Saving and Investment by Degrees

Sam Levey
10 min readSep 6, 2020

There is a massive amount of confusion, both in popular discourse and in actual academic economics, around the concepts of saving and investment. Much of this trouble hails from a very deep misconception of what “finance” is.

Finance refers, in the most basic sense, to your ability to pay for stuff. In an economic context we’re often talking specifically about paying for investment goods (eg. factory machines and inventories), and this is because a basic dilemma in capitalism revolves around the fact that production takes place before sales — that is, businesses have to pay to produce a good, that only later will they sell to earn revenues that will recoup their costs. In the meantime, how will this production be financed?

We understand intuitively that borrowing of some kind needs to happen here. And the way we understand borrowing is that you take something now and sometime later you will return that thing (possibly plus interest), eg. I borrow your car today, I’ll give it back tomorrow. This is the basic mental model that gets brought to the savings/investment debate, only with money: I go out and find somebody who will give me $10,000 today, and I will return it to them next year, plus interest. I’ll call this the “finding money” model.

But really, even at this basic level, we’ve already gotten the story wrong. We’ve combined the pieces in the wrong way, and scaling it up to the whole economy will only compound the error.

The basic problem here is that this story privileges “money” above other financial instruments, assuming that money and the monetary system pre-exist the need to shift resources around: there’s “money” out there, and if I want to invest in my business, I need to go out and find some. But this story conflicts with claims put forward by Modern Monetary Theory that money isn’t special or superior, but rather is just another form of financial instrument, like any other, and comes into existence in the same ways. So we need a different story.

For MMT the basic unit of finance is a promise, an IOU. The standard story used to be that humans were forced to use barter to distribute resources within society prior to the origin of money, meeting to negotiate prices in spot transactions where one good was traded for another. Historically we now know that this is false; but the reason it’s false is deviously simple: there’s no need to stop and haggle if we’re neighbors and I can just “get you back later” for whatever you give me now. In the modern parlance, we can just write each other IOUs. This is a better candidate for the building block of finance, as this kind of direct credit predates formal money systems by thousands of years.

“Writing an IOU” as we’ll see is a very different puzzle piece for our finance story then “finding money.” The most significant difference is that in the “finding money” story, the money needs to pre-exist the transaction — it’s out there already, and we have to go get it. And who are the people who have money? It’s the people who earned it but haven’t spent it yet, i.e. savers. This is reinforced by some basic accounting identities you pick up in Macro 101 which say that Saving and Investment must be equal. The simplest possible version of this starts with the GDP definition for an economy that has no government or foreign sector, so that sources of income (Y) are Consumption (C) and Investment (I), so Y = C +I, and uses of income are Consumption and Saving (S), so Y= C+S. This means that C+I=C+S, and voila, I=S. This is interpreted in the “finding money” model to mean that savers of money provide the finance that gets used for investment.

So the “finding money” story focuses our attention on saving and savers as the ‘prime mover’ of investment. But what is their role in the “writing an IOU” story?

Suppose again that I need goods/services for investment in my business: the simplest way to finance this is for me to write an IOU directly to my suppliers. They give me the goods now, I give them the promise of something of equivalent value in the future. This isn’t a standard terminology, but for convenience I’ll refer to this as “first-degree finance.”

There are certainly some inconveniences here, but probably less than you might think, and in fact this is done all the time in the real world. If money didn’t exist you might presume that the problem would be the old “double-coincidence of wants” problem, eg. if my business makes vacuum cleaners, am I going to repay you by giving you 100 vacuum cleaners that you don’t need? But issue evaporates if we make the IOU divisible and transferable: you can keep my IOU for 1 vacuum cleaner and sell the other 99 to other people. Really the problem is creditworthiness: do you, and the people you might sell my IOU to, trust me to actually make good on the IOU, so that you get repaid? This has something to do with my personal character, but also has a lot to do with the kind of business I’m running and the environment I’m running it in.

There is an innovation that can make this easier to handle, and that is the use of a third party. If my suppliers aren’t sure they want to accept my IOU because I might not make good, then I can instead find a outside person, some entity trusted by both of us, and we can offload the risk onto that person — I write the IOU to the third party, and then the third party writes an IOU to my supplier. This alters the risk dynamics, because now my suppliers don’t have to worry about whether I will honor my promises, that’s the middleman’s problem. This intermediary then might specialize in evaluating the creditworthiness of people who want to issue IOUs.

This third party is called a “bank,” and this is the basis of all bank lending. The bank accepts the IOU of the borrower (the “loan”), and issues its own IOU to the supplier (the “bank deposit”). The catchphrase here is “loans create deposits,” and because these bank IOUs are so widely-accepted we tend to think of them as “money.” I’ll refer to this kind of thing as “second-degree finance”: if the supplier won’t accept your personal IOU, you can find a bank to be your IOU-intermediary.

The “finding money” story really revolves around what we might call “third-degree finance,” in which already-outstanding IOUs get ‘recycled’ for new transactions. One way I could come into possession of previously-issued IOUs would be to earn revenue from normal sales, where my customers paid me using bank deposits, perhaps drawing on their checking accounts. If I still have these lying around as retained earnings, then I can use them to finance investment. Alternatively, I could acquire these outstanding IOUs by borrowing them: I find somebody holding bank deposits, and offer them my IOU in exchange. Then I use these bank IOUs (“money”) I have just obtained to pay my supplier for my investment goods.

