Take for example how supply and demand influence price. I.e. more supply means reduced price, more demand means increased price. And 90% of the time, that hold's true, but it's that 10% where it throws the models off, and the cause could be something fickle like masses of people arbitrarily decided that the product has gone out of style and they don't want it anymore no matter what price it is sold at.
I've written theories largely based on cost, and handwaved price as a market economics topic. I believe that's a valid stance.
In my economic theories, the basis of productivity improvement is labor time reduction: if you need 10,000 man-hours to produce food for 10,000 people, each one person must work 10 hours to eat. As Adam Smith observed, you can compartmentalize this: 2,500 people can work 40 hours to feed everyone, and the other 7,500 can do something else. Adam Smith's observation was flawed in that he claimed division of labor was the only way to do this--that you had to add new people handling smaller parts of the task--and thus claimed you couldn't have the *same* people or the *same* number of roles invested in doing different tasks requiring less time and producing the same output. For example: he discounted that a power tool maker could design a better power tool, and discounted that something like cellular manufacture would have any gains (cellular manufacture is a rearranged assembly line to reduce the time spent carrying intermediate products around).
That productivity improvement implies a lot of things. Your theory of "Supply and Demand" has implications such as something called "Scarcity", which I can explain. Scarcity occurs with superlinear growth of labor requirements.
Let me demonstrate.
It takes 2,500 people to feed 10,000 people. It takes 5,000 people to feed 20,000 people. It takes 10,000 people to feed 30,000 people. It takes 30,000 people to feed 40,000 people. It takes 60,000 people to feed 50,000 people.
Somewhere between a population of 20,000 and 30,000, it started taking more people--more labor-hours--to produce additional food for one person. That means you can feed up to 20,000 people with 10 hours of labor invested per person; but when you get to 30,000 people, you're averaging 13 hours of labor per person--which means those last bits of food are averaging a lot more. If it's the last 10,000 people requiring the scaled-up effort, then you're paying 10 hours per person for the first 20,000 and 20 hours per person for the last 10,000.
Eventually, you need more labor than you have available: making things is just impossible.
Scarcity starts when it starts taking more labor per unit output to produce an increased output of goods.
My theories of wealth growth stand not on labor hours, but on labor costs. Labor costs are labor-hours multiplied by labor price. The primary method for reducing labor costs is to reduce labor hours; I recognize that increasing labor price has serious economic effects, and that decreasing labor hours both decreases productive scarcity and decreases labor costs as two separate economic factors. In other words: lowering the labor requirements to produce a good produce one set of effects by the same mechanism as reducing wages, and another set of effects stemming from the addition of available workforce labor. It's self-referential in that second bit: think of it as "like cutting wages plus other stuff you don't get just by cutting wages".
Prices can go as low as costs, sustainably; they can't go any lower in the long run. If it costs $550/tonne to produce rice, you can't sell rice for less than $550/tonne for very long. You can sell it for $1000/tonne if no other market factors drive the price down, of course.
A lot of market factors drive price toward cost. There's direct competition (ten rice suppliers; better push rice down. Oh, we can make it for $180/tonne now, so let's undercut that $550/tonne price and sell it for $200/tonne), which is very fast; and there's inflation pressure (prices climb slower than inflation), which is very slow. (The buying power of unit currency is the total income divided by the total production--the total production being the total buying power--which is how we get inflation.) To your point, however:
the cause could be something fickle like masses of people arbitrarily decided that the product has gone out of style and they don't want it anymore no matter what price it is sold at.
There's a market behavior of indirect competition. People are no longer interested in overpriced fancy shoes; tech is in-vogue, and they want smart phones and tablets. Because they can't afford both tablets and fancy shoes, they stop buying fancy shoes. To compensate, the fancy shoe producers reduce their prices closer to cost, within the affordability of the consumer base after buying smart phones and tablets. If the fancy shoes cost more to make than that affordability, they vanish--or get replaced by poorly-made imitations; otherwise they come down sharply in price, because shoes are competing with iPhones now.
That explain your observation of an apparent link between supply and demand? It's a valid observation; it's built firmly on more fundamental economics that nobody has yet theorized, and those demonstrate its validity and explain its quirks.
I'm working on explaining all this, but it takes some time and I'm lazy. I've also been blogging some stuff lately--we'll see how long that lasts--to get some notes down that I'll later feed into the paper. Much of the unwritten theory is already in my head, and I'm using a lot of it to extrapolate further theories and observations.