Hacker News Reader

Comment by dcow | Hacker News Reader

Is it just me or do “markets” feel like a really dumb AI that is incapable of any sort of strategic planning? It feels like it has the maturity if a child in grocery store swooping up all the candy and scoffing at all the vegetables. Undoubtably markets are great for a lot of things but simply devastating when ineffectively applied to mission-critical goods and services that involve contextually aware knowledge and meta-level thinking. It seems like theres a missing component that would observe swings in demand and try to smooth out the impacts by e.g. subsidizing industries where demand has tanked or taxing industries where demand has skyrocketed in a tapering way (tax that reduces over time as the rate of change of the price of the commodity changes) to avoid something like shuttering a sawmill because of a temporary swing in market economics.

View on HN

They're as dumb / clever as people are.

Unless we had a crystal ball we were unlikely to do better than the market and we would have likely predicted demand to go down during the pandemic as well and decommissioned some of the sawmills that we need right now.

If you predicted what the article is citing as the causes for the shortage, you would have started a sawmill in 2020.

Do you trust the central planners to plan better than participants with a vested interest in the outcome?

In general, I’m deeply skeptical that central planning and government “picking winners” would achieve an overall benefit. At a minimum, it does not appear to have a strong record so far.

Greed is a miserable planner.

It is the best so far. Show me one that works better on a large scale. The economy makes up of billions of transactions that are not planned but are there to fulfill greed, each in their separate ways end up acting as a choreographed system that the central planners could never replicate.

The naive solution I had in mind isn’t a centrally planned economy. It’s some type of addition to the concept of a market to deter volatility in specifically commodities markets where volatility is often disruptive and potentially catastrophic. I don’t trust a gov’t to centrally plan an economy, thats for sure, but I do believe society could iterate and improve on the way it applies markets to solve various problems.

There is a vast space of possibilities between (and orthogonal to!) the line between "unregulated market" and "centrally managed economy." Just because someone says "this didn't work" doesn't mean they would say "let's do 100% the opposite."

Surely, in this case, there are better ways to smooth out these spikes and dips in flows than just letting the price go sky high so that only the rich can build or repair housing for a year or two.

You could have some kind of entity that buys up supplies when the price is low and sells when the price is high.

In fact, it’s legal for you to create one today, and if you operate it successfully, you’d make a lot of money.

Keep in mind that public opinion is turning against these "speculators" that you describe.

Blackrock Investment is doing everything you describe, but for housing. They are being heavily derided by the media for doing so.

Oh believe me, I am aware. Though that’s a great point.

Housing isn't cheap right now, is it?

> You could have some kind of entity that buys up supplies when the price is low

How do you know the price is low and to start buying or wait because the price might fall down more?

You'd also be subject to inventory tax on your stored stuff no?

The government wants to make sure things are moving in the government, not sitting around

Central planning has a great track record in corporate governance. Walmart employees 2.3 million people and as far as I know is centrally planned. Central planning can at least scale up that far. Sixteen states have a population less than that. I don't see any indication that scale is holding Walmart back.

Markets can strategically plan though. It's why turkey doesn't cost several thousand dollars a pound every November in the US.

That is because the same happens every year. Even a bad AI or child (using OPs example) can plan for that. But when it comes to events that happen sporadically markets seem to have issues where most sane adults would be able to handle it.

It's probably a sane choice for most products to -not- prepare for very rare disasters like a hundred year pandemic. After all, markets optimize for profit, not convenience or safety.

These shortages will all smooth out relatively soon. It can cost a lot of money to be prepared for everything.

Sometimes the most efficient option is not the safest or happiest or winning option. If our prime goal was cheapest lumber 99% of the time, then sure a traditional market is very good at that. I would happily pay for 1% more expensive lumber all the time to avoid paying for 1000% more expensive lumber (read: not being able to buy lumber) when I really need it.

Would you pay 40% more for lumber to avoid 1000% price spikes though?

>> But when it comes to events that happen sporadically markets seem to have issues where most sane adults would be able to handle it.

The market is made up of sane adults. It's a feature, not bug. Fortunately. Or unfortunately.

Markets allow the type of behavior that we see “sane” adults engaging in by design. So if you update the design or the rules of the market, you will observe different behavior from its participants.

“Sane adults” only do well when predicting in hindsight, where algorithms are unbeatable.

