(a) Lending - creating interest bearing loans;

(b) Spending - to buy (say) government debt; to pay staff or other costs; and to pay dividends to shareholders.

In every case bank credit creation gives rise to a matching demand deposit, and the sum of these deposits is - with notes and coin (and now QE) - the 'fiat' money in existence. There is of course plenty of other (trade etc) credit in existence.

There is no reason at all - other than pure ideology - why Treasuries acting directly, or indirectly through Central Banks, cannot spend, as well as lend, money directly into the provision of productive assets in the public or private sectors.

This spending process would need to be managed by a service provider with a stake in the outcome, and would also need to be accountably supervised by a monetary authority.

Once productive assets are complete, then the QE/Public credit used to create them could be refinanced by existing or new long term (eg pension) investment, and the QE would be retired for recycling.

Investment in the individuals and enterprises necessary to create these assets would be taxed, and part of this tax would again retire and recycle the QE investment.

The 'Big Lie' is that public credit/QE is 'inflationary' when private credit is not. In fact both are potentially inflationary, particularly if applied to existing productive assets, but private credit is self evidently more inflationary than public credit to the extent that it includes excess management etc payments and dividends to shareholders.

Neither public nor private credit has any cost at the time of creation. Both come with a cost of service/platform provision and default costs.

"The future is already here -- it's just not very evenly distributed" William Gibson