Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Markets are Efficient if and Only if P = NP

papers.ssrn.com

Post reply on HN
Print view View original

51–60 of 86 posts

Re: Markets are Efficient if and Only if P = NP

#51

Post by aphyr »

CWuestefeld wrote: post #44
aphyr wrote: post #35

Earlier quoted context omitted.

Jerf, this is a surprisingly insightful and concise statement--one that I largely agree with. I'm curious, though, and want to ask CWuestefeld as well: If we accept that the free market can act in ways which are detrimental to the global objective function (without loss of generality, let's pick your favorite, e.g., "Everyone gets food", "Individuals have class mobility", "I can shoot and eat the homeless for sport"…

That's a fair question, but don't let it suck you into a "false choice" fallacy. There are plenty of real-world examples of this. In all cases, free trade will make all parties better off on the whole . However, the effect of doing so may be that individuals at the margin lose out, e.g., an unskilled factory worker may find that someone in China is doing what used to be his job. It's pretty well decided now (although…

Perhaps we misunderstood one another. When I said "without loss of generality", I meant that for almost any conceivable objective function, free markets may not maximize that function. For a social objective like "the weighted average of the self-actualization of all people in their individual ways", a market may be non-optimal. I hesitate to construct an example for fear of getting bogged down in one particular case, but Jerf's efficient pricing of slavery comes to mind. Another might be the well-being of nonparticipants; for instance, habitat preservation for wildlife.

I argue that the only function that a free market can optimize for is the market itself; all other consequences such as happiness, fulfillment, distribution of experiences etc., are side effects which arise from coupling the market to the world via the decisions of participants, regulations, etc. Only if you assume perfect coupling of the market to the global objective is the system morally optimal.

In all cases, free trade will make all parties better off on the whole

This, however, makes it sound like you are assuming perfect coupling in the absence of any regulation, or that your objective function is the execution of the free market itself, and not, say, the well-being of the individuals and societies participating in it. I must not understand you correctly.

Re: Markets are Efficient if and Only if P = NP

#52

Post by arethuza »

CWuestefeld wrote: post #50
arethuza wrote: post #47

Earlier quoted context omitted.

"I find the argument reasonable." Care to share a reference that backs up that argument? I'm not doubting that there may be an argument, just interested to read that point of view.

I think the parent post overstates it a bit. I wouldn't say that the Fed was a primary cause, although they were certainly complicit by keeping interest rates so low, artificially increasing the demand for mortgages. If you're willing to generalize the comment to say that the crisis was primarily due to the heavy hand of the federal government, then I offer these two links. They're both op-ed pieces, so shouldn't be…

Ok - I can see that those articles make some good points about the US housing market. However, I am not in the US, and the housing market in the UK suffered from very similar problems and, as far as I am aware, we never had any overt government support for mortgages so our mortage market looked rather more "free market" than the US one and we still ended up with similar problems. Northern Rock being the most obvious example:

http://en.wikipedia.org/wiki/Northern_Rock

Re: Markets are Efficient if and Only if P = NP

#53

Post by prodigal_erik »

CWuestefeld wrote: post #40
jbooth wrote: post #31

Earlier quoted context omitted.

Well, I certainly can get behind your definition more.. but a gigantic 10 year bubble sticking out to the side and then bursting doesn't seem efficient by that definition. You'd think the wizards of the market, with the billions of dollars in pay they're collectively bringing home, would've corrected for that sooner, no? This concept that in 2008 the market was responding to incentive misalignments created by governm…

This concept that in 2008 the market was responding to incentive misalignments created by government regulation.. what regulation? All we did was deregulate from 92 onwards. This is false. The Community Reinvestment Act, passed in '92 and periodically ratcheted upwards thereafter, forced banks to increase loans to minority groups. It issued guidelines on how lenders should evaluate borrowers for these purposes: Affir…

The Big Short tells us "In Bakersfield, California, a Mexican strawberry picker with an income of $14,000 and no English was lent every penny he needed to buy a house for $724,000." Nothing in the CRA ever demanded that obvious inability to pay be disregarded. Once Wall St. started believing bogus CDO ratings and it became possible for mortgage banks to unload 100% of the risk, they weren't being pressured into writing stupid loans, they were eager to stuff the channel by writing as many as they possibly could before the music stopped, and used every trick they could think of (teaser rates, negative amortization) to get each buyer out the door and through the first months of a loan they knew was doomed.

Re: Markets are Efficient if and Only if P = NP

#54

Post by CWuestefeld »

aphyr wrote: post #51
CWuestefeld wrote: post #44

Earlier quoted context omitted.

