[tsimionescu]

> They are largely based on supply and demand, but most companies need to predict supply and demand months and years in advance because converting raw material into a sale to the end user is a multi month to multi year process, so a natural result is that the prices are based on speculation of supply and demand, months to years in advance.

But that practically means they are based on speculation about possible future supply and demand, and we are more or less taking it on faith that that speculation is correlated to actual supply and demand.

[avianlyric]

> But that practically means they are based on speculation about possible future supply and demand, and we are more or less taking it on faith that that speculation is correlated to actual supply and demand.

This statement almost seems tautological. Of course prices are based on speculation about possible supply and demand, nobody has a crystal ball that can tell them “actual” supply and demand.

In a competitive and liquid market, you expect prices to rapidly approach “optimal”, because otherwise there’s an opportunity there for someone make money out of the market inefficiency, buy correctly predicting when supply is high, and buying, then selling when supply is low. Which is exactly what future etc do, except without the need to actually move the physical commodity around.

[tsimionescu]

> Of course prices are based on speculation about possible supply and demand, nobody has a crystal ball that can tell them “actual” supply and demand.

Exactly, no one has such a crystal ball. And this actually generally applies not just to future supply and demand, but even to the past: it's actually impossible to measure supply and demand across any significant industry, to check whether prices matched it or not.

And yet, economists and economical theory enthusiasts talk about the "law of supply and demand" as if it's some scientific observation, and not just a simplistic model that seems intuitive.

> In a competitive and liquid market, you expect prices to rapidly approach “optimal”, because otherwise there’s an opportunity there for someone make money out of the market inefficiency

You might expect that if you believe in the law of supply and demand, but as we were discussing, that is not how prices are actually formed, and anyone betting based on observed supply and demand (to the extent that it is actually possible to observe them) will be beat in general by others who are betting based on current speculation, which is how prices are actually formed.

For example, if you are betting that gain prices will increase in the winter because that's what you think they did every year, and ignore some prominent pundit predicting that they will decrease this year, you may well lose the bet if everyone else believes the pundit. And that will be true regardless of whether grain will be in low supply or not.

[avianlyric]

> You might expect that if you believe in the law of supply and demand, but as we were discussing, that is not how prices are actually formed, and anyone betting based on observed supply and demand (to the extent that it is actually possible to observe them) will be beat in general by others who are betting based on current speculation, which is how prices are actually formed.

But we do see prices approach optimal. Your argument is predicated on the idea that the law of supply and demand is only a useful model if the actual price always follows observed supply and demand. But for that to ever be true, it would require speculators to have a crystal ball, otherwise there’s no reason to believe the speculated price will always match the “optimal” when the point in time being speculated about actually occurs.

The law of supply and demand tells what market systems will trend towards. But like complex control system, having a governing idea about long term trends doesn’t mean momentary perturbations don’t occur, it just helps you understand what the system will do after the perturbation.

To claim that the law of supply and demand is useless, is like claiming that Hooks Law is useless for understanding how suspension systems in cars work, because perfect springs don’t exist, and cars don’t remain stationary.

Perfect markets don’t exist, perfect information doesn’t exist, so why would anyone expect real markets to perfectly follow the law supply and demand? And clearly markets do follow the law of supply demand at the macro level, when long periods of time are considered, otherwise commodity pricing would be entirely arbitrary and wouldn’t in anyway reflect the value of the commodity to society at large.

[tsimionescu]

> But we do see prices approach optimal.

Where do you see that? What does it even mean, how can you objectively tell what is the optimal price for a good or service, so that you can later say that the market converged to it?

For example, is 1000$ the optimal price for an iPhone? Or is it simply the price Apple chose? If they sold it for 500$, would they make more or less money? How do you know?

> And clearly markets do follow the law of supply demand at the macro level, when long periods of time are considered, otherwise commodity pricing would be entirely arbitrary and wouldn’t in anyway reflect the value of the commodity to society at large.

