Saving, Investment, and Lending in the Real Economy (Graphs). S=I?

With all the chaff that’s been flying around (recently, and for years now) about saving and investment, dissaving, and lending/borrowing, I felt the need to go back to the numbers and see how they’ve played out over the decades in what we tend to call the “real” economy — domestic households and nonfinancial business. Click for larger:

Update: The signs were reversed for lending/borrowing. Graphs corrected and updated.

Screen shot 2013-04-16 at 7.17.56 AM

Here’s the lending/borrowing broken out for you:

Screen shot 2013-04-16 at 7.36.17 AM

This is all from the Fed FFAs. Saving is household/nonprofit net saving (after-tax/transfer income minus expenditures) + undistributed business profits (after-tax/transfer income minus expenditures and distributed profits [distributed profits are part of household income]). Details/spreadsheet on request.

I’ve actually written at least three (long) posts on this in course of building out these graphs, but now that the graphs are complete I find myself fairly flummoxed. Saving seems to always be wildly insufficient to fund investment (and no, lending/borrowing + saving has no relationship either). S=I seems to provide exactly zero illumination here.

And the post-1990 lending/borrowing swings I see don’t fit with any real-sector saving/dissaving story I’ve heard (or can remember). We see borrowing spike during the internet boom, dive following the bust, then spike again during the real-estate bust.  ?

So I’m going to leave this open to my gentle readers for the moment. What in the heck is going on here? What story (or stories) would you tell to explain what you see?

If anyone wants to see earlier periods zoomed in to get a better feel what’s going on, let me know. I’m thinking 1946-1975 (to see what seems like a period of consistency), and 1970-1990 (from the fall of Bretton Woods to the start of the internet bubble and the Clinton surplus).

Cross-posted at Angry Bear.