Does Saving "Fund" Investment?

If Asymptosis has any tiny claim to any important influence, it might be that anonymous and magisterial commenter JKH used the comments section here to first bruit his insight (both tautological and profound) that S = I + (S – I).

He revisited that construct and concept again recently, and I’ll leave it to you to explore his very interesting thinking.

But I do want to address a central issue in that discussion: the notion of “funding.”

JKH quite properly uses standard flow-of-funds accounting terminology to explain that private-sector “saving” “funds” both its “investment” (buying/creating drill-presses and such), and its acquisition of newly created financial assets.

Quite properly, but: it’s important to understand what that key accounting verb (“funds”) actually means. It describes an after-the-fact and arguably largely arbitrary accounting allocation of income streams to outflow streams.

Imagine 2011, a year in the life of BFC Corp.:

Profits (revenues – expenses): $100,000
Net Borrowing (borrowing – loan payoffs): $100,000

Investment (spending on drill presses and such): $100,000
Dividends paid to shareholders: $100,000

Looking back: Of the $200K in inflows, which part “funded” the investment spending on drill presses? What funded the dividend payout? The accountant’s allocation decision, absent any other information, is after-the-fact and completely arbitrary. Funds are fungible — especially when viewed in retrospect.

Before-the-fact conditions and restrictions might well give justification for a given after-the-fact accounting allocation decision. If BFC decided in 2010 to spend X% of profits on drill presses in 2011, and that X% came to $100,000, an accountant after the fact might quite reasonably say that the drill-press purchases were “funded” by that year’sprofits.

Alternately: Imagine BFC “set aside” $100,000 from 2010 profits for future drill-press purchases by “funding” a drill-press holding account on their books, debiting their 2010 profits to “fund” that holding account. (Maybe it even created an actual external bank account to hold and segregate those funds, though that’s not actually material to this discussion.) It then spent down that holding account in 2011 to buy drill presses. Were those purchases “funded” by 2011 profits or borrowing? The proper accounting answer here is “neither.” Looking backwards you might/could/would say that they were funded, ultimately if somewhat arbitrarilly, from 2010 profits, or from the holding account. Either is accurate, depending on how you telescope your “funding” pipeline, in both time and account-space. (This is all rather like discussions of the Social Security Trust Fund.)

When we say, in a backward-looking flow-of-funds statement, that “X funded Y,” that is an ex-post description that is informed, and arguably justified — but not fully or authoritatively determined — by knowledge of before-the-fact intentions.

So when we say that “…the marginal dollar borrowed by a nonfinancial business [post-’85] was simply handed on to shareholders, without funding any productive expenditure at all,” we are making a statement about what “funds” what. We’re saying that all the borrowing went to payouts, and all the profits went to investment. The reverse could be equally accurate, given that shareholders from ’04 to ’08 were paid about $200 billion more than their companies earned in profits.

Let’s try this on the level of national/international accounts, and sectoral flows. Here’s mythical 2011 accounting for Bandalaria:

Assume (purely for simplicity in explaining the “funding” concept) that:

1. There is no net trade surplus or deficit, and the country’s capital account balance sheet remains unchanged.

2. The central bank does not increase or decrease its holdings on net.

3. The financial system does not increase or decrease its loan book to the private sector.

That leaves two sectors, with (looking back) no accounting impact from the above:

• Federal government (Treasury)

• The nonfinancial private sector (nonfinancial firms and households)

What do Bandalaria’s net money flows look like?

From Treasury -> Private
Deficit (purchases minus taxes): $100 million

From Private -> Treasury
Treasury bond purchases : $100 million

Looking back, how would you describe these flows? Are are the bond purchases “funding” the deficit, or is the deficit spending “funding” the bond purchases?

The correct answer is “Yes.”

Likewise: when JKH says (my words actually) that saving (income – expenditure) by the private domestic nonfinancial sector “funds” both its investment spending and its net acquisition of new financial assets (including government bonds), his description is perfectly correct.

But he would equally correct if he said that government deficits (less trade deficits) “fund” some of the investment, or (all of) the acquisition of new financial assets (notably government bonds), by the private domestic nonfinancial sector (or some of each).

Obviously, the two accounting-based descriptions, both accurate, have very different rhetorical implications.

This just reiterates the point I made in the post to which JKH responded with his revelatory identity: accounting tells us nothing about economics, except that it often tells us when economic thinking doesn’t make any logical/arithmetic sense.

I guess my main point here, perhaps obvious to many, is that accounting descriptions — choices about how to describe the past in accounting-speak, especially regarding “saving” and “funding” — are, inevitably, rhetorical hence normative. Or at least, those choices of descriptions have inevitable rhetorical hence normative implications.

Or to put it simply: accounting is normative.

My impression is that many economic discussions and disagreements, especially in the “MM” worlds, are at their root disagreements about what “funds” what (frequently compounded by imprecise sector definitions with different parties using different implicit definitions), and the rhetorical hence normative implications of those competing descriptions.

Cross-posted at Asymptosis.