Enough money to pay interest? Steve Keen says “yes”
So one of the surprising things I learned from the Money from Nothing video is that when money is created through debt, enough money was created to repay principal, but not the interest. This made logical sense to me, and so I held this as an article of faith right up until yesterday, when I watched the following explanation from Steve Keen in a talk he gave sometime back in 2012:
From Keen's economic model, it was clear everything worked out just fine, and there was plenty of money to pay interest AND principal. How is this possible? It seems mathematically impossible.
He ran through it really quickly, but I played it several times and then thought about what he said. Primarily, it's a stocks vs flows misunderstanding. Let me try and explain what I learned.
So first, the MFN video and Keen both agree – the total credit money stock is borrowed into existence through a promise to pay – money is created through loans, and it is destroyed through repayments and defaults. But then they diverge. Let me explain why I think Keen is right.
First I will use a simplified example to show how a given money flow can result in satisfactorily paying the interest load on a given amount of borrowed money stock.
First, let's assume the total loans in the system are steady state. Money has been loaned into existence, and everyone has some, and then it doesn't change. This is just to keep things simple.
Now then, take the case of one borrower. For our purposes let's make it a company that produces widgets, on which it makes a profit. To finance its inventory, it borrowed $1M from a bank @ 5% interest per year – as an "interest only loan" (again, for simplicity). The loan is static. Money supply is static. According to MFN, the company should die eventually.
However, in our example, the company turns over its inventory 5 times per year, and makes a 3% profit on each inventory turn. Based on these numbers, can the company pay the interest?
3% profit margin x $1M x 5 turns = $150k profits… yes, the $50k annual interest payment is no big deal. Clearly, MFN is wrong.
The key trick here is "flow" part – in this case, the "flow of profits" (inventory turns x profitability per inventory turn) is definitely fast enough to overwhelm the cost of the interest payments. Another way of looking at this is the following equation:
Money Flow Per Year ($5M) x Profit Margin (3%) > Loan Stock ($1M) x Annual Interest Rate (5%)
As long as this statement is true, all is well. However if the money flow rate drops too low (i.e. if inventory turns happen only 1/year), or if profitability tanks, then things blow up for the company.
How can money flow be so much higher than money stock? That's just how things work. Most workers bank accounts don't have a bank account containing their entire annual salary. People living paycheck to paycheck "turn over" their bank account every month, assuming they have one month's pay in their bank account. That's a flow-rate of 12! $36k in annual salary (flow) with a balance (stock) of $3k. So clearly, the economy doesn't need to have $1 in money stock for every $1 in annual activity. The same is true for companies.
Check this with our actual economy. M2V is 1.52, which say that $11T in money stock (M2) supports $17T in total activity (GDP) by "moving through the economy" (i.e. changing hands) 1.52 times per year. And so as long as profitability x flow rate > interest burden from the debt, all is well.
Note: this assumes that interest payments are recycled, rather than being taken out of the system and squirreled away on some desert island. Interest payment flows from borrowers are turned into bonuses for bankers and dividends for shareholders – which are either spent on assets like houses in the hamptons for the bankers and salaries for the servants, or lands in the bank accounts of the shareholders – bottom line, every dollar flowing through the system (profits, wages, interest, etc) lands in a bank account at the end of the day, where it then sits until the owner decides to either spend or invest it. The key takeaway is – money flowing through the economy is neither created, nor destroyed, unless through loan default or repayment.
And its flow rate x profitability that define the interest burden each participant can bear, not the total size of the money stock x its initial interest rate at creation.
So that's how the MFN video was confusing stocks & flows. I'm not sure I made this clear; perhaps listen to Keen, maybe he did a better job of explaining it than I did.
Of course, this is not saying debt is fine. Its not, and Keen talks about this at length in that same video. Turns out, our money stock of $11T is vastly dwarfed by our total debt of $59 trillion. There are almost $5 in debt claims for every $1 in money stock! Its another warehouse receipt problem, but this time, for paper! And for each claim, there is an interest flow attached. That's a lot of interest flow. Most of that flow goes through finance sector. As Keen says, these guys are very motivated to sell debt.
The whole video series is worth watching, if you have a few hours to kill and don't mind hitting "replay" when he goes by things really quickly.
Steve Keene is being a complete egghead on this one, by which I mean utterly divorced from simple real-world realities.
I agree with the highly simplified and utterly unrealistic set up which has all flows of money coming back into the bank and then flowing back out into the world, perfectly balanced with all stocks, and no accumulations of said stocks at any particular points.
Under those conditions of idealized and perfect stocks and flows it's theoretically possible to make it all balance out.
However, out there in the real world, where there's $57 trillion in debt, it's impossible to have all $2 trillion in debt remittances flow into the bank and back out as wages in a manner that prevents exponential growth in the money system.
By way of evidence I have charts of both debt and money spanning many decades with near perfect exponential growth. R^2 of 0.99, baby!
