I've come up with a way of building cooperative networks of digital money transactions that I think has a lot of potential. It might be difficult to understand and believe the concept if you don't have a strong background in network dynamics; but I'll try to explain it simply.

Simplified explanation:
It's like an automated pay-it-forward system. Say, a user buys lunch and adds a voluntary 10% to 'help' the network – without expectation of return –, goes home and, by the end of the week or so, little by little, distributions of these additionals made from others within the network have recouped back 100% of the base transaction. Not exactly a free lunch, but one that initial user could have again. The network 'helps' back greatly as a compound effect to those that 'need' it.

Technical explanation:
The math may seem simple and, simply stuck at a given state, yet it hides the overall dynamics that can only be interpreted as a whole with lots of activity within. The way it works is that when a transaction is made with a voluntary fee, it goes to an auxiliary account (B') of the receiver. Transactions are registered reinforcing or weakening incoming and outgoing links between accounts (Li & Lo) and a 'metabalance' (V) is defined for each account. Weighted distributions of the auxiliary accounts weaken incoming links trying to match each account's balance up to its metabalance, emulating an extremely high yield rate, though bounded to a modified balance equation: B' + B = Li - Lo + V. At anytime the sum of all balances is equal to the sum of all metabalances.

Since both balance (in the base) and metabalance (in the additional) are 'transacted' in the same operation, there's the option to make transactions as both (B,V) forward, one forward and the other backwards, only balance forward or only metabalance forward. This enables the possiblility to define goals within the network, for example one, to try to equate metabalance across, by sending the metabalance of the transaction to the party with the least, this would prevent 'demand collapse by liquidity strain' of base consumers, a sort of dynamic basic income.  

A playlist on the mechanics of the model can be found here. A paper with these mechanics can be read here (the way it handles links and routings is optional, but recommended). And a mockApp showcasing how would a user see it (highly sped up) can be seen here.

Applications:
This could be setup as a 'spendings account' in contrast to a 'savings account'. It wouldn't have a certain periodic yield based on the amount held, but a 'gradual cashback' based on the amount of the additional on transactions made and proximity to commerce with higher network activity and spending.

It'd be great to see Neobank FinTechs emerge from this technology or as a new product within traditional banking. I'm in the rush myself of pitching from zeroes (lacking the skills and budget to get to an MVP/Demo) both to VCs and Innovation Centers, although I'm not particularly interested in leading such ventures. There's also the crypto-space possibility. I'm sure it could be implemented in a single SmartContract. Up for grabs!