I have just finished the very intellectually stimulating course by Perry Mehrling on Money and Banking. A key point of this course is to move away from the standard economics and finance view about the ‘veil of money’ and money neutrality, which holds that money is fundamentally a neutral unit of account, and has no real effects on the economy, and to move towards the ‘money view’ which argues that money has its own crucial dynamics and processes which can have real effects on the economy. These effects are primarily mediated through the ‘money market’, which consists of several interlinked markets for short-term lending and borrowing. The key way that money impacts the real economy is through liquidity constraints. Every single day your cash inflows (including borrowing) must be sufficient to pay your cash outflows, else you go bankrupt, no matter the intrinsic solvency of your enterprise. However, on average there is no guarantee that your cash inflows and outflows are precisely lined up in time to allow you to meet your liquidity constraint every single day. This means that the continued functioning of the economy is largely dependent upon the money market – or short term borrowing – to provide liquidity to allow firms to push off such liquidity concerns into the future where (hopefully) there will be a cash inflow to pay off the present cash outflow. When this money market stops functioning, as in a financial panic, many normally solvent firms may be forced into bankruptcy by liquidity issues, thus having a deleterious effect on the real economy.

Perry also proposes that in any economy at any time there is a strict hierarchy between different types of money. These different types of money are things like cash (i.e. banknotes), bank deposits, reserves at the Fed, or loans due to you. Credit sits at the bottom of this hierarchy, which then moves up to deposits in a bank, to bankers’ interbank deposits, to the ‘best money’ – US treasury bonds or federal reserve notes or, under the gold standard, gold. Importantly, although some of these types are strictly credit (such as mortgage loans) and others are appear to be money such as bank deposits, it is important to note that it is all some form of credit except for the ‘best money’ – banknotes from the fed. Your bank account may seem like money but it is in fact a loan to the bank, which the bank promises to repay on demand. In good times all these types of money are relatively easy to convert into one another. You can withdraw money from your bank account. If the bank makes a loan to somebody they can sell it to somebody else for cash or deposits. Banks will accept deposits at other banks. Some of these forms of credit can be converted into the ‘best money’ (cash) at par, such as bank deposits. Others trade at a discount such as loans, as they are recognised as risky and the buyer demands a discount, or risk premium, to take on the risk of default. Importantly, although some forms of low level ‘money’ such as bank deposits typically trade at par with real money (cash), and thus are treated as one and the same, the bank deposit is still credit. When a crisis occurs (such as a run on the bank), the differences between real money and credit becomes very apparent.

With this perspective, we see that the money supply is thus not just a single number but a hierarchical pyramid of levels of credit, where each layer of credit can be payable in terms of money higher up within the hierarchy. In good times, even with a fixed supply of ‘best money’ at the top, the lower levels of credit can expand to allow for growth in economic coutput, while in bad times the system collapses and there is a flight to ‘better’ money as first you realize that the loans you bought may not be safe, and secondly, if the crisis is bad enough, that your bank deposits may not be safe either.

Importantly although Perry’s model contains multiple layers of credit, it also contains a fixed ‘best money’ at the top of the pyramid. In a crisis, the ‘best money’ is always safe. However what exactly this ‘best money’ is, and why it occupies the position it does, is unclear. In Perry’s model the best money is either liabilities of the Fed (cash,notes, Tbills) or else gold (under the gold standard). But why is this the best money? And, moreover, why is there a hierarchy at all – i.e. why is it better to be paid in the ‘best money’?

Ultimately, I will argue that the best money is not arbitrary or some distinct category, ontologically distinct from credit. Rather, as you ascend the hierarchy, you instead reach increasingly generalised levels of credit. That is, credit that more and more people will accept. Ultimately the ‘best’ money is the credit that anybody and everybody will accept. All money is credit or debt. Even the money printed by the Fed 1, and even gold. The key concept is that of generalised debt – a debt which anybody will accept in return for real goods or services, and anybody will accept it because it is a Schelling Point. The best money is simply the strongest Schelling Point in the system. It is not absolute and this Schelling Point can even collapse under sufficient economic stress. From this perspective, we can then understand theories of the value of money – chartalism and metallism – which have often been considered to be incompatible, and see that both are incomplete and can be subsumed within the Schelling Point theory. Each simply describes the advantages and dynamics of one possible Schelling Point for the best money – state fiat money, or gold.

To begin, consider the purpose of money. The standard story is that money is a store of value, a unit of account, a medium of account, and a medium of deferred payment. Typically this latter function is dropped. However, I argue that being able to make generalised deferred payments is actually the core function of money. Think about it. You make a transaction. You pay somebody to get real goods. You get goods, they get money. But what is the value of money? It has no intrinsic value, only what goods another person are themselves willing to trade for it. But this trade will usually happen in the future, so it is a deferred payment.

Consider the lower forms of money further down Perry’s hierarchy. These are uniformly forms of credit. They are promises to pay “later” in return for something now. Now importantly cash money is this as well. When you pay for something with money, they give you something now. But what do they get in return? Money. But money has no intrinsic value, so why do they do it? They do it because they expect to be able to use that money to buy real goods and services for themselves later. Note, that money is used to get goods now for goods later. It is a credit. But it is a special kind of credit. The people who pay money for goods and now and later are different people. This is the crucial differentiator from the normal conception of debt which is bilateral – i.e. you pay someone now in return for more money from the same person later.

