In this section we will construct a trivial fiat currency. The purpose is to understand:
- Why the currency has value
- How the issuing authority controls the value (can it stop inflation? can it stop deflation?)
A lot of simplifications are made in this section for clarity. In the next section, we will kick complexity up a notch to make the system a lot more realistic.
Step-by-step construction of the system
- We start from a world were there is no money, no government, no taxes, etc.
- There is only one asset class, call it “housing”
- The folks in our simplified fictitious nation are divided into two categories: the wealthy, who own two houses each, and the not-so-wealthy that own no house. Historically, the latter have been renting a house each, from the former and paid for rent in kind, since there is no money
- One day, a credibly altruistic entity called “the central bank” pops out of the ground and offers to broker private credit deals
- In such a deal, a not-so-wealthy agent purchases a house from a wealthy one. The wealthy agent gets credited money on his account at the central bank, while the not-so-wealthy one has “negative money”, or debt. On a net basis, the central bank is always flat and the transaction does not change the level of wealth of either agents.
- Every day, the central bank (altruistically) decides the rate of interest paid by the borrower (not-so-wealthy) to the saver (wealthy). In this simplified model, we will ignore the credit spread needed compensate future defaults, or the issue of term (all loans are overnight and rolled)
- Now scale this up and imagine there are many borrowers and just as many savers performing this operation. Loans and cash are fungible and intermediated by the central bank. That is, a borrower does not owe directly to a saver. The borrowers all owe to the central bank and the central bank owes to all the savers.
- If the saver no longer wishes to be part of this system, he can simply go out and buy a house in the open market with his credited money
- If a borrower no longer wishes to be part of this system, he can simply sell his house and pay back his debt to the central bank. If the buyer of the house is a not-so-wealthy agent, then the loan effectively just gets transferred to him. If the buyer is a wealthy agent that paid cash, then the loan gets terminated and netted off against the cash that the wealthy person was previously sitting on
Value of the Currency
In step 5, the wealthy agent agreed to replace a real tangible asset (the house) by some electronic credits on the central banks hard drive. Who’s to say that these will be worth anything? The central bank publicly commits to a certain price-level target. For the sake of this example, let’s just say it commits to keeping house prices flat forever.
Because the central bank was assumed to be altruistic, we know it has the will to achieve its target. But does it have the means?
The only lever the central bank control in this whole system is the overnight interest rate. Let’s look at whether that knob is powerful enough to provide a credible price-level control:
First, let’s assume the currency starts to run away and hyperinflation sets in at say, 200% per year (making it extreme just to illustrate). The central bank, in an effort to contain this sets interest rates to say, 500% per year. In that situation, the trade-off between owning cash and a house is as follows: if I own a house, I will still own that house in a year (so 0% capital gains in terms of “houses”) and I will have collected rent, however much that is. If I own cash instead, I will own 6 times more cash at the end of the year (500% nominal interest rate) but it will be worth 3 times less in terms of houses (200% inflation). In other words, I will own the equivalent of two houses. Unless real rental yields are stupidly high (100%), this is an absolute no-brainer and every rational agent should rush into it. As those guys sell houses and hoard cash to benefit from the high real yield, the inflation spiral will be thwarted.
Conversely, let’s assume the system enters a deflationary spiral, the central bank can always lower the real yield enough that houses will look much more attractive than cash. Any time real rental yield (in terms of houses) is higher than real cash yield (also in terms of houses since that’s how we defined inflation), people should be flocking from cash to housing, thereby stopping the deflationary spiral. Note that that this step is predicated on the central bank having the ability to set the nominal interest rate to arbitrary large negative values. Without this, the central bank does not have a credible way to stop deflationary spirals and the system will be prone to get stuck in liquidity traps. More on that below.
Combining the last two results, we conclude that the central bank can credibly achieve whatever inflation target it wishes. Since the system is credibly stable, there is no reason for nominal rates to ever need to get as humongous as in the above examples, although leaving that door open is necessary for the system to be able to resist any arbitrary shock.
If the central bank wishes to target something other than house prices, say the price of the CPI basket, it can do so as long as there’s some elasticity between the price of houses and the price of the basket. That is, if they achieve a doubling the price of houses, the price of the CPI basket will not remain unchanged. Any well behaved utility function will lead to a higher relative consumption of the CPI basket if its price relative to the house halves, so this is a pretty innocuous assumption.
