This is mostly a response to James Gardener's posts, so you might want to read them first. This was written with a high degree of ignorance about actual bank cost structures and revenue streams, so criticism is both welcome and deserved

Whatever the shape of *banking*, as a monetary function it will need to exist and it will be a core value-adding activity, regardless of who does it. If nothing else the government needs a formal money supply to collect taxes in. But it is perfectly possible for the big box *banks* to fall to the edges. Over the last three years I've made more visits to corner shops to pay bills and take out money than my bank's branches. It might be a massive change to the financial system, but it would be a fairly modest shift in most consumer's behaviour.

Where banks will be, depends partially on the regulation of banking as monetary function, which is the governments decision, partly on how banks deal with the end of the Product Farm business model, and partially on the acceptability of "unofficial" credit channels.

Banking is a monetary function. It's not an institution, it's the activity of extending credit against collateral as new "offical" money. Which raises two questions: What system would regulate the creation of that new money? And, to what extent do we now need that new money to be "official" legal tender?

And I think the answers are: government will try to have whatever system pleases it, the use of bank credit might drop sharply, and the Product Farm model is transitory and will die.

The primary long-term concern of governments in banking should be a stable tight-money system to avoid free-riding undercapitalised banks. In practise, it's to support favoured sectors (estate agents in the UK, plastic toys in China, cows in France, etc) with soft loans. Japan has proved that the short-term option can last for a very long time. But I don't believe the extent of capital and human mobility in Britain would allow that. The resources to maintain Chinese-style policy banks simply won't be there. Not without killing non-bank financials with regulation with one hand, and imposing capital controls with the other.

When the government works itself around to a sane position [more competition, higher capital requirements, accurate accounting, numeracy test, etc.], there are going to be two important questions: Could banks continue to lend on-demand deposits as committed long-term loans? And, how would depository institutions be allowed to place client money with one another?

If banks can lend out current and savings accounts for long periods, banks can be a substantial source of new capital to business. (Or government...) The interest from these loans would allow high rates for savers, and as banks are part of the state (see Martin Wolf of the FT) those profits are safe. (The German model.) If they can't, then banking would be restricted to trade credit and overdrafts. It would be too dangerous to let out money that might be needed tomorrow to finance a 10 year project. Fresh investment capital would have to be raised in the stock and bond markets. Those markets will have to expand to allow economic growth.

The extent to which banks can place deposits with other institutions, is the extent to which business lending and retail activities can be separated. If a new bank can simply pass customer deposits into the money markets, it doesn't need an army of loan officers. The army needs to exist, but not on their payroll. Conversely an investment bank can fund loans that are "bank grade" with retail deposits, even if their investments are usually speculative. Disintermediation through the money markets has already created pure savings banks like ING; it could allow money service firms like Moneybookers to become banks without direct lending arms. (Long-term it might be sensible to disperse bank funding activities into big "normal" companies. A small bank might fit nicely on the side of Unilever for example.) But money markets and inter-anything are not fashionable post-Lehmans.

If the government chooses to force a return to "traditional" banking, new institutions would need to create *balanced* lending and depository arms. The lending arm would be much harder to scale and would determine the rate of growth. Lending requires relationships and specialist knowledge: making judgements about future cash flows is fundamentally harder than servicing a consumer.

And from this we can make matrix (w00t).

In any of these four scenarios, innovation in and outside of banks will continue, perhaps faster at the top left than the bottom right, but it will happen. Secular shifts in technology determine the cheapest way to send a statement or to clear funds. These will be adopted to increase revenue regardless of any policy environment.

So, can the Product Farm model of big box banking survive? The Product Farm being a horizonless call centre outside Milton Keynes, where task workers sip 65p instant cappochinos [it's a number 52 in most Milton Park offices IIRC] from regularly spaced vending machines, meeting daily re-mortage sales quota. There's some instinctive organisational actions around deposit taking and payments, but it isn't a high margin activity. (The guy responsible for FastPay sits in the corner looking alienated, no one speaks to him.) The Product Farm is going to die. The people working there have better things to with their lives; I have better things to do than talk to them. And packaged personal loans & payment protection insurance are dead now.

More importantly, financial arrangements are not products. You sell a product, take one payment and ship it. All the money is upfront, all the cost occurs before the sale, and physics makes mass customisation impractical. In most financial deals, money can flow both ways, most of the work occurs after the point-of-sale, and customisation means altering a few numbers. A family going to Peru once a year, do not carry the same risks as a single male heading off skiing for 6 months. They really should have individual travel insurance contacts, and it's not too difficult to give them that. Any provider who wants to capture these efficiency will end up passively exposing their pricing, and asking the customer to pick the "cut" that fits them. Not doing this will mean leaving money on the table, both for the provider and the customer.

(Myself and Antony Evans had a long muse about the "Platform Bank" as a potential end point of all this, which you can read here.)

This certainly wouldn't mean the end of sales and marketing, but when the customer comes to you, it becomes more a case of informing people about what's there. Traditional sales forces aren't going to be a big factor. I am not totally convinced that without their use as sales channels, most banks would maintain a serious high street presence. I can pay my gas bill at PayPoint, get cashback at most shops, Western Union has more agents on my street than banks have branches. So it's provenly unnecessary to have your own high street frontage to service customers.

Finally, there's the question of how much money needs to be legal tender and run through the "official" banking system. Ignoring baby sitter vouchers and LETS schemes as utopian; lots of potential alternative money systems have been put in to use. Switzerland has the WIR payments network of mutual credit between SMBs; many parts of Russia developed de-facto local currencies (often around electricity and local taxes); Argentina ran parallel local government currencies for a while; Ireland survived a bank strike for years by just endlessly reassigning paper checks. Yes, all these "anomalies" happened during crises, but they weren't created specifically because something bad happened. They were created because of sudden shift in the relative price of using bank money or creating your own. That shift can come through changes in technology. The dominance of the government within the Western economies might mean its money would always dominate, but even in the UK that still leaves, maybe 25-45%, of the money supply 'in play'. (For an explanation of why multiple currencies are desirable - read Hayek.)

So in summary: Banks will be reshaped by whatever monetary arrangement the government settles on and this will determine their structure going forward. Those changes create intense competition from new entrant, or almost none. In any case financial service providers will need to start offering individual arrangements, not selling pre-packed "products". And there's a giant wildcard out there of what happens when anyone can run a monetary institution from their Nokia...

----
If you liked this rant, subscribe to my RSS feed, and next month I'll be talking about Twitter (like a proper blogger).