That's a fair question, but don't let it suck you into a "false choice" fallacy. There are plenty of real-world examples of this. In all cases, free trade will make all parties better off on the whole . However, the effect of doing so may be that individuals at the margin lose out, e.g., an unskilled factory worker may find that someone in China is doing what used to be his job. It's pretty well decided now (although…

Perhaps we misunderstood one another. When I said "without loss of generality", I meant that for almost any conceivable objective function, free markets may not maximize that function. For a social objective like "the weighted average of the self-actualization of all people in their individual ways", a market may be non-optimal. I hesitate to construct an example for fear of getting bogged down in one particular case…

Actually, I rather like your proposed objective. Except I'm not sure what you mean by "weighted average": what would be the weighting factor? I realize you said that you didn't want to get bogged down in the function itself, but I think it may be important.

If we can agree that capital is the means by which a person can achieve fulfillment (I don't just mean by "being rich", obviously. It may enable the person to donate to charity; or that person's labor, which which has a monetary value, may achieve some end; or to obtain medical care to cure an ailment; etc.), then I can see an answer.

In the free trade example, opening borders to trade expands the economies of both nations. That's pretty much the same as saying that more capital is available within both nations, which means that -- across the whole population -- people are better able to move up their utility function (i.e., achieve self-actualization). In other words, because money -- capital -- has the power to buy or enable other things, then our positions may be compatible.

Of course, some individuals may benefit more than others, and indeed, some may find themselves worse off. That's where some would claim that we as a society owe compensation to some, to pay for our own gains.

Looking back at what I just wrote, it seems the crux of the argument, then, is the degree to which you can believe that capital can provide the answer to whatever your objective function is.

I'm going to reject the example of a market in slavery. This is the very epitome of a non free market. The slaves themselves are being forced into the transaction, and since a free market, by definition, requires rational actors entering the transaction of their own free will (because they all expect to be better off in the end), it cannot be considered free.

The classic examples of market failure are "tragedy of the commons" problems due to externalized costs, and public goods that would suffer "free rider" problems, like national defense. The externalization problems are due to our legal framework for property rights; a system that fully allocates property rights (e.g., some person owns this river; some person owns the air) would fix this, as demonstrated by the Coase Theorem. The free rider problem is stickier. They do turn out to be much rarer than people believe (e.g., the classic example is the lighthouse, but historical research shows that in the past, these have been provided by markets), but they're not entirely fictional, and I don't have a good answer to the overall question at the moment.

Re: Markets are Efficient if and Only if P = NP

#55

Post by jbooth »

gsmaverick wrote: post #48
jbooth wrote: post #38

Earlier quoted context omitted.

The, uh, the united states doesn't have a central bank. I don't want to be nitpicky here, but this is a conversation about not turning your brain off. Are you referring to the federal reserve? They don't regulate anything. The SEC? They're in no way similar to a central bank and have been regulating progressively less over the last 70 years. What changed in 2008 that made their regulation more burdensome? If you're m…

The United States does have a central bank; it's the Federal Reserve. They kept interest rates abnormally low for much of the last decade which meant that borrowing was incredibly cheap causing many consumers to overextend themselves. Regulation, government intervention they're the same thing. It is undeniable the government intervened in the housing market and this was discussed earlier in this thread[1]. [1] http:/…

"Regulation, government intervention they're the same thing."

Again. This subthread is not about turning your brain off.

The fed changing interest rates is not regulation. It's not writing laws, it's not forcing anyone to do anything.

Re: Markets are Efficient if and Only if P = NP

#56

Post by knowtheory »

tmeasday wrote: post #34
zeteo wrote: post #13

Even if the problem that markets are trying to solve is NP hard, this doesn't mean anything. Market forces are always subject to random factors. For practical purposes, randomized approximation algorithms are actually a great choice for tackling NP hard problems.

Sometimes. I think the simple answer to that is 'it depends on the problem'.

That's not an answer, that's an IOU for an answer, unless you can begin to describe broad classes of problems and how they differ.

Re: Markets are Efficient if and Only if P = NP

#57

Post by scott_s »

jerf wrote: post #39
jbooth wrote: post #31

Earlier quoted context omitted.

