Recent Posts

Sunday, May 25, 2014

Savings Equals Investment And All That

One of the key points of debate in the current environment is the direction for corporate profit margins. They are currently elevated and the question is whether they will revert to a lower level or not. In this article, I discuss the interpretation of the national accounts accounting identities that relate to this debate – most importantly, the savings equals investment ("S = I") identity.

This article by Sam Ro at the Business Insider summarises one corner of the debate. On one side, James Montier and John Hussman argued that high profit margins are the result of government deficits, and thus shrinking government deficits will weigh on margins going forward. On the other side, Sam Ro quotes the strategist Jim Bianco, who objected to that argument in this fashion:
S&P profit margins are high from structural reasons, not because of the deficit.
... The crux of this specious argument that profit margins are boosted by a high government deficit and low household savings rate rests on the over simplification of “Profits = Investment + Consumption – Wages” to “Profits = Investment – Household Savings”.
This construct assumes that no savings are recycled as investment. This is not a small matter. It represents a major conceptual flaw in this framework, which taints the entire analysis. The equation above would only be correct if all savings were stuck in a Keynesian liquidity trap. This is neither the point of the argument nor the general condition, thus the equation above fails to recognize that: Investment = Savings. 
... In our view, profits are not a function of the degree to which households give back their wages as expenditures on consumption. This is a very flawed concept. This framework suggests zero sum economic terms and that savings are unproductive. In our preferred conceptual framework, investment drives growth. If households consume less then they invest more. Thus, profits are a function of cumulative past savings and the return on such capital stock as determined by risk and the competitive forces between labor and capital.

I disagree with Jim Bianco's assessment, as I do not believe that savings are (always) "productive".

The Kalecki profit equation

The Kalecki profit equation is:

    Profits = Net Investment - Household Net Savings - Government Savings - Foreign Savings + Dividends.

This Wikipedia entry describes this accounting identity. It has a lot of moving parts, and I will have to discuss it in more detail elsewhere. Within the context of this article, the key component of the equation is the fact that investment is a source of profits. Jim Bianco is correct to emphasize the importance of investment in determining the level of profits. The issue is the role of household savings within the equation.

(Note: For simplicity, I am being loose here with regards to the distinction between net and gross savings and investment in this article. Switching between net and gross amounts will affect the equation. For example, if you want profits before depreciation, you would use gross investment - net investment plus depreciation.)

What Is Investment?

There are two common usages of the word "investment". These two definitions correspond to two different concepts.
  1. Financial investment - the purchase of financial assets.
  2. Real investment - fixed capital investment or inventory investment.

Difficulties in understanding the savings equals investment identity is often the result of mixing of these two concepts of investment. The formal definition of "investment" within economics corresponds to what I refer to here as "real investment".

Investment In A Barter Economy

The core of mainstream economic models are based on barter relationships. One can visualize this by imagining that one is on a desert island, and you are trading coconuts for fish. If you do not eat the fish or coconuts, you are accumulating real goods which are some form of investment.

The identity stating that savings equals investment is obvious in this environment. What you do not consume is assumed to be an investment. Therefore, not consuming is taken to be virtuous, as it is assumed that you are deliberately investing in productive capacity.

Investment In A Monetary Economy

The savings equals investment identity also holds before a non-barter economy. But the logic is more complicated. In a monetary economy, households rarely directly invest in capital goods or inventories. This only occurs amongst households that own businesses. Otherwise, households normally save via the accumulation of financial assets. Typical examples include bank deposits, bonds, equities.

If you look at workers in a capitalist economy, their wages are an expense, and their consumption is revenue for the business sector. If workers increase savings out of their income, this means that revenue falls.  Bank deposits for the business sector will fall and the bank deposits for the household sector will rise correspondingly.

But this does not mean that profits necessarily fall. What will probably happen is that revenue will fall short of forecasts, and there will be an unplanned rise of inventories. The increase in inventories will be treated in the national accounts (as well as in financial accounting) as a form of investment. In this sense the rise in savings by workers results in higher investment. But this was not "good" fixed capital investments rather it was in the form of a "bad" unplanned rise in inventories.

This "bad" form of investment cannot happen within mainstream economic models. The assumption is that involuntary inventory accumulation is impossible because firms are supposed to lower their prices to clear excess goods holdings. But Post-Keynesian economists view that assumption as questionable, as they argue that inventories are used to buffer shortfalls in demand - not price changes.

Rising unwanted inventories cannot be sustained forever in the face of weakening consumer demand. Therefore, it will be necessary at some point for the business sector to deliberately cut back production. This planned cutback in inventory investment will then result in lower profits (based on the Kalecki profit equation). This mechanism explains why many Post-Keynesian economists do not view personal savings as being inherently "productive".

Household Savings Versus National Savings

Another point to note is that the savings term in the savings equals investment identity is national savings. In the Kalecki profit equation, we see the various savings by sectors, for example households and government savings. Corporate savings does not appear, as it is embedded within the definition of profits.

