> I hope others see that our markets are designed. They are human artifacts. They are not natural laws to be accepted as-is.
“Markets” are not the same thing as “the market”. Markets are human artifacts; the market is an emergent behavior. It’s governed by laws beyond our control as surely as the behavior of materials are governed by thermodynamics.
If the behavior of your markets fails to match the reality of the market, the best you can hope for is to only introduce a small amount of inefficiency. Of course, utopians who ignore the real nature of the market usually find themselves lethally encumbered by deadweight loss.
> your goal is to allow companies to be judged on their own merits, instead of someone else's.
What is this supposed to mean? Why would a company’s subjective evaluation of themselves be a better basis for market valuation than the evaluation of the people who actually buy and sell interest in that company?
The way you've put this bears more resemblance to religion than anything else.
"Markets" are real things. I can walk to them, or navigate to their website, or call them up.
You are claiming that there is an invisible, metaphysical "the market", which exists in the Platonic world of forms, that governs the behavior of these genuine markets.
This is like reading "The Gods of the Copybook Headings", and founding a religion which worships and offers sacrifice to those gods!
This "the market" you refer to is an abstraction of the behavior of real markets. It can give us all sorts of guidance as to what will happen when various rules are applied, or not, to these real markets; but comparing these broad heuristic rules to thermodynamics has me suspecting you don't understand either thermodynamics nor macroeconomics.
Every market has rules. Invariably, you are not allowed to stab a vendor and take their stuff. Sometimes there are things you can't sell; sometimes only Dutch auction is allowed, and so on ad infinitum.
This is simply a market with different rules from those governing other stock markets. We're all free to speculate on the consequences those different rules will have; perhaps they will summon literal Gods of the Copybook Headings, which will return upon us with actual terror and slaughter.
The safe bet is that prediction and reality will be divergent. This is economics, not thermodynamics, after all.
>We're all free to speculate on the consequences those different rules will have
To echo what my sibling commenter said, those consequences are what the grandparent comment is referring to as the emergent market - because it's not possible to control or accurately predict the behavior of every single actor.
It was a rebuttal to a statement saying we don't need to consider markets as pure forces of nature, because how we set them up can affect the outcome. I think everyone in this conversation agrees that it's both true that the design of a market affects how people act in it, as well as it being impossible to predict or control perfectly. The question is to what extent it actually is possible.
On the contrary, he/she is claiming that "the market" is an emergent behaviour of the dynamics of a system. We set up "market" and instigate rules. "The market" is what happens subsequently.
If you think emergent behaviors are "invisible" and "metaphysical", this suggests a modeling deficiency on your part rather than anything interesting about the real world.
> This "the market" you refer to is an abstraction of the behavior of real markets
The market isn't an abstraction at all - it's a handle for the emergent behavior of the sum total of all economic activity.
"The market" hardly exists in a bubble entirely outside of human sentiment and human institutions.
Human institutions; the prevailing ways that capital moves; legal institutions; customary behaviors, mores, etc, all have a profound impact on not just markets but also The Market (tm). The edifices we create are important.
> Markets are human artifacts; the market is an emergent behavior.
If "the market" is a 2d array of 1 and 0 that emerges from a cellular automata, then "markets" are the rules for that automata. The human artifact defines the rules and incentives that cause the emergent behavior seen by aggregating all market participants.
There is no Platonic ideal "markets" independent of the human structures where they operate any more than you can define how fast a pendulum swings without specifying its length.
"Markets" (locations or brokers designated for trade) represent a very small fraction of all human economic activity. Conflating the rules of NYSE with the rules of the market is a mistake.
> There is no Platonic ideal "markets" independent of the human structures
You need to understand that there absolutely is a single "the market" and it's not platonic at all. It refers to the sum behavior of all human economic interaction, which is absolutely a physical reality.
> It’s governed by laws beyond our control as surely as the behavior of materials are governed by thermodynamics.
It's funny you make this comparison. Because so many of these supposed laws about "the market" violate an effective equivalent to the second law of thermodynamics for information systems: P=/=NP. The claims often made about the transcendent nature of the laws governing "the market" violate the computational properties of reality as we understand it [1]. If "the market" behaved as so many claim then we would have proof of P=NP.
Evolution is also an emergent behavior, of biochemical systems. But we also understand how it functions as an algorithm. So much so we now use it to aid in the design the wing shapes of our planes, the design of new high efficiency radios, and to find new algorithms.