Negative Interest Rates and Liquidity Traps
As seen in the above example, without the use of negative nominal interest rates, the economy can get stuck in Liquidity Traps. In a nutshell, return prospects are so bad that even at 0% nominal yield, agents do not believe they can profitably borrow to invest. Japan has been stuck in this kind of vicious cycle for decades and other developed economies are joining the party.
When negative nominal interest rates get mentioned, there’s always someone to throw his hands up in the air and say: “this cannot work as people will just stash physical cash to circumvent the interest rate penalty”.
While this is a correct observation, the “money under the mattress” option can easily be killed by either of these two solutions:
- Eliminate physical currency altogether. It’s no longer necessary given the electronic alternatives that we could put in place, from anonymous keychain money holders to strongly authenticated mobile phone transaction. As a side benefit, this puts kidnappers, drug dealers and bank robbers out of business. Or:
- For the libertarians among you that prefer to keep a physical currency in circulation, banks can use a conversion ratio when withdrawing/depositing physical currency at banks (businesses will have to use the same conversion ratios when transacting with physical cash). The conversion ratio will have to be consistent with the compounded overnight interest rate. Obviously, the central bank can just compute and publish that number.
It is important to note that (other than the physical currency issue which is a technicality), the 0% boundary is purely in our head. There is nothing special about 0%. Chinese mathematicians started using negative numbers some 2000 years ago, it’s time for central bankers to embrace them too.
Limitations of the Trivial Model
Clearly, this model is very simplistic. Let’s outline the main limitations:
- The central bank has to make all credit worthiness decisions
- There is only one asset class
- No credit spread is charged to borrowers to reserve against losses
- All loans and all deposits are overnight
Moving right along to the Modern Banking System section where we address those limitations.
I’ve gradually come around to the “backing theory” of money. Money has value because it’s either convertible into something valuable, or theoretically could be convertible into something. This encompasses floating exchange rate fiat money, because although fiat money is by definition not convertible, there is nothing permanent about that. Offering to convert is a “sure fire” method of stopping a hyperinflation. Raising interest rates is not, because if money is worthless, it doesn’t matter how much more of it you will have in the future.
Try this analogy. Stocks are valued for the dividends they pay, and yet a stock like Berkshire that has a policy of never paying dividends still has value. That’s because even though it doesn’t pay dividends, nothing prevents it from paying dividends. Likewise, nothing prevents a solvent central bank from giving its “fiat” money value by offering to convert it. That possibility makes the central bank’s commitment to price stability credible.
“Offering to convert is a “sure fire” method of stopping a hyperinflation.” ->
Offering to convert to what? Most of what the central banks owns is nominal assets. If you fear inflation, you don’t care too much that the central bank is redeeming your money and giving you another asset exactly like it in return.
“[..] if money is worthless, it doesn’t matter how much more of it you will have in the future.” ->
This is one case where the difference between actually worthless and almost worthless is very important.
For money to become worth exactly nothing, then at some point in time, the rate of inflation must have become infinite.
If the rate of inflation never becomes infinite, then the central banker can always set a nominal rate that’s sufficiently higher than it, such that the real rate of return on money is high enough to make people want to hold it.
Now, could the rate of inflation actually become infinite? In theory, it would mean that the holders of money would have to completely throw in the towel, and voluntary hand it over in exchange for nothing. Why would they ever do that? If it’s worth nothing they might as well keep it, no?
Besides, remember that for every unit of currency created, there’s a guy who borrowed it from the central bank. That guy must either repay the principal of his loan or continue to pay interest to the central bank. Lots of it if rates are very high. The only way for him to obtain principal or interest is to get it from the money holders (who receive the interest from the central bank). He’ll be willing to give them something of value, to avoid defaulting. And there’s your demand for money.
“Offering to convert to what? Most of what the central banks owns is nominal assets. If you fear inflation, you don’t care too much that the central bank is redeeming your money and giving you another asset exactly like it in return.”
It’s OK that most of the assets are money IOUs as long as *some* of the bank’s assets are real.
Consider this modification of your scenario. Instead of starting out with nothing, assume the bank owns a house. It offers to sell the house for $100. Now money has value, and the bank’s owner can purchase up to one house worth of goods for money – assuming he can be trusted to regard the money as his debt.