Money is a generalised credit. It allows you to obtain goods from any person in the economy, in return for the credit. This is crucial. Because why would any person accept this ‘credit’? They accept it because they believe that they can take the money and exchange it to any other person in the economy and so on, ad infinitum. The value of money, therefore, arises from the expectation that it can be exchanged at value for goods or services by anyone in the economy. This property of being a property everyone else accepts is a Schelling Point. Thus the value of money arises from its status as a Schelling Point for making deferred payments.

Using this perspective we can understand both metallism and chartalism. Both of these theories are not theories of money, they are theories of one potential Schelling Point for money. Chartalism explicates the dynamics behind the Schelling Point of fiat-money backed by the power of the state.

The power of the state can provide a powerful impetus for the formation of a Schelling Point. If people are legally obliged to accept a currency (legal tender laws), then if the currency is reasonably intact and usable then it will probably take precedence, since the alternative would be risking the wrath of the government for no reason. This power is also enhanced by the state’s ability to levy taxes in its own currency. Importantly, the strength of the Schelling Point decreases as the power of the state wanes. It is not absolute. If the state’s money is widely seen as worthless, or else likely to lose value, this will be reflected in rapidly rising prices, which in effect represent the risk premium of holding the state’s credit.

This is an interesting way of looking at hyperinflation. Here we have that the state’s money is so risky that nobody would willingly hold it, yet they are forced to by law. The result is that pressures escape through rising prices (i.e. very low conversions of currency into real goods). They can force you to accept the currency but not the price at whicih you accept it. If the price is regulated as well, the market mechanism breaks down and you get shortages and a black market which effectively replaces the government official currency with some other currency.

Similarly, metallism explains the mechanisms behind Schelling Point of gold as currency. These are it’s intrinsic scarcity, durability, moldability, aesthetic components etc, and difficulty to counterfeit. These have some force as a general Schelling Point. The Schelling Point of gold, or precious metals, is enhanced by the general and extremely strong cultural instinct to view precious metals as ‘real money’. Of course it is important to note that gold has very little value as money in and of itself. It shares the exact same problem as with fiat currency. Sitting on a giant pile of gold does you little good if nobody is willing to trade things for it.

So ultiamtely what does this theory mean? First, it provides a unification of Mehrling’s pyramid. All ‘money’, even the ‘best money’, is credit. What we see as we ascend the pyramid is a.) credit going from particularised (a bilateral loan between two entities) to generalised (money which can in theory be traded at par with any agent in the economy). and 2.) the ultimate creditworthiness improves as we move up the system. It takes a much bigger economic shock to have people stop accepting the Fed’s liabilities (dollars) than to stop paying their mortgages.

We can also understand the dynamics of how monetary systems shift (such as to or from the gold standard). The fundamental shift is one of Schelling Points. When Nixon left the gold standard in 1971, we see the Schelling Point shift (from the mostly theoretical gold) directly to the liabilities of the Fed (dollars). It explains why fiat money works in the first place – it is a Schelling Point “encouraged” by the power of the state, but in theory separate from it. It explains why alternative currencies like Bitcoin can work even without any state backing, simply because they are potentially standalone Schelling Points.

It explains how the US dollar is the reserve currency: it is ultimately the strongest Schelling Point in the global economic system such that even if people in other countries do not accept their own currency, they will accept dollars. It will explain what happens if another currency takes over to become the world reserve currency – for instance the renminbi.

The theory provides a natural view of hyperinflation, distinct from the standard quantity of the money supply view, and potentially allows you to conceptualise “normal inflation” as effectively the risk premium of holding cash vs real goods. Although this is probably not a useful way to look at it and it can likely be more explained by quantity of money factors. In general how this Schelling Point theory of money interacts with changing the money supply and monetary policy in general is still an open question, and something I have to work out in more detail.

(1) Money printed by the fed is ultimately a promise that the Fed will pay you when requested. In Mehrling’s taxonomy this is automatically taken to be the best money. But why? All it is is an agreement from the Fed, that if you take this agreement to them, they will pay you x pieces of paper. It is right, but largely meaningless to say that the Fed cannot default. It is true they cannot – they will always be able to produce the x pieces of paper and give them to you. But ultimately you don’t care about that. You care about what people who are not the Fed will give you for these pieces of paper. You can’t buy groceries from the Fed. What the Fed is giving you for your piece of paper, is in effect a generalised credit on the rest of society. That you have the right to demand from them, real goods or services, equal to the value of the paper, in return for the paper.

And when would they accept this? Only when they believe they can also get equal value of real goods or services from anybody else in society for the bit of paper you have just given them. This belief is the real creditworthiness of the fed, not whether they can actually produce physical pieces of paper (which we assume they always can). It is when this belief is lost that we see events like hyperinflation occuring, which is precisely this loss of trust in the value of the ‘best money’ in the system. Importantly this is entirely about beliefs, it is not necessarily directly related to the quantity of money. Trust can be lost even if the quantity of money is unaffected. Suppose there is a war in which the US is defeated and conquered by some foreign power. The value of the dollar could decline (or hyperinflate) even if there were no actual increases in the physical amount of dollars available, because nobody would (under pain of death) accept dollars in return for payments, so the Schelling Point is destroyed. The value is lost, irrespective of the actual quantity. Of course the quantity does interact with the value in interesting ways, but the value is not solely determined by quantity.