Supply and Demand
Way back in my sixth form days I took an economics class. It was there that I was formally introduced to supply and demand for the first time. I had heard of it before, and had a vague idea what it was about, but now I could say I had properly studied the theory and the various scenarios one could potentially model.
It's a very simple model. You start with a graph in the 2D plane. Draw a line with negative gradient representing 'demand' and another line with positive gradient representing 'supply', and the point these lines intersect determine the price you're interested in. You could even shift the lines around to represent supply and demand changing, and change their gradient to represent whether a market was 'flexible' or 'inflexible'.
I had a number of issues with this model right from the start. Which I am going to list right now =P
1) This one is a minor semantic point, but it really bugged me at the time. The economics theory had a concept analogous to gradient but defined it the opposite way round to my maths class ie. Change in X over Change in Y. This meant that a line getting flatter was increasing in gradient and aaaarghlfflbfrdfl;g
2) Now, for a more serious point, what were the axes supposed to represent? Consensus agreed the horizontal axis was 'price', but the vertical axis was more controversial. It could maybe represent 'quantity' or 'market willingness' or maybe it was just abstract units because really, what units could you use to measure 'demand' anyway?
3) The concept of 'shifting' (or translating) a line was, in my opinion, introduced entirely the wrong way round. The idea was to represent an increase in demand by shifting the demand line to the right ie. an increase along the price axis. What is that even supposed to mean? We all collectively decide that a higher price would be better?
On later reflection I realised you could simply think of it as a shift upwards rather than a shift to the right. It would be the same mathematically, but conceptually make much more sense. That is, the total number of people willing to buy an item at a given price increases. Higher demand. Simple.
3.5) Speaking of which, who says supply and demand have to be straight lines anyway? Why couldn't there be a curve or jaggedy line? It seems to me that literally any shapes would work just as well so long as they intersected in the right place =P (tease)
Now for my three biggest problems with the model:
4) It was simply an abstraction taken to extremes. The people who actually set prices, shop managers and the like, don't have a convenient graph with lines drawn all over it to determine their price. They have to consider various factors like how much it cost to make, how much profit they might make, and what prices others are charging for similar products.
In fact, the concept of profits or revenue seemed totally absent from the whole model. Yes a higher price might turn some customers away, but I would be prepared to accept that if my overall revenue was greater as a result. How does that fit onto my graph?
5) From a scientific perspective, the whole concept seemed grossly unmeasurable. Even if you could pin down the dimensions of the vertical axis, it's just not possible to measure our abstract concepts of supply and demand in the real world, especially when any real price that a transaction takes place with counts as an intersection for the purposes of the model. What about all the space on the graph that is not the intersection? It's conceptually impossible to get real data on those regions.
6) Again from a scientific perspective, the model ultimately seemed to have little to no predictive power. You can observe a rising price and say, aha, demand must have risen. Or, erm, maybe it was supply that shrunk. Or maybe the retail outlet played a clever psychological trick, repackaging/remarketing a product to convince people that it's better than before, so even though the same number of people are buying the product and the same amount is being produced, they're prepared to accept a higher price. Maybe we can just lump that under an increased demand?
Ultimately the model seemed to say little more than this:
Price is positively correlated with demand, negatively correlated with supply.
Of course, all of the above was just the ravings of my sixteen year old self, in response to a limited economics teacher =3 I have since learned that model does have some merit and some of my objections had some reasonable responses that I shan't be going into here =P
I've been inspired to write about this today by the online game I've been playing recently, EVE Online. It has a genuinely player based market ie. if you're trying to make a profit off trading, you're going to have to outsmart human competitors, much like in real life.
However, it has two major advantages over real life for the purposes of an economics case study =3
1) The items in question are fundamentally identical. There's no room for clever sleights of hand convincing you that my loaf of bread is better than his loaf of bread, they function the same and can be compared without caveat.
2) The players are given perfect information over the state of the market. All of the people who are willing to buy, or sell, in what quantities in which region and at what price are listed in a neatly formatted database =3 So there's no opportunity for a big company to get away with selling expensive pizza when a guy down the road is selling the same thing for half the price.
With this system in place, it's actually possible to observe how supply and demand impact price directly and immediately. Allow me to explain.
First, define 'demand' as the total number of people who are willing to by an item at a given price. Except that's not quite right, as they should be weighted by how much each person is willing to buy. So maybe call it the total volume of goods that the playerbase as a whole would purchase (over an arbitrary unit of time) at a given price. Similarly define supply as the total volume of good that people would be prepared to sell.
So imagine you've come across some item in the EVE universe that you want to sell. That is, you intend to increase the supply. What do you do? You have two options.
1) Find someone with a 'buy order' and fulfil that order. Obviously, you want to sell it to the highest bidder, because to do otherwise would be burning money (remember that you have equal access to every potential buyer). But in fulfilling this buy order, it is complete and no longer present on the market. So the second highest bidder is now the highest bidder, and the price has fallen.
2) If none of the buyers are offering a price you find satisfactory, you can set up a 'sell order' instead. Here you can set any price you like and wait for someone who is interested. But what price to set? If you're like me, you'll like at the current cheapest sell order and price your item a little lower, so as to be the most attractive and the quickest to sell. Again, the price has fallen.
There is a caveat here. You might instead choose to price your item at higher than the cheapest sell order. What happens then?
Well, assuming the players buying the product aren't intent on burning their monru, they will ignore your sell order in favour of the cheaper ones. After all, they are identical in quality and equally accessible.
This is not to say your order is worthless. If you wait long enough, chance might have it that all of the cheaper sell orders have been taken and now yours is the cheapest. That is, the price has risen up to the point that your sell order has become relevant.
So ultimately, making a sell order at higher than the current cheapest price (or making a buy order at less than the current highest bid), is a declaration to the player base. It says "I refuse to participate in this market until the price has reached a point I am comfortable with."
For the purposes of making graphs of supply and demand, these orders are amazing. They don't involve the current market price but instead tell you how much supply/demand there would be at a hypothetical alternative price. If you define a sell order as "I would be happy to sell at this price or higher" and a buy order as "I would be happy to buy at this price or cheaper" you can construct actual, data driven graphs of supply and demand that actually mean something holy shit. They might not be straight lines, but they would at least be monotonic.
There are one or two small problems with this idea, but I'm too tired to go into them. Oh well =P