The entire universe is governed by laws beyond our control. That doesn't prevent us from understanding them and bending them to our will. As we have done with ropes and levers, atoms and proteins, so too can we understand and engineer markets.
As it happens markets are likely in a similar class of algorithm as evolution: efficient approximations of NP problems. This is why people use markets as the basis of things like prediction markets, scheduling algorithms, and bin packing.
"The market" is a claim to the "god is in the gaps" no different than intelligent design is. It even does the whole song and dance of elevating an approximation algorithm to some unknowable divine will. An approximation algorithm we already understand and use in engineering everyday.
[1] Make no mistake, bits of information are physically connected to reality, a non-trivial NP algorithm running on human timescales would boil the oceans. See: https://en.wikipedia.org/wiki/Landauer%27s_principle
This isn't known to be true. In terms of computational complexity, market clearing is PPAD (https://en.wikipedia.org/wiki/PPAD_(complexity)). It may turn out that PPAD is computationally tractable even if P != NP, though it is thought to be unlikely.
I would argue that reinforces my point rather than weaken it. My point being that the market is an understandable thing becomes more reasonable if it turns out it's not even an intractable problem that we are discussing.
But I also think you are misunderstanding the argument I was replying to in the first place, because a lot of the claims about "the market" are about the second order allocation of resources that markets cause and not (just) the clearing of them. For example market clearing is how we arrive at the price given the buy and sell orders, but the original goal was the efficient allocation of resources. For which markets are obviously not a perfect solution (it's relatively easy to create order books which end up with unfulfilled orders despite money/weight being available to argue for that allocation), even if they are a good approximate solution (via market clearing). The claims about "the market" are market fundamentalism, that the market is the ideal allocation structure, that we can't find a better or even equivalent solution, because that's just how good markets are. That was what I was rejecting. Not market clearing.
> The claims about "the market" are market fundamentalism, that the market is the ideal allocation structure, that we can't find a better or even equivalent solution, because that's just how good markets are.
It's more usual to accept a weaker claim: that markets are the most efficient approximation to the perfect algorithm for price discovery, that is, they are the best way we know of to approximate perfect pricing given the constraint of time.
Insisting that no better approach is possible is a strong claim; you'll find it in the wild, but it's always dicey to make confident predictions about the future of human ingenuity.
But at the same time, there's no reason to waste much breath on technology which doesn't exist: markets are what we have, and they're what we should use, with the burden of proof firmly on the inventor of anything new to demonstrate that it has greater efficiency.
> It's more usual to accept a weaker claim: that markets are the most efficient approximation to the perfect algorithm for price discovery
This claim is just as problematic. You may view it as weaker, but algorithmicly speaking it's the same nonsense. Algorithms in similar complexity classes are usually translatable to each other, which is to say they can be framed in terms of each other. You appear to be claiming that markets are the most efficient variant of an algorithmic class. Which is, again, nonsense. It might have some very nice qualities compared to other algorithms within it's class, but to claim it's the most efficient is, again, problematic.
A claim like that would require proof, not just the absence of contradicting evidence.
> Insisting that no better approach is possible is a strong claim; you'll find it in the wild
Ok, but I do believe the poster I was originally disagreeing with is arguing that:
> the market is an emergent behavior. It’s governed by laws beyond our control as surely as the behavior of materials are governed by thermodynamics.
> If the behavior of your markets fails to match the reality of the market, the best you can hope for is to only introduce a small amount of inefficiency. Of course, utopians who ignore the real nature of the market usually find themselves lethally encumbered by deadweight loss
This appears to be a strong claim that nothing can be more efficient than "the market" without introducing inefficiency or "lethal[.]" dead-weight loss.
> markets are what we have, and they're what we should use, with the burden of proof firmly on the inventor of anything new to demonstrate that it has greater efficiency.
This is a fallacious argument against my point.
To begin with, a new algorithm in the same class as markets is likely to have the same or similar efficiency. And it will be the other properties of it that are more interesting.
But the real issue is that the double standard people argue with the efficient market hypothesis (and especially the economic calculation problem) is problematic because it misunderstands computation. You simultaneously claim (in this post) to have the best approximation due to a lack of contradicting evidence, but then also want alternatives to provide proof of beating that physically impossible standard.
My main response though is: just because markets are the best thing we have does not mean we should accept absurd claims about them. Just because solar energy is effectively unlimited, does not make it a perpetual motion machine.
> This claim is just as problematic. You may view it as weaker, but algorithmicly speaking it's the same nonsense.
