fresh_42 said:
It is a bit more complex than this. E.g. at stock markets around the world, we have a sufficient number of arbitrageurs and yet, the prices aren't equal. It's better, but still not perfect. I assume, that in reality the transaction costs are also affected by ordinary taxes, costs for participation in the markets, liquidity and probably some more constraints or even soft conditions like customers' prejudices.
There's an issue here of what most people are using the term arbitrage to mean, and the stricter sense of it's meaning.
Law of one price should be a very close approximation of stock markets, if you are in fact getting the same 'good' in different markets and there are
actual arbitrage mechanisms available.
The issue is that I think you are in effect pointing out is that these goods have different properties in different markets. Liquidity is probably the most subtle one of these -- so nice catch there.
However, if you can effect an exchange between markets (i.e. actual arbitrage) then relative liquidity does not matter -- at a high level.. at a lower level there's a lot of nits on how much product can be bought or sold without moving prices considerably which is a liquidity concern.) Customer prejudices don't matter either in this case.
Something the press messes up a lot is substitution is another, weaker, way to get prices down but in any pure sense of the term: substitution ain't arbitrage. (There are generalizations of the term arbitrage like 'statistical arbitrage' but that is a very impure term and not at all what I'm referring to here. Things like Merger Arbitrage are a more pure extension but it is largely about risk pricing and ultimately a distraction.)
For example: the fact that there is an exchange -- i.e. arbitrage-- mechanism is why etfs (in liquid markets) closely track underlying prices and you really do see the law of one price in action. Going long an underpriced instrument and short an overpriced 'copy' of said instrument is not in any pure sense arbitrage. If you can make an exchange and net the positions, then it is.
I of course agree that ordinary taxes and costs for participation, but they really get lumped under 'transaction costs' or 'frictional costs' which is the umbrella that covers shipping costs as well... There's always a question on general of language to use.
To fair, "arbitrageurs" tends to mean substitutors these days, and it's my strict usage that may be out of date. The strict usage is the right way to interpret law of one price though if we want stronger claims. And if we want weaker results, then interpret the 'law' with weaker mechanisms like substitution, as acceptable. Seems reasonable enough to me.
- - - -
I'm not sure what "ASS generics" means and I am a bit concerned to search. (Maybe Over The Counter, OTC, generics?) Generics have branded competition though, which breaks the abstraction typically used in law of one price.
fresh_42 said:
I'm not arguing against patents, because I know about the costs of development, approval, and the need to protect research.
A couple years ago
The Economist actually ran a scathing piece that argued for, in essence, getting rid patents. I can post a link. I still am ambivalent on the article and mulling it over.
- - - -
edit:
full disclosure: the term used in econ is "law of one price" so I used it, though I cringe at it a bit.
The "approximation of one price" or "tendency toward one price" would sit better with me, but I guess they don't sound as nice.