A fun concept

 The expected utility of an agent's risky decision is the mathematical expectation of his utility from different outcomes given their probabilities. 

Take a closed system, the pedestrian district nearby.  Three theaters, five night clubs, four restaurants, and two coffee houses. There is also on large supermarket. Which sector has the highest utility at equilibrium for a closed patronage? The sector with the most floor space.

This is not Kelly betting.  All of these sectors draw from the same customers set and customers split time between them.   The Kelly betters in this case are the landlords, betting interest charges vs building capacity. The pedestrian sector is a market, unto itself. Like the revenue sharing, Congress does not make the Kelly bet, Treasury does.

Kelly bets are across independent sectors, and adaptive Kelly bets try to force independence in a portfolio rebalance.

The reality is a bit mixed. The pedestrian will often take a different route, grabbing breakfast on credit because the crowds were small.  Or sometime skip the store if it is packed.   The pedestrian is adjusting daily bets based on a innovations in expected payoffs. But normally, the pedestrian daily wad of cash gets through the day.