Did you read the article? The lumber companies DID plan ahead using their knowledge and experience, it just turned out to be wrong.

I'm curious to see what economic system would be able to foresee a drastic increase in the demand for lumber during the pandemic-induced depression, despite all the evidence suggesting that the opposite would happen.

That’s not how I read the infographic. I read as: wood bug ruins a bunch of trees then human speculators shorted the lumber market which caused sawmills to close. The problem here is humans being able to profit off of volatility in commodities during a crisis. I’m entertaining a solution that prevents humans from being able to do this by including built-in buffers that apply consistently and equally without human intervention.

Demand didn't skyrocket, it simply stayed the same but because humans tried to preempt things sawmills closed and boom now there’s no supply to meet the existing demand so prices skyrocketed.

Plain old supply and demand instantaneous markets are dumb,as you say. However futures markets are just as smart as the humans who run them. If there isn't a futures market for lumber (or microprocessors, or RAM, or whatever else has big fluctuations in price) then whoever starts one is in line to make a killing. Manufacturers want to be able to sell at stable prices so that they can do long term investments in efficiency.

The whole point of a market is that if you think its wrong you can profit by correcting it. If you think prices will fall, have you shorted it? If they're going up I assume you're long?

So a bunch of people think the lumber market is going down during the pandemic and take out short positions on lumber futures because they want to profit and then lumber mills start closing in response. The problem is the lumber mills closing. That’s not part of the picture that can be easily corrected. Now the market participants, who were wrong, have impacted the production capacity of lumber and we cant get any new sawmills because it’s too expensive to spin one back up? Maybe the info graphic isn’t telling the whole story, though?

Enabling more flexible capacity would fix this problem by introducing additional "contextually aware knowledge and meta-level thinking" into the market. Higher prices would spur additional market entrants to come online faster. To do that you'd need to reduce regulation or at least make regulatory processes more efficient.

According to Bloomberg lumber dropped ~18% this past week, the highest drop on record since 1986. [1]

[1] https://www.bloomberg.com/news/articles/2021-06-11/lumber-pr...

My observation is that there are very little need to understand about market, you only need to know the basic where a young child could understand as lead time and basic cost.

There is a whole lot to understand but possibly never be able to fully understand about human nature.

The market, after all is just a sum of all human behaviours.

Economics is the same, human are highly irrational.

So viewing the whole thing from a rational perspective will make you feel like the market is really dumb. But if you look at it from a human perspective, in the seat of those managing their company from Fear, Greed, Risk and Security everything makes lot of sense.

because of a temporary swing

You only know it is a temporary swing retrospectively. It is impossible to know if it is a temporary swing. It seemed more likely a year ago that we would still be waiting for a vaccine in 2021. If someone was trying to guess the future and massage the market, they just as likely might have assumed even less demand and subsidized shutting down more mill capacity to reduce virus spread which seemed like a bigger problem than lumber prices.

You can know that the rate of change of a given commodity over some window is higher than an acceptable threshold and tax/subsidize it using funds collected from all transaction made on the market or something until that window passes or until the commodity’s price stabilizes (is no longer unacceptably volatile), whichever is shorter.

The idea I’m trying to capture is to specifically not apply arbitrary human sentiment and planning to the daily operations of the market and instead have the impetus to keep e.g. prices stable built into the market itself (and only in scenarios where a less volatile economy is desirable) so that the market remains like an AI but just operates a little better for the given problem it is solving. I think it’s fair to treat markets like software that can be changed and improved over time in order to better solve the problems they’re deployed to address. Maybe a more volatile market is better for some other set of goods or services? I think it is up to humans to determine the parameters and desired outcomes of the market they want governing the exchange of goods and services in different areas of society.

There are smart actors that plan, but that is rare, since most humans are short term thinkers (or are negative term thinkers, e.g. dwelling too much on their past failures to plan ahead).

Why does this matter? It's an unpredictable short term problem that is going away. If you can build an AI that can predict the increased demand during the pandemic then the market won't need the rest of the AI.

From my understanding amount paid for the raw trees by mills hasn't gone up, so it's clearly on the mills that are capturing the excess profit; makes me wonder if there needs to be a mechanism for society to capture much of that unusual profit to then redistribute as one such mechanism.

We already have a system that captures a part of both usual and unusual profits and redistributes them. It even scales with the size of the profits. That’s how taxes work.

Reply on news.ycombinator.com