This third-degree in general and that last transaction in particular are the cookie-cutter mental models of investment finance for proponents of the “finding money” story, who have in their head something like the bond market. In some cases these people deny the existence of the first and second degrees, misconstruing what happens when banks make loans; more commonly, the person will acknowledge that first/second-degree finance can happen, but regards them as some sort of “perversion” or “distortion,” as if the third-degree is some “natural” baseline that we all ought to be subjected to which the rapscallions engaging in first- and second-degree finance are skirting somehow.

Either way, admitting the existence of first- and second-degree finance forces us to rethink the savings/investment relationship. In the third-degree transaction, the “saver” is usually identified as the person who supplies the money, i.e. holding the pre-recycled bank deposits. First, it’s worth pointing out that this person need not actually meet the conventional definition of “saving,” which is to consume less than your income — it’s equally possible that they received the bank deposits by borrowing them from a bank, which would have created them from thin air. But more importantly, that person doesn’t even exist if investment is financed through the first- or second-degree! So how do we interpret the Saving=Investment equation in these cases?

What we have to understand is that these accounting constructions are meant to capture production, and not necessarily financing. “Investment” is counted as increased when the factory machine is produced — initially this is counted as “inventory investment” on the part of the producer. And these accounting constructions are purposely designed to omit transfers of already-existing assets: when the factory machine is sold, the “investment” of the purchaser is canceled out by the “negative inventory investment” of the supplier, so that total investment and GDP are unchanged by the sale. Because this production constituted income (“real income”), but nothing has been consumed, the act of producing the investment good was itself an act of “saving,” in that the supplier received income that they haven’t consumed. When the factory machine is sold with first-degree financing, the supplier still has the same amount of savings, but now it has changed form from a physical asset to a financial asset — it is the newly-issued IOU of the borrower that now constitutes the savings of the supplier!

We can see pretty clearly in the case of first-degree finance the old Keynesian adage that it is investment which creates saving, and it is not the case that we need prior saving in order to finance investment. And it really doesn’t change at all when we move to more complicated financial schemes. Even in third-degree finance, 1) as above, third-degree finance does not actually require prior saving, and 2) even when the provider of funds is a “saver,” it is not their saving that is referred to in the Investment = Saving equation! To see that, remember that the accounting constructions in GDP are all defined for a particular period, eg. Q1 2020. I=S implies that both the I and the S are happening in the same period, but the person who does the lending need not have came across the funds in the current period at all. Meanwhile, the act of producing the investment good, as before, is real income and saving for the producer, which is then converted from a physical asset to a financial asset when the sale is made. At least initially, the accounts would show that the saving of the supplier as exactly equal to the amount of investment, regardless of how it was financed. And in fact, the savings of the lender would not be affected by the loan: it merely changed to a different type of financial asset, where initially it was a claim against the bank but now it’s a claim against the borrower. Still just IOUs either way.

So far we’ve seen that saving and financing are distinct from each other, and that it is investment that creates saving. From this latter point follows a very important fact: individual decisions to save more do not increase total saving! What they do is transfer savings around. Suppose I receive income, and then I’m choosing to either consume or not consume — obviously, if I refrain from consumption then my saving will be larger by exactly that amount, whereas if I consume then it won’t be. But what happens to the rest of the world if I consume? Suppose I purchase a service like a haircut: my purchase creates income for the producer, which, at least initially, increases that person’s saving by exactly the same amount that mine falls¹! My act of consumption didn’t decrease total saving, it merely transferred some savings to another person. Likewise, that person now can choose to consume or to save, but choosing to consume wouldn’t reduce total saving, it would just transfer it to somebody else.

Why does all of this matter? Because if you’re looking for a “finding money” story then you’re ultimately going to come to incorrect conclusions about what determines interest rates, and therefore their role in producing and/or stabilizing the business cycle. Internet Austrian economics people are especially guilty here, but mainstream economics falls prey to this as well. If your mental model is that third-degree finance is the base case (and only the variant of third-degree that actually involves prior financial saving) then you’re eventually going to reach a story where interest rates have something to do with saving (either they do, or they should, or both). On the other hand if we treat first-degree finance as our mental base case, then it becomes pretty obvious that interest rates are the price of finance, not saving, and we need not have prior saving in order to provide finance. And when we factor in that individual saving decisions don’t determine total saving, we realize that it is not possible for individual saving decisions to determine interest rates.

We can see from the first-degree case that both the saving and investment are created at the moment that the physical good is produced. Other transactions might shuffle around balance sheet entries, or change the form of the savings or who has them, but they don’t change the basic relationships. This absolves us from the fallacy of pre-assuming the existence of money.

In fact, money creation provides a good analogy for all of these transactions. If we use the strict definition of “the money supply” that statisticians use, then it’s true that only second-degree finance literally “creates money.” But if we zoom out and consider all financial assets as having some degree of “money-ness,” then all three kinds of finance invoke this magic of money creation in some sense: in each case, the borrower receives real goods against only a promise. While third-degree finance does additionally make use of pre-existing money, it only really comes in as like a catalyst, allowing the chemical reaction (the transaction) to proceed, while emerging unchanged at the end, available again to catalyze other transactions.

— — —

¹ It’s a bit more complicated if I purchase a good instead of a service. The income was actually created when the good was produced as inventory investment, which also created an equal amount of saving. My purchase of it to consume decreases my saving by an equal amount (and the producer records negative inventory investment), so that the change in total saving is still zero at the end. Because producers of goods don’t want to produce goods that nobody is going to buy, inventories of consumer goods won’t tend to grow faster than consumption on average, and so it’s still good enough to say that consumption spending transfers savings.