**Previous:** On printing money

Price is the amount of money one pays for something. However, economists have a more elaborate concept of price: It is the amount of money people willingly pay for something when demand and supply are in balance. In this article I am going to talk about the latter concept, about what one would call "a fair price".

The interesting thing about fair price is that it almost never occurs in the real world. Your best chance to spot it is to peek into an economics textbook.

There are many reasons why fair prices are mostly confined to mythological grounds but today I would like to focus on one that is rather interesting from the theoretical perspective.

Imagine a market for some kind of goods. Say cars.

After reading an economics 101 textbook one would expect the price of a car to stabilise at a point where demand for cars and supply of cars match.

If you look at the real world though you'll discover that there are Tata cars for $1500 and Rolls Royces and Ferraris for half a million or more. That's a difference of more than two orders of magnitude. How come?

The answer is pretty obvious: Tata and Ferrari are not the same type of commodity. We should think of it as of two different markets. A market for cheap cars and a market for luxury cars.

But when you look at the market for luxury cars, the spread is still rather big. Say one order of magnitude. So there are really two markets for luxury cars, one for cheap luxury cars and one for expensive luxury cars, n'est-ce pas? But then, why not divide the range over and over until you get a category for itself for each type of car?

But that's not the end of it. There's a price difference between new Tata Nano and used Tata Nano. Let me see: There's probably a different market for used cars. But then a price for 5 year old used car may be different form price for 20 year old used car. And maybe pink cars come at different price than black cars? And what about cars with a scratch on the front left fender?

The point is that things in the real world come in different flavours and with different nuances. They can be classified into infinite number of categories.

I've chosen cars as they are approximately in the middle of the "commodity level" scale.

On the high end of the scale there's stuff that's quite commoditised, like flour or sand. On the low end of the scale come stuff like houses. Each house is so different with respect to its location, surroundings, size, style and so on, that one must wonder how can the market in houses exist at all. At the absolute low of the scale is an unique item, such as Rembrandt's Night Watch. At that point it's silly to even talk about a "market for the Night Watch".

The observation from all the above is that you can either think in broad categories, like cars or houses, but then the prices will differ so widely that even the idea that there's a fair price for 1 house and that you can get 10 of them for 10x that much is pretty ludicrous.

Or you can look at increasingly finer categories like "old 2 bedroom houses in south-east Sussex with a fireplace and a Wisteria climbing the front wall" where the amount of things matching the description becomes so small that a "fair price" — which is really a statistical concept — is hardly applicable.

All in all, it seems there's some kind of uncertainty principle, vaguely resembling the uncertainty principle in quantum physics, at play here: The closer you look, the more is the price distorted by small sample size. The wider the picture the more is the price distorted by varying nature of the priced good. Generally, it looks like the amount of distortion stays the same whatever you do.

But wait! There are money and there's bullion! There's a lot of one dollar bills and they are all exactly the same, interchangeable. Also, bullion is a bullion is a bullion. It doesn't matter whether you own this gold bar or that one. They are all the same.

As one can observe, the uncertainty principle can be circumvented by defining a virtual good, a purely abstract measure of value.

However, given that such virtual goods are good for nothing by themselves, one has to exchange them for an actual good at some point and that's where the uncertainty principle kicks in again.

**Martin SÃºstrik, March 23th, 2016**

**Previous:** On printing money

Economists ought to stop emphasizing supply and demand curves on their basic teaching. The problem is that those curves do not exist, they are model artifacts, they are unobservable and much more complex than a basic course can convey.

Anyway, you are missing product substitution. Unified markets do not really exist, what exists is a set of people with their own preferences, that map better or worse on the products available for sale. And distribution is almost never optimized.

ReplyOptionsHow does product substitution solve the fuzziness problem?

As I see it, it's just restating the problem in different terms.

Imagine a multi-dimensional space of all possible substitutes. When pricing a particular good one can look only at its close vicinity (i.e. good substitutes, differing only slightly from the target good). In that case there's the small sample problem. The estimated price is quite random.

On the other hand, one can look at larger vicinity (less good substitutes, differing from the priced good in significant ways) in which case the prices will vary simply because they refer to disparate stuff.

I get that economic theory imagines world with infinite number of samples to base the pricing decision on and smooth price gradient throughout the substitute space, but that's not really the case in the real world.

ReplyOptions## Post preview:

Close preview