Corporate savings are undistributed profits or retained earnings. (Dividend payments are distributed profits, and acts similarly to wages in the equation, except that the payments are going to capitalists, not workers. If the owners save the dividend payments, the cash does not return to businesses, hence it lowers revenue in a similar fashion to savings from wages.) Retained earnings are a major source of finance for investment. In the current environment, major corporations are buying back their shares, rather than increasing their fixed investments. This is a sign that there is currently no shortage of finance for investment in aggregate.

But these aggregate financing relationships do not apply to every single company. For example, small startup companies need to find finance, as they do not have retained earnings to work with. Aggregate relationships cannot capture mismatches for individual entities. The role of the financial sector is act as an intermediary to bridge these mismatches.

Comparison Of Views

In summary, I feel that Jim Bianco's argument is based on questionable assumptions that are common within mainstream macro models. He relies on the savings equals investment identity to argue that all accumulation of financial assets by the household sector will add to productive capacity.

It is entirely possible that there is a shortage of capital investment within an economy, and it is necessary to encourage financial savings to make room for the increased fixed investment. But in the current environment, I do not see this as being a plausible concern. We  already have an abundance of financial capital desperately searching for viable projects to finance. Any increase in personal savings is likely to just result in an inventory buildup.

But to be fair, there is no way to determine whose view is correct. The answer depends upon the cyclical context. This is a fundamental problem with attempting to use accounting identities to make arguments about the economy. We know that they will always hold, and a scenario which violates the accounting identities is obviously wrong. But accounting identities do not give behavioural information. In this case, we simply do not know whether rising personal savings will just increase inventories, or whether they are needed to make room for fixed investment.

Final Note

My comments here are not particularly original. This is an old debate. I believe there was a quite similar explanation of this mechanism on the Fictional Reserve Barking blog a few years ago, but I could not find the article.

See also:
(c) Brian Romanchuk 2014


  1. I would agree that "Investment = Savings" in the following way: When banks make loans, they create both money and financial property. The money is in the form of a new deposit which becomes "savings" to someone, at some time. At the same time that the loan deposit is created, a loan document is created and becomes "Investment". Both money and loan documents are financial property.

    We can see now that by making a new loan, a bank has created TWO pieces of financial property of two distinct flavors. The same thing occurs with loans to government.

    The money side of a new bank loan does not disappear from the economy until it is actually used to repay and retire an "investment". Because it does not disappear, it must become someone's savings, even if that person was not a saver. The money from a loan must reside someplace, like a "hot potato", until the money is used to retire an investment made by a bank.

    1. The formal definition of investment within economics is the acquisition of new real assets, either fixed capital or inventories. This will create income flows, and add to GDP. Taking out a loan does not create income (other than residual banking fees), and does not add to GDP. It therefore is also does not count within Investment. The proceeds have to be used to finance the acquisition of new assets to count as investment.

  2. Thanks for your reply.

    Please let me explain further why I find the connection to investment difficult.

    GDP is a measure of both labor and goods that exchange in a time period, usually one year. Both labor and goods are "property" and are counted, upon exchange, as part of GDP.

    GDP is measured from both expense and receipt side of the exchange, with both measures being identical.

    Exchanges by government are a very important part of GDP. Government expenses are unlikely to be equal to government receipts with the difference being a change in financial property, resulting in a change of government debt. If there is a change in government debt, then either the expense calculation or the receipt calculation of GDP MUST include the pure debt number to make GDP (expense) identical to GDP(receipts).

    If the pure debt number must be present for government, the same thing must be true for the private sector GDP calculation. In the usual formula presentation, this is labeled as "Investment". BUT, you can see that "Investment" in the private sector is going to have a character similar to what we find in the government sector - not investment in the traditional use of the word.

    Finally, on the receipt side of the GDP calculation, we frequently see "Savings". Savings would simply be the sum of Investment and the government deficit and would be a pure financial number.

    "Savings" is a very inappropriate name for the borrowing and change in national financial balance sheets that drive year to year changes in measured GDP. The increase in money supply driven by bank lending and government printing of money is certainly not savings in the traditional sense of spending-less-than-income. This macro-savings is the result of money supply expansion and an increase in the quantity of financial property.

    If you have followed my logic to this point, thank you for your patience. I recognize that my logic is unorthodox but it seems correct to me. I appreciate any further thoughts or critique of logic.

    Thank you for this and many other thought provoking postings.

    1. I like looking at these things transaction-by-transaction. You categorise each transaction, then the aggregates take care of themselves.

      The only government transactions that add to GDP are those that involve the purchases of goods (or labour). They are either classified as government consumption or investment. (I would need to look up how government-provided services are handled, like free education.) Transfers and taxes do not directly affect GDP, but they will implicitly raise or reduce the consumption/investment of other sectors.

      I have to run, so I cannot look too carefully at what you wrote. But I think that if you think about how government counts in GDP, you see that the financial aspects (debt issuance) is somewhat distinct; it shows up by its impact o other sectors.

  3. A bit late to comment, but let me thank you for this post. It clarifies things. :)


Note: Posts may be moderated, and there may be a considerable delay before they appear.

Although I welcome people who disagree with me, please be civil.

Please note that my spam comment filter appears to dislike long "anonymous" posts. I get no warning about this, and only go through my "spambox" infrequently. The best bet it to keep comments short, and if you think the spam filter struck, let me know with a short comment.