Well, I certainly can get behind your definition more.. but a gigantic 10 year bubble sticking out to the side and then bursting doesn't seem efficient by that definition. You'd think the wizards of the market, with the billions of dollars in pay they're collectively bringing home, would've corrected for that sooner, no? This concept that in 2008 the market was responding to incentive misalignments created by governm…

"but a gigantic 10 year bubble sticking out to the side and then bursting doesn't seem efficient by that definition." Where was the enormous arbitrage opportunity left behind, and why didn't you (or anybody else) exploit it? Maybe you can answer that, but if you can't, you're not arguing for "inefficiency", you're still arguing "not good". You also appear to be conceiving of regulations as something you simply have "…

I slogged through much of the Financial Crisis Inquiry Report (http://fcic.law.stanford.edu/). One of the conclusions was that while mandating loans didn't help, they weren't numerous enough to be a major cause. Once the mortgage-backed securities engine got churning, there was an enormous demand for more loans to feed it. That provided most of the incentive for the bad loans, not government mandate.

Re: Markets are Efficient if and Only if P = NP

#58

Post by scott_s »

CWuestefeld wrote: post #15
jbooth wrote: post #8

Earlier quoted context omitted.

Was the fall of 2008 "extremely close" to optimal? I'd submit that worldviews based on 1-dimensional criteria like "market!" or "hayek!" fall pretty far short of the mark. Although I can see the attraction. It's nice to simplify things to a level where a human being can actually have the answers with a high degree of confidence. Both the tea-party-hayekians and the linked paper fall into this trap.. what about inform…

I'm not sure you understand what the free market is, or how it works. The fall of 2008 wasn't a free market. In particular, the government was forcing lenders to accept more risk (viz sub-prime borrowers) than they would otherwise have done. Also, the lenders themselves incorrectly modeled their risk exposures. None of your objections: what about information asymmetry? What about borked incentives? What about just pl…

In particular, the government was forcing lenders to accept more risk (viz sub-prime borrowers) than they would otherwise have done.

As I pointed out above, the number of loans made because the government said-so is not enough to account for the problems they encountered. Because of the mortgage-backed securities industry, there was incentive for banks to make sub-prime loans and offload the loans (and the risk) elsewhere. In other words, the government's requirements did not lead to the glut of sub-prime loans.

I also find your comment about lenders incorrectly modeling risk strange - it's a true statement, but I don't see how that means the market wasn't "free."

Re: Markets are Efficient if and Only if P = NP

#59

Post by jacoblyles »

jbooth wrote: post #49
jerf wrote: post #39

Earlier quoted context omitted.

"but a gigantic 10 year bubble sticking out to the side and then bursting doesn't seem efficient by that definition." Where was the enormous arbitrage opportunity left behind, and why didn't you (or anybody else) exploit it? Maybe you can answer that, but if you can't, you're not arguing for "inefficiency", you're still arguing "not good". You also appear to be conceiving of regulations as something you simply have "…

The original contention was "extremely close" to optimal. I'm merely contending that the markets fuck it up big occasionally here, pretty low hanging fruit for me. As for why I didn't exploit it? I don't know, I don't work on Wall Street, don't play that game? I could've told you that the mortgage thing was a problem long before the ratings agencies, apparently, but everyone knew in hindsight. Anyways, 10 years is pr…

jbooth, if you're going to accuse others of ignorance and "turning off their brain", you should probably educate yourself about the issues you are talking about. The size of Fannie and Freddie's balance sheets were around $500 billion and taxpayers have so far lost about $300 billion on the bailouts of the GSEs. So, they were not insignificant contributors to the crisis, and reasonable economists disagree about the share of blame that the GSEs own. The mortgage interest tax deduction is about another $100 billion of subsidy to the mortgage market annually. Some people question the wisdom of these measures and the government's role in propping up the bubble.

Housing subsidies directed at minorities were increased over time in the Clinton and Bush administrations, but weren't large compared to the massive overall portfolio.

Re: Markets are Efficient if and Only if P = NP

#60

Post by aphyr »

If we allow that large, computationally capable entities are likely to form in free markets, it is probable that individuals will be faced with a market in which the prices of major goods vary wildly from minute to minute, from person to person, and are predicated on both fantastically complicated loyalty programs and data collected about the individual and current market circumstances, processed in statistical models that no human has ever understood or evaluated. A market in which machines algorithmically game humans and each other by adjusting prices and the nature of the goods themselves to exploit cognitive or algorithmic flaws in other actors.[1]

The classical model of efficient, free markets ignores computational complexity and time bounds, as the article notes. These were reasonable approximations for a hundred or even twenty years ago, but I can forsee environments in which these factors lead to significant asymmetry. I question whether humans can really compete successfully in this environment, and whether the process of doing so is really best for us.

[1] Cases in point: Amazon, insurance companies, grocery stores, every major retailer, Google, high-frequency trading firms, airline ticket sales.

Post reply on HN