Part of the problem is just that the systems involved are so complex, of course -- the economy has complexified and the human brain has not kept up.
Another part of the problem, though, is that none of the traditional economic theories (neither the mathematical ones or the qualitative ones) embrace this complexity in principle.
I.e., the very fact of the increased complexity of the economy has direct implications, which seem to require introduction of new approaches for economic analysis.
One way to look at this is in terms of what I semi-seriously call "Nietzschean economics," where a more abstract notion of economic power replaces the traditional notion of money.
Put simply: it seems that the modern economy has somewhat obsoleted the concept of money, which has economists (and many others) confused...
For instance, China has a lot of US dollars right now, in principle ... but they can't really spend most of it, because if they traded their dollars for goods on the open international market, then the dollar would collapse in the international currency markets, the US economy would tank, demand for Chinese goods would collapse, and China would risk massive internal unrest etc.
It's not exactly that the Chinese government has money they can't use; but they have money that comes with severe restrictions on the ways it can be used. Having this money confers great economic power on them; but, this economic power can't necessarily be used in the traditional manner of buying stuff.
In general: with the world so interconnected, the notion of a unit of currency as being something that can be exchanged for a certain amount of stuff, anywhere, doesn't really apply...
More complexly, the same sorta phenomenon applies among and within investment banks and other financial institutions. E.g. a large bank holds some portfolio of financial instruments ... but if they sell a lot of one kind of instrument, this impacts the markets in a way that affects the value of the others, etc. So one can hardly assign each instrument they hold a value independently of the others.... The values of their various portfolio items are interdependent in the manner of a system of simultaneous nonlinear equations.
What this growing, rampant interdependence means is that the traditional concepts and tools of economics don't closely apply to the international biz/finance world anymore...
A lot of this was foreseen, on a qualitative level, in Galbraith's book "Economics and the Public Purpose" from the 1970s ... he talked about the "market economy" versus the "technostructure" and pointed out that the latter (being a complex of large corporations, governments and other institutions) follows quite different rules.
Well, now the nonlinear dynamics of the technostructure is running rampant -- but economists are still mainly studying it with tools designed for studying markets.
One step I think that needs to be taken, on the theory level, is to view actual buying power as Level 1 of a hierarchy of types of economic power. I'm thinking of:
- Level 1 = power to buy goods or services [right now, or at some future point(s) in time]
- Level 2 = power to influence others' Level 1 power
- Level 3 = power to influence others' Level 2 power
- .. etc. ...
The ultimate extreme of all this of course would be
- Level infinity = power to influence others' Level infinity power
which was basically Nietzsche's view of the driving force of the universe ... and I do think that international economics and politics boils down to this sometimes: power for it's own sake, rather than being tied to ultimately influencing the acquisition of goods and services.
Traditional economics is based on the notion that everything boils down to buying power ... but in a world where no one is smart enough to calculate what their decisions will imply in terms of buying power, this sort of economics seems to have limited applicability...
What we need is a nice, elegant, pragmatically applicable theory of the nonlinear dynamics of Level k economic power under conditions where
- the computational complexity of recognizing important high-level patterns in an economy
vastly exceeds
- the computational capability of even the smartest individual participants in the economy
This might be a fun thing to work on, but I've got a thinking machine to build, so hopefully somebody else will do it ... or we'll have to wait for the AI to solve the problem. (Of course, an appropriately constructed AI could also palliate the problem, due to having increased capability to recognize economic patterns, either due to possessing greater-than-human general intelligence, or due to combining human-level general intelligence with specialized capabilities for economic analysis.)
4 comments:
So we started out as hunter-gatherers.
Then we invented barter.
As the need for a medium of exchange for trade increased, we invented money in about 1350.
Are we now at the end of the monetary system? Are we ready to move into a resource-based economy?
This means realizing we actually have the power to provide basic food, shelter, and medical care for every person on the earth, as a matter of resources and production.
This ability will only become greater in the future as we create more useful robots, and eventually thinking machines.
Is what we are experiencing now the birth pangs of this resource-based economy?
A less abstract example:
Level 1 = workers
Level 2 = employers
Level 3 = banks
Level 4 = US Treasury
Now the Treasury is trying to spend $789 billion in such a way as to get the banks to fund companies that will employ people. They have some ideas about how to do this, but certainly not well quantified. There are no formulas that tells us how much to spend to generate so much employment.
Someone in a private email response, asked me what my expertise in finance is; and also posited the view that the key to the current financial crisis is to eliminate derivatives.
Here was my email response to them.
My background regarding finance is deep but spotty ... I studied some mathematical finance in grad school (stochastic differential equations based models, etc.) and I much later spent a few years directly involved with quantitative finance, especially the creation of directional trading systems for currency and commodity futures and equities. I never got rich from any of this though ... so I can't rate myself a real expert ... nor a financially lucky person ;-)
Regarding your reply: I think it is untrue that the "efficient market hypothesis" held until ~2000.
In fact, I think, the markets were very inefficient before then. My experience is that fairly simple pattern recognition algorithms can succeed in predicting pre-1980 markets, for example. The markets back then were more easily predictable than more recent markets, because the use of sophisticated automated pattern recognition algorithms had not yet been "priced into the market."
Regarding your proposal to eliminate complex derivative financial instruments ... I agree that this would simplify the economy, but it would also eliminate a lot of efficiency and slow down technological progress. These complex derivatives exist in order to allow the hedging of risk in various ways, and this indirectly allows society to take MORE risks on investing in advanced technologies of various sorts. Without derivatives, we would have more stability, but less risk-taking and hence IMO less progress.
I think the US's system of regulating derivatives got fucked-up in a complex way, due to a combination of stupid business practices and stupid government regulations; but I disagree that eliminating derivatives is the answer ... **unless** one's goal is to slow down progress and revert closer to a steady-state society.
I actually do think that AGIs could manage derivatives way more intelligently than humans. We are not evolutionarily tuned for processing financial data, but rather for processing visual and acoustic data, right? We try desperately to grok financial data by projecting it into visualizations, but these are low-dimensional projections that leave out most of the information. An AGI, even if it were inferior to humans in many respects, could almost surely way outdo humans in this domain.
The only reason we *don't* have such an AGI yet, I believe, is that the folks who hold the finance-industry purse-strings in the US are obsessed with pursuit of short-term gains, and haven't wanted to make a (say) 5 year research investment in the development of AGI's to comprehensively analyze global market data. But this will happen eventually.
Quite peculiar, eh? Once the dynamics of the system are figured out, a new level of complexity is injected into the mix. Sometimes I believe that the only guaranteed way to understand and manipulate a global system is to become the most powerful agent in that global space. I don't see that happening, though =)
Post a Comment