Even if each individual agent isn't rational it is possible that, on average, we are.
Think of a room full of gas that you slowly heat. All sorts of quantum effects and chaos affect individual particles. But, on average, we can predict the net outcome with fairly good results using very naive models (laws of thermodynamics, ideal gas law, etc).
And you're absolutely right about profit. Standard economic theory predicts that, in a free market that is at a stable equilibrium, each firm won't earn any profits exceeding the marginal cost of production (economists say that a firm is earning "Normal profits"). But, obviously, no markets (that I can think of) are perfectly free or in perfect equilibrium (and economists know this, and have far more complex models for real world situations).
Although, if memory serves, this hypothesis has been empirically verified by many almost-free markets (i.e., commoditized goods sold to intelligent/large consumers). You have a bunch of sellers doing business, barely making ends meet, until something happens that disrupts equilibrium (new technology, another company's bankruptcy, sudden changes in demand, etc), where they can hope to earn supernormal profits in the short term, before the market returns to equilibrium.
When each individual agent is not rational the group is not rational. Take car buying, a major car company that cut it's advertising budget in half they could try to sell a higher quality product for less money. At which point all other company's would have to limit their budget to compete etc. However, this does not happen because consumers are not rational as a group.
You can even measure the level of rational behavior by comparing stable markets. EX: Gasoline vs Bottled water, there is vary little to distinguish the products, but brand name water carries a huge price premium.
Think of a room full of gas that you slowly heat. All sorts of quantum effects and chaos affect individual particles. But, on average, we can predict the net outcome with fairly good results using very naive models (laws of thermodynamics, ideal gas law, etc).
And you're absolutely right about profit. Standard economic theory predicts that, in a free market that is at a stable equilibrium, each firm won't earn any profits exceeding the marginal cost of production (economists say that a firm is earning "Normal profits"). But, obviously, no markets (that I can think of) are perfectly free or in perfect equilibrium (and economists know this, and have far more complex models for real world situations).
Although, if memory serves, this hypothesis has been empirically verified by many almost-free markets (i.e., commoditized goods sold to intelligent/large consumers). You have a bunch of sellers doing business, barely making ends meet, until something happens that disrupts equilibrium (new technology, another company's bankruptcy, sudden changes in demand, etc), where they can hope to earn supernormal profits in the short term, before the market returns to equilibrium.