It's an existential claim, not a mathematical claim.
The claim is simply that markets are more efficient than other mechanisms, proposed or extant, which actually exist in the real world.
Clearly to quantify this claim requires some real mathematical analysis, but that's quite beyond my talents or interest. I'm just here to point out that the weakened claim is "markets are the best we've got", not anything theoretical nor abstract.
And I'd suggest reading my reply to the absurd thermodynamics bit before assuming too much about my thoughts on it!
> And I'd suggest reading my reply to the absurd thermodynamics bit before assuming too much about my thoughts on it!
I mean. My issue is multi-fold. I have an issue with market fundamentalists who treat it like a religion. But I also have an issue with economists who use methods of analysis that aren't empirical.
> It's an existential claim, not a mathematical claim.
This isn't just an issue of philosophy and mathematics in the abstract. Computation is a physical phenomena. The claim of efficiency being made here, even in the weaker version, is not congruent with reality.
It's an existential claim that would be no different than chemists shouting "this chemical reaction will continue forever" (a violation of the second law of thermodynamics).
> I'm just here to point out that the weakened claim is "markets are the best we've got", not anything theoretical nor abstract.
The weaker claim is no different than doctors who once believed in the four humors. It's not technically an incorrect categorization of reality given current knowledge in the field. But it's a piece of general economics knowledge that has no backing in reality, nor is falsifiable, nor provides predictive use.
And every time an economist spouts that sort of nonsense I can't help but remember their field must be primarily pesduo-science. I don't know how they expect me to take them seriously when they are still yelling about four humors and perpetual chemical interactions.
Because existential claims like this one are incongruent with reality. I don't know many times I can emphasize this point, the claim, even the weakest version I have seen, is not physically possible, hence it would require extraordinary evidence. It would upset the entire field of computation if it were true. It stands in opposition to modern pillars of computational reality, no different than claiming one had a perpetual motion machine would physics. Humanity would be like gods if it were true. You can continue to claim it all you want, but it's as misguided as claiming the earth is flat.
There is one prime difference between computation and economics - and for that matter, material and social sciences. Computation doesn't study humans.
Wouldn't a "unified theory" of economics (a deterministic one at that) effectively rule out free will? After all, what is "the market" if not the aggregate behaviour of all human actors in it? And if you accept free will, doesn't that conversely create a problem where the best you can do when predicting markets is correlation, not causality?
>an effective equivalent to the second law of thermodynamics for information systems: P=/=NP.
This is not true. For example, the second law is only a statistical claim - the larger the system, the more likely it holds. But the second law is not a law against entropy ever going the other way. It's entirely possible, but not likely that another universe pops out of a quantum fluctuation at any moment. See the concept of a Boltzmann Brain for where physicists think this leads.
And P != NP is not a law at all - it's a guess, since no one knows if it's true. Complexity classes certainly changed with the introduction of quantum computer, allowing (very) few problems with known exponential classical Turing time to be done in polynomial quantum Turing time. It is also known that if closed timelike loops are allowed in computation then P=NP, but it's not known if we can engineer such physical items.
>The claims often made about the transcendent nature of the laws governing "the market" violate the computational properties of reality as we understand it
As evidence of the falsity of the above claims, TQFTs routinely compute NP hard problems, which is why for some time Friedman has tried to leverage them to solve NP hard computational problems.
Here's [1] but one paper showing this to be true. "Non-Abelian topological quantum field theories exhibit the mathematical features necessary to support a model capable of solving all ⧣P problems, a computationally intractable class, in polynomial time. "
Roughly, TQFTs perform certain "computations" on knots in polynomial time that are known to be NP hard in classical or current quantum computing models.
[2] is another nice take in Annals of Mathematics, the most prestigious journal in math.
Thus reality routinely solves NP hard problems. Thus your claims are not true.
“Markets” are not the same thing as “the market”. Markets are human artifacts; the market is an emergent behavior. It’s governed by laws beyond our control as surely as the behavior of materials are governed by thermodynamics.
If the behavior of your markets fails to match the reality of the market, the best you can hope for is to only introduce a small amount of inefficiency. Of course, utopians who ignore the real nature of the market usually find themselves lethally encumbered by deadweight loss.
> your goal is to allow companies to be judged on their own merits, instead of someone else's.
What is this supposed to mean? Why would a company’s subjective evaluation of themselves be a better basis for market valuation than the evaluation of the people who actually buy and sell interest in that company?