Sectoral Balances: (Mis?)understanding NAFA and Net Lending/Borrowing

NAFA, or Net acquisition of financial assets, is an important measure, often rather a crux, in Post-Keynesian, MMT, and other economic thinking and writings. It’s central to understanding sectoral balances graphs, for instance. So it’s surprising to find widespread and long-standing confusion about what that label means.

In footnote 26 to a recent paper, I highlighted how Wynne Godley and Marc Lavoie (G&L) in Monetary Economics use a different (and IMO better, clearer) label for the measure. But of far greater issue: “NAFA” there refers to a different measure from the one most economists think of. This post is an attempt to sort that out clearly.

The national accounts measure being graphed in sectoral balances is “Net lending (+) or borrowing (-)” for each sector’s Financial account. Here’s that measure and its subcomponents in the Integrated Macroeconomics Accounts household table S.3.a. (See also the Financial Accounts’ household Transactions table F.101.)

40    Net acquisition of financial assets
67    Net incurrence of liabilities
77  Net lending (+) or borrowing (-), financial account (lines 40-67)

Most people naturally think that NAFA refers to the top line here. But in G&L, it’s actually referring to the bottom line. And the longhand version of NAFA in G&L is called net accumulation, not acquisition. Both differences matter.

To begin with, Net lending/borrowing is a problematic and confusing label. Look at the table for any year, and imagine three possible counterfactuals.

  1. Government transfers an extra $1T to households. Nothing else changes. Households’ assets and so net lending/borrowing increase — with no household lending in sight.
  2. The household sector spends an extra $1T buying goods from firms (which households consume). Again, nothing else changes. Households have less assets, so less net lending/borrowing, while no actual borrowing happened.
  3. The household sector borrows more from the financial/government sector(s). No other changes. Households have more assets and more liabilities — an “expanded balance sheet” — with no change to Net lending/borrowing.

A change in net lending/borrowing, with no lending/borrowing. New borrowing, with no change to net lending/borrowing. No surprise if people are confused.

It’s probably best to understand Net lending/borrowing as a stylized label, term, usage, rooted in a particular understanding: “You can only increase your financial asset holdings if some other unit increases their obligations to you, their liabilities. So one unit’s asset increase is effectively ‘lending’ to other units.” There are some problems with that understanding, but it does let us see where the terminology comes from. “Effectively” is invisibly carrying a lot of water there. (“We don’t mean ‘lending’ literally!”)

“Acquisition” is another potentially confusing label. To many, it may seem to imply purchases and sales of financial assets. Couple that with another widespread misconception, and confusion is likely. Many think somewhat vaguely that buying (acquiring) bonds or equities increases households’ stock of “savings” or some such. But really that’s just portfolio rebalancing, a dollar-for-dollar asset swap: fixed-price M2 assets exchanged for different, variable-priced assets (or vice-versa; it’s a swap).

So if households “acquire,” purchase, more equities for example, they’re just rebalancing the sector’s portfolio mix. There’s no change to total assets. Hence, likewise, Net lending/borrowing is unchanged.

Godley and Lavoie nicely avoid that confusion; their NAFA means Net accumulation of financial assets. See for instance their Appendix 12.1 (pp. 490-492), which lays out the sectoral balances identity though not by that name.

But the far larger confusion remains: the NAFA in G&L isn’t our usual NAFA. It’s actually our familiar Net lending/borrowing, the bottom-line measure, not the assets measure. 

Lavoie has acknowledged as much in a recent private email (which he’s given me permission to quote):

Yes, you are right, in the Godley and Lavoie book we used the terminology that Wynne had used, which was not always consistent with that of the national accounts, in particular in the case of NAFA. I used the proper terminology in my 2014 book on post-Keynesian economics. [Post-Keynesian Economics: New Foundations.]

He does indeed address this issue (see eg pages 260 and 515), but not with the succinct clarification I’m hoping to provide here. To that end, here’s a proposed revision to the national-accounts labeling (slightly verbose, for clarity).

40    Net accumulation of financial assets (gross inflows minus gross outflows). NAFA.
67    Net incurrence of liabilities (gross new borrowing – loan payoffs and writeoffs). NIL.
77  Net net accumulation of financial assets (lines 40-67). NNAFA.

The double “net” serves to nicely clarify the whole construction, in my opinion. This is at least worth keeping in one’s head, to metabolize the web of accounting identities/definitions at play.

Note that these Financial-account measures are all changes in “volume,” versus “valuation” (capital gains/losses). The latter, which are far larger than the volume changes, and consistently positive over more than six decades, are tallied separately, in the IMAs’ Revaluation account.

  1. No comments yet.
  1. No trackbacks yet.