So what does it matter if it's theoretically possible under heavily constrained conditions in a stripped down spreadsheet to make stocks and flows balance for a couple of turns of the crank? Stocks and flows are never ideal or perfect and, because of this, you get the exponential behavior we see in the real world.
Chris, I encourage you to sit through the entire set of videos, although they are long, so perhaps you don't have the time.
In the remainder of his talk, Keen is not saying "things work out" with our 57 trillion in debt. He's saying that on this one specific claim – "its not possible to pay the interest on the debt stock beacuse the money isn't created" – it is a stocks/flows fallacy.
In the rest of his talk he shows that bankers have this innate desire to create more debt, and this desire (and their eventual control over government through their profitability) coupled with the ratchet effect will inevitably and repeatedly drive us into crisis as debt continues to grow until we finally blow up from a massive debt bubble – exactly as you pointed out. You and he are in complete agreement.
He did NOT say that the stocks/flows situation means there is no exponential growth in debt. But he did seem to suggest it wasn't because of the interest – but rather, the ratchet effect. No – wait – it was the whole Minsky ponzi finance that first drives everything up, and then breaks down when the "fundamental buyers" start to sell assets because they can no longer cover their interest payments with their cash flow, and that triggers the pop.
I suspect I explained his case poorly. I really was only targeting the one narrow claim: "there isn't enough money to pay the interest." Nothing else.
I posted this in "economy wonks" because this is kind of a wonky, narrow subject in the first place. I've been working on an overall model of the economy so I can understand how things function from a capital flows point of view, and that's why I discovered (and sat through) this talk in the first place. Points that doesn't matter so much from a "boy our debt is massive and growing exponentially" point of view turn out to matter quite a bit more when you're developing a model for how stuff really works.
This guy took a kind of naive approach.. not understanding the nature of debt-money but feeling his way around the situation. He decided debt money was a dysfunctional con – there is a nice chart in the piece if you want to look at it. ;
Total Private sector debt is $89.27 trillion and Total Public Sector debt is $47.62 trillion, making a total debt level of $139.89 trillion.
Adding together the money supplies of all the countries together produces a total of $68.34 trillion. That is exactly half the level of debt.
In other words, even if every last cent was added together, we could still only pay off half the debt. In other words, the 2.5:1 ratio of debt to money supply that I noted for the Eurozone is a pretty typical case.
The last column gives the ratio of debt to money supply for each country. For some reason, Scandinavian countries like Norway, Sweden and Denmark all have very high ratios of well over 4:1. But the USA is also up there with a debt to money supply ratio of 3.5:1.
There are only five countries that actually have a money supply large enough to cover their debt (coloured in green in the table). Apparently Mexico is one of them, but this seems extremely odd – maybe an error in there somewhere.
The other four are China, Hong Kong, Saudia Arabia and Luxembourg. That seems to make sense. But even if you combine all their money supply surplusses, you still only get about $18.5 trillion. So, even they are totally unable to help pay off the mountain of debt that the world has amassed.
I find these figures quite incredible. They demonstrate quite clearly that those who have been lending the money that we owe can't possibly have had the money they lent. The whole thing is a complete con.
So this was the puzzle I was trying to solve. How can this be? How can it not collapse instantly?
Its a wonky narrow question, but one that really puzzled me.
Again, its a stocks & flows thing.
As long as there is profit in the economy, all that debt CAN be paid down over time. I can come up with a model that shows you this. This is hard to wrap your brain around. I encourage you to try, its an interesting experiment that will end up giving you insight if you stick with it.
We have 59.4 trillion in debt. 11.7 trillion in M2. How could we ever pay it all down?
From a macro viewpoint, according to Keen, Aggregate Demand = GDP + change in TCMDO.
So this means we'd have to go an an extended aggregate demand diet. We'd have to forego one trillion dollars in consumption over a 59 year period (not including interest, of course). Nasty effect – aggregate demand drives employment. A persistent drop in aggregate demand like that would whack employment for 59 years.
Here's something else I figured out as part of this process. Not all debt is money – credit money, I mean.
Bonds – not money. Only bank loans are money. If you repay all the bank loans then yes, no money. But government debt? Not money. Its just a marker that says money was taken from someone's bank account and spent into the system (landing in yet another party's bank account). And of course, an interest payment is attached to the marker.
How about repaying the bond? That's just money going from every taxpayer's bank accounts back to the original lenders bank account that bought the bond in the first place, thus extinguishing the bond. No money was created OR destroyed by the repayment of the bond.
Money is only destroyed when bank credit is repaid or defaulted upon. Its the magic of the bank's balance sheet that lets them do this. If bank credit isn't involved, the total money in the system (from a total bank credit point of view) is unchanged.
Stock market is the same from an M2 viewpoint. If a stock is sold, it is "money supply" neutral. Money is transferred from buyer to seller – neither created, nor destroyed. If the stock drops by half and is then sold: only half the money is then transferred from buyer to seller. Total account value drops for sure – but stock market valuations are not part of M2. Regardless of bull or bear markets is bank credit neutral. Stocks are just a money transfer vehicle, at some level. Just like bonds.
