRecommended ReadingStuff You Should Read

EVERYONE Funds Their Spending

One of the most confused (and confusing) elements of endogenous money is the idea of “funding”. Endogenous money is not a new theory, but it is not well understood even to this day. Even many supposed endogenous money theorists, like the MMT people, misunderstand it and as MMT has gained some popularity I am seeing increasing misinterpretations. It took me a while to get this so let’s dive in and see if I can’t explain this more succinctly and clearly.

Endogenous money is the fact that anyone can expand their balance sheet from nothing so long as they can find a willing counterparty to agree to that balance sheet expansion. For instance, I can write an IOU on a piece of toilet paper for $100 and as long as someone is willing to accept that toilet paper I have essentially created a new contract that agrees to deliver $100 to someone else at some point in time. We have created new financial assets from nothing. In other words, the contract is endogenous. It doesn’t come from existing money or the government or something like that. We created it internally. Okay, that’s very simple.

But we have to acknowledge another essential fact of ALL viable financial contracts – they involve counterparties. You cannot create viable financial assets without a counterparty or someone else to take the other side of your agreement. Endogenous money involves two parties and requires both a buyer and seller. Otherwise, there is no market. But having a market does not fully reflect the scope of funding. After all, you can have a market for your assets and still have a high cost of funding. For instance, there is a market for junk bonds, but they are junk specifically because the market for those bonds is thin. These firms have “funding”, but the cost of that funding is expensive because the risks associated with it are higher.

Now, much of the confusion on this concept revolves around the cost of funding for banks and governments. I have described on many occasions how banks do not necessarily expand their balance sheets by sourcing existing reserves. That is, banks do not lend money from a fixed pool of loanable funds like we are taught in textbooks. Banks don’t take in deposits or revenue and loan out that fixed pool of funds, per se. Instead, loans create deposits, but deposits also fund loans. What this means is that banks operate much like the toilet paper maker above. But their financial asset creation is contingent on being able to leverage a certain capital position. Banks achieve this by having liabilities that are less expensive than their assets. For instance, if a bank has deposits that cost 1% and loans that pay 5% then they have a net interest margin of 4%. If their other operational costs are 2% then they should earn a net profit of 2%. In other words, the combination of their cheap liabilities and higher paying assets make it affordable to accrue a positive capital position which allows them to leverage their balance sheet into more loans. So their loans create deposits, but their cheap deposits also fund their loans. The kicker is, when their capital position deteriorates the cost of their funding will rise which will make it increasingly difficult to fund their balance sheet.

Governments do not operate that differently, but have a lot more flexibility in their nominal price setting because they tend to be the most credible entities in the economy due to their vast income via taxing authority. Like banks, governments do not take in a dollar and rely on being able to lend out that dollar. Instead, a government can be thought of as leveraging its domestic output in much the same way that a bank leverages its capital or I leverage my net worth. For example, I have a printing press in Cullen Roche debt. I can go to any bank and obtain a line of credit that will allow me to expand my balance sheet. But I require a counterparty who will price the terms of that money printing to account for credit risk, interest rate risk, inflation risk, etc. If demand for my money declines I will have to borrow at both higher nominal and real rates. Importantly, my income and assets play an important role in the demand for my money creation. If my income plummets my credit will become more expensive because the demand for my money will decline.

Governments are just very credible entities with massive income streams who can reliably settle their payments at par because they can’t be determined to be bankrupt by anyone but themselves. But governments very much rely on counterparty funding and they cannot force people to buy their financial assets at a viable real price. The market influences the cost of a government’s funding since anyone can re-price government money against other assets. For instance, when a government’s income/output collapses they often undergo a decline in demand for their money which shows up as a rise in the rate of inflation. The real cost of funding increases.¹ Governments issue financial assets at par value so their prices change in real terms. That is, the market doesn’t re-price government assets in nominal terms because the government cannot go bankrupt in nominal terms, however, the market does re-price government assets in real terms.² A government that prints money and throws it on the street still relies on demand for that money at a real price. This price is the effective cost of funding. So having a printing press might keep you from having a funding crisis in nominal terms, but it does not mean a government can’t experience a funding crisis in real terms.

So, for instance, when MMT people say things like “taxpayers don’t fund spending” or cite old (incorrect) Federal Reserve research that says “taxes for funding are obsolete” they are making a rather basic misunderstanding of endogenous money. In fact, they are clearly contradicting themselves when they claim that inflation constrains spending, but output doesn’t fund spending. Since income is indicative of any entity’s funding capabilities and the demand for domestic output/money it should be obvious that taxes fund spending and make a government’s balance sheet leverage increasingly viable even though the government does not need to bring in a dollar of taxes for every dollar spent.³ In other words, we can say that taxes/income reflect how much policy space a government operates with since taxes reflect the ability to redistribute existing productive income which gives the government the ability to increasingly leverage its balance sheet with new endogenous money/debt creation.

The bottom line is, in an endogenous money system, everyone funds their spending, but the cost of that funding and the causes of it will vary depending on the specific entity and the specific environment they find themselves in.

¹ – Inflation and insolvency can both be viewed as a relative decline in the demand for money, but the causes of a very high inflation are quite different from the causes of an insolvency. This is crucially important when understanding this discussion since high levels of government spending and debt do not necessarily lead to high inflation in the same way that high levels of household debt might increase the risk of household bankruptcy. 

² – It’s important to understand that being able to sell bonds (or cash) to your own Central Bank does not mean you have viable funding. For example, Argentina does not currently have liquid and viable funding sources. This isn’t because the demand for their domestic debt can’t be purchased from the Central Bank. It is because the non-government sector has priced cash at a level that is extremely expensive for the government to print and so their real cost of funding has created a nominal funding constraint (since issuing more cash/debt will likely cause even higher inflation).  

³ – As I’ve stated before, this is an obvious and rather basic operational error in MMT. But while it sheds some doubt on their operational understandings of endogenous money and the financial system it does not necessarily “debunk” MMT because MMT has never really been tried anywhere.