If the bank is owned by a government, it can use its tax power as an asset. It can say: give us your house, or give us $100. It can then spend $100 on one house worth of goods.
This theory neatly explains why hyperinflation happens in practice. It happens when a government (and with it its central bank) is insolvent, and the insolvency is cured by inflation rather than by defaulting on bonds.
Debt monetization in normal times is not inflationary, but monetization when the government is broke is inflationary, because it forecloses the possibility of default.
“It’s OK that most of the assets are money IOUs as long as *some* of the bank’s assets are real.” ->
What happens when the bank has sold all of its real assets to prevent inflation and is now left with only IOUs? Now we’re back in my case.. so either it must work with all IOUs, or it will need all real assets. I think it works with just IOUs
“If the bank is owned by a government, it can use its tax power as an asset. It can say: give us your house, or give us $100. It can then spend $100 on one house worth of goods.” ->
If the idea was to tax the $100 to create demand for money, spending them back immediately means undoing that. On a side note, MMT is deeply flawed but I’ll write a blog post on that soon.
“This theory neatly explains why hyperinflation happens in practice.” ->
You haven’t told me what’s wrong with my theory
I think inflation happens when demand for nominal assets is less than supply because nominal rates are too low. It is true that taxation could create some demand for money, but lowering the nominal rates by some amount can always offset that. Likewise, government spending can help prop up inflation, but that can always be offset by the central bank setting the nominal rates appropriately higher. Ultimately, the in/deflationary impact of any fiscal actions can always be undone by a brave & independent central bank.
“Debt monetization in normal times is not inflationary” -> Agreed.
If the CB buys floating rate bonds issued by the government (assuming such a thing exists) and default is not a concern, that does nothing to in/deflation in a full reserve system.
If rates are at 0%, then that does nothing in a fractional reserve system either (since the banks could already source reserves at 0%).
If the bond is a fixed rate bond, then the central bank is pushing swap rates lower and that does something, but it also puts the taxpayer at risk.
Of course if the government is in fiscal troubles, having the central bank holding a ton of government debt puts the banking system at risk unnecessarily.
“What happens when the bank has sold all of its real assets to prevent inflation and is now left with only IOUs? Now we’re back in my case.. so either it must work with all IOUs, or it will need all real assets.”
If the bank has sold all its asset (both IOUs and real) and still has money outstanding, then the value of the remaining money is zero.
“If the idea was to tax the $100 to create demand for money, spending them back immediately means undoing that.”
The idea is not to create a demand for money. That is assumed to already exist. The idea is to create a value for money. In this example, $100 liquidates a tax debt of one house, which establishes a value of $1 = 1/100th of a house.
I made it too complicated. Here’s a simpler and better way of looking at it.
The bank values houses at $100. It offers to lend house owners $50 (50% of the value of the house – less than full value to protect the bank from loss) and it receives a lien on the house.
Now the bank’s money is backed by real assets – house liens. The bank’s valuation of the houses determines the value of money. As long as it never changes its valuation, the price of houses will be stable. (At least I think so – it makes intuitive sense, but it cries out for a formal proof).
Note that if the houses were completely destroyed, the bank’s liens would be worthless and so would the bank’s money (no matter what the interest rate!). The liens are a real asset which backs the money. Or in other words, if the collateral is destroyed, everyone is delighted to default and the bank is incapable of removing its money from circulation.
I’m sure I’m reinventing stuff that some long dead economist figured out, but no matter – it’s fun.
Nice blog, will add it to my reading list.
BTW, have you heard of Soft Currency Economics? It explains how the currency fiat currency works. The initial version is found here http://www.gate.net/~mosler/frame001.htm
The author has since updated it and now available as an ebook
http://moslereconomics.com/mandatory-readings/soft-currency-economics/
Hope you will find some time to read them and comment on them.
thanks again
- Nathan
Hi Nathan,
Thank you for the vote of confidence.
I have indeed read Soft Currency Economics and studied MMT fairly closely. After thinking about it long and hard, my conclusion is that it is deeply flawed but not for the reasons that most critics cite. I intend to write a series of post exposing why I think that is. Hopefully, the first one will come out within a week. Stay tuned