Again, a narrow wonky thing that only matters to model-builders.
I'm still trying to figure out what it all means from a macro viewpoint.
Dave, I don't have time for an extended dialogue right now.. but you pushed one of my hot buttons right here;
Bonds – not money.
This is not true. Gov't bonds represent their borrowing. This borrowing is spent in to the economy.. to pay for everything the Gov't spends money on. This borrowing is therefor MONEY. When the Gov't borrows money to pay for deficit spending, then a small part of that deficit spending goes, for instance, to Raytheon to pay for some cool military hardware. That money.. the money that was created as the bond, then flows through Raytheon and ends up in the bank accounts of the engineers working for Raytheon, and the suppliers who provided the components… etc. This is real, spendable money. It is not some abstraction as you would like people to believe based on the fact that reserves are created simultaneously and kept at the FED… no, there is real money being spent, and ending up in banks, in the accounts of engineers like me even.
Relax, I'm not saying what you think I'm saying. 🙂
We almost completely agree. Please bear with me.
Yes, govt bonds represent government borrowing. That borrowing is spent into the economy. It most definitely adds to velocity of money. It has a material (inflationary) effect on things. Its bad.
But from a tracking standpoint, the bond itself is not money – the creation of the bond by the government doesn't increase M2. All it does is transfer bank credit from the buyer of the bond to the government's bank account, so they can then spend this (previously created) private bank credit into the economy.
Think of the steps in the transaction:
Government offers bond
Buyer buys bond (I did that just last month); bank credit is sucked out of my bank account, and transferred to government bank account.
Government then spends what used to be my bank credit to buy a bomber; Boeing workers get paid, money gets dropped into their bank accounts. No money was created thereby – my money ended up in Boeing workers accounts. And I own a bond, which is just a marker for the government's debt to me.
When it comes time for the bond to expire, government transfers bank credit from its account to mine – money it has taxed or borrowed from someone else. My bond is paid back. But again, no credit was destroyed by the repayment – just transferred from a taxpayer to the government and then to my bank account to repay that bond.
Note: US treasury bonds can be traded for money, but they are not themselves money. They are not bank credit money. Not M2.
Again, this is a model thing. I'm trying to figure out how to use the various measurements of the system to track economic activity. That requires I really understand what's being measured by which measurement. Its perhaps too technical or too wonky for most.
But at least I think its interesting.
Government creates massive debt: M2 remains unchanged, however velocity goes up.
Banks make massive loans. M2 skyrockets, because through the magic of that banking charter, the thing created by a bank loan is – by definition – credit money.
What happens when US bank credit goes overseas – say to buy trade goods? That's one I haven't figured out yet! How do we track US dollar deposits overseas? Someone probably has to.
Again, US dollar deposits at a Chinese bank is materially different than (say) Chinese ownership of US treasury bonds, or US stocks, etc.
Simple proof: you can buy pretty much anything with bank credit. Legal tender, all debts public & private. But you must sell a treasury bond before you can buy anything with it. Therefore, T-bonds aren't money.
They can be TRADED for money, but they aren't money. They simply represent the promise that the owner will receive a transfer of money at a date certain, assuming no default occurs.
Thanks for you interesting writings, however I am not with you on this one:
"money is created through loans, and it is destroyed through repayments and defaults."
Why is money destroyed when defaults happens? The money appear as an loss in the banks balance sheet. It is not destroyed in my opinion. Can you explain what I am missing?
Pellepan from Sweden
Response #1: I base my contention that bank credit money is destroyed by defaults on certain timeseries provided by FRED. Specifically, there is one labelled "Total Bank Credit" and it is the sum total of all outstanding loans on the bank balance sheets in America.
When a loan is defaulted on, the bank writes the loan down, and that writedown makes TOTBKCR drop by that amount. Same thing happens when the loan is paid off.
Since bank credit is money, and when bank credit declines, money vanishes, anything that results in total bank credit declining destroys money.
Response #2: If #1 is too circular for your taste, try this on for size:
Deposits are inextricably linked to loans on the balance sheet through reserves. Losses on loans hit reserves directly.
1) If enough loan defaults occur, reserves are wiped out, the bank goes under, and the bank deposits (the liability side of the balance sheet) are reduced by the same amount as the losses in bank assets. That's one very straightforward transmission mechanism for destroying bank credit.
2) In a less severe case, losses on assets hit reserves, but not enough to take the bank down. In that case, bank reserves must be rebuilt over time. The only way this can happen is by the bank siphoning off income (i.e. cash flow through the bank) and rebuilding reserves, which removes that credit money from circulation. Normally, interest payments flow through the bank to shareholders, who then either deposit them or spend them into circulation.
Thanks for your explanation Dave, makes sense.