I would argue that it often is, at least for many non-essential products. The price of many non-essential goods is much much higher than the price of essentials, even when those non-essential goods are cheap and easy to manufacture (say, softdrinks or many cosmetics). The price of most energy resources is largely controlled by non-market forces, even on the face of it.

Also, why restrict this discussion to commodities? If we switch to investments, the price of stocks is quite obviously arbitrary as well, with "market makers" often controlling the allowed prices (or at least, price volatility) for stock. The price of real-estate is often determined to a large extent by the price of borrowing, and that is quite explicitly set by banks and the central bank based on nothing related to supply and demand. Services are even more complex, with huge variations in price based on entirely subjective factors.

[avianlyric]

> Also, why restrict this discussion to commodities?

Because they’re generally pretty fungible, and have a larger number of sellers and buyers involved in the market, hence their markets are more likely to behave like an ideal market. The same does not apply to housing, or services.

> Where do you see that? What does it even mean, how can you objectively tell what is the optimal price for a good or service, so that you can later say that the market converged to it?

For commodities, I would point to the reasonable price stability that exists. As evidenced by the fact that basic goods don’t frequently suffer from repeated shortages or gluts of supply. Strongly indicating that the price is both high enough to incentivise production, and stable/low enough to allow for relatively low risk long term investment in production, because continuous long term demand is expected.

> For example, is 1000$ the optimal price for an iPhone? Or is it simply the price Apple chose? If they sold it for 500$, would they make more or less money? How do you know?

These is nothing about the iPhone market that suggests it’s anything close to an ideal market (for one Apple have a monopoly on iPhone sales), so I don’t know why you would expect it to behave like an ideal market.

> The price of real-estate is often determined to a large extent by the price of borrowing, and that is quite explicitly set by banks and the central bank based on nothing related to supply and demand. Services are even more complex, with huge variations in price based on entirely subjective factors.

What’s your point? Of course a law describing how ideal markets work doesn’t correctly describe markets well know for being extremely distorted and non-ideal. Next you’re going to tell me Newtons laws of motions are all useless because they can’t help you model the behaviour of objects travelling at relativistic speeds.

You seem to be struggling with the idea that a model doesn’t need to be perfect, or applicable to every real world scenario, to be useful. All models have their limits, that no surprise to anyone. That doesn’t make them useless, it just means you need to be aware of limitations, and adjust expectations appropriately.

[PinkRidingHood]

The pundit's claims are part of the speculation. If you speculated that the pundit's words would matter, you'd short or sell out grain, making the pundit's words matter. Supply and demand are only sometimes good measures of people actually using or producing these products (and when they fail bubbles occur), but are always good measures of the spending of both people actually using the product, and the investors buying it in wanting to sell it.

In any case, as the investors buy the product they eventually have to sell it, meaning over the long run demand and supply are good measures of "actual" usage and production. You know this because there's never a bubble that doesn't collapse and return to equilibrium.

[tsimionescu]

> In any case, as the investors buy the product they eventually have to sell it, meaning over the long run demand and supply are good measures of "actual" usage and production.

That's not necessarily true. It depends on the product and industry, but it's absolutely possible to horde products for long periods of time (see diamonds and gold) or to choose to destroy products rather than sell them at a price you don't like (see public transportation in many US cities in the 50s, or fancy food items).

[signatoremo]

> you may well lose the bet if everyone else believes the pundit

Can you give a real life example of when a pundit giving a predict against the actual supply that had everyone buy-in? Says, oil price goes down even when demand is hot?

[tsimionescu]

The price of oil is a particularly bad example, since it is heavily controlled everywhere in the world, as is supply. The international price of oil is largely set by OPEC or the USA or a few other countries (depending on geographical region).

Also, this type of pundit influence is common with the price of stocks. It also happened with the price of natural gas in Europe last year, when it increased based on lack of confidence in reserves that turned out to be misplaced, and never really recovered.

[PinkRidingHood]

An inverse example would be any bubble, for example the housing bubble that happened over a decade ago. Prediction of house prices going up eventually reached a point where house prices were going up, despite low "actual demand" (of course, if you count "demand for investment", then prices are absolutely a measure of that).