So the bank is out of the bag. Today (24 September) at the Lib Dem party conference, Vince Cable announced the establishment of a state bank to get credit flowing to businesses. But as the last five years have taught us, there are all sorts of banks: good banks, bad banks, casino banks and zombie banks. Will this new bank be the right sort of bank, and will it help business to grow?
There was much to like about the speech. Vince's vision of an industrial strategy based on "long-term investment in innovation, skills and science" will be familiar to anyone who's read Nesta's PIan I for innovation-led economic growth.
When it comes to banking, the Secretary of State's heart is in the right place. He believes that if we want banks that truly support the UK's businesses, we need to split 'utility' from 'casino' banking, a policy we floated with John Kay in 2009 in Narrow Banking. He noted the banking sector can be anti-small-business (at Nesta we'd add 'anti-growth business' - the 2010 report Vital Growth suggests banks' lending models discriminate against high-growth firms). And of course, he launched a government-backed business bank, something argued for in Plan I.
But when it comes to state banks, the devil is in the detail. Good state banks, like Germany's Kreditanstalt für Wiederaufbau or Brazil's BNDES, provide a basis for business growth and economic dynamism. Bad ones... well, we know how that story ends.
Today's speech gives some clues about what sort of organisation the new bank will be - but they're mixed.
On the one hand, it sounds like we are looking at something relatively hands-on: an actual bank rather than an extension of the funding for lending scheme. On the other hand, it's not clear whether the bank would make its own loans or simply lend money via the existing banking system.
None of these are bad choices.
But there's the danger that we end up with something neither fish nor fowl. One clear option is to set up an expanded government guarantee programme. In our 2011 report Beyond the Banks, we showed how the government could guarantee portfolios of bank loans, allowing the Bank of England to buy them from existing UK banks, freeing up their balance sheet to lend more. This is a straightforward answer to the problem of banks' tightening balance sheets.
If we also want to tackle how banks lend, then we need to do more than just provide guarantees. Plan I sets out a proposal for a true state bank that has the power to lend directly. As the British Chambers of Commerce argued in an excellent report, if the existing banking system denies credit to creditworthy firms, we can't simply pour more money into a broken system. Korea and Brazil provide good examples of active state development banks that provide patient capital to help businesses grow.
Finally, amid all the talk of banks, it's important not to forget the really disruptive innovators in the world of finance. Platforms like MarketInvoice and Crowdcube, Funding Circle and Angellist are toiling away to transform the way businesses get finance. I don't know which if any of them will succeed, but it's essential that we give them the space they need to try. This means not crushing them beneath inappropriate regulation and not crowding them out with other schemes.
So, Vince's speech holds out some hope for innovative businesses looking for finance - but how it's implemented will make all the difference.
Follow Stian Westlake on Twitter: www.twitter.com/@stianwestlake