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

A detailed exposé on how the market is rigged from a data-centric approach

nanex.net

Post reply on HN
Print view View original

141–150 of 153 posts

Re: A detailed exposé on how the market is rigged from a data-centric approach

#141

Post by VikingCoder »

harryh wrote: post #139
VikingCoder wrote: post #138

Earlier quoted context omitted.

Please don't be so rude, it doesn't add anything to the conversation. There are many markets where you list a product for a price, and are legally bound to sell them at that price. Those markets function, proving that withdrawing quotes is not necessary to make markets work.

What markets are those? I can think of no markets in which you're not allowed to change the price of whatever goods you're selling.

Best Buy runs a flier with a price. You walk in, and Best Buy tells you they're not selling that product at that price... but they do have it in inventory at a higher price.

You drive up to a gas station, listing one price on its sign. By the time you pay, they've changed the price.

You see a house for sale. You make an offer, at that exact amount. There are no higher competing offers, but they want to back out.

You have legal recourse in each of those situations.

Re: A detailed exposé on how the market is rigged from a data-centric approach

#142

Post by harryh »

VikingCoder wrote: post #141
harryh wrote: post #139

Earlier quoted context omitted.

What markets are those? I can think of no markets in which you're not allowed to change the price of whatever goods you're selling.

Best Buy runs a flier with a price. You walk in, and Best Buy tells you they're not selling that product at that price... but they do have it in inventory at a higher price. You drive up to a gas station, listing one price on its sign. By the time you pay, they've changed the price. You see a house for sale. You make an offer, at that exact amount. There are no higher competing offers, but they want to back out. You…

1) That's not because Best Buy isn't allowed to change prices. There's even fine print on their marketing that says the prices on the flyer are marketing only and may or may not match pricing when you get to the store. They just choose not to most of the time for marketing reasons.

2) Actually, at least in New York there are no regulations that discuss the large roadsign pricing signs at gas stations. The law just says the price on the pump must match what you get charged. If they change the price at the pump but get behind on updating their big sign you are SOL.

3) People back out of selling houses all the time.

And for all of these it's worth pointing out that what we're really arguing about is time scale. No one would argue that the gas station (for example) wasn't allowed to sell gas for a different price today than it charged yesterday right?

Bid and ask prices for securities just change prices faster than what we're used to for retail products.

To really expand the scope of this discussion it's worth noting that time scales for retail product price changes are actually shrinking. Walmart has experimented with electronic labels in stores that allow them to change prices in real time. Uber changes prices in real time based on demand. It's interesting to watch consumer reaction to these new trends.

Re: A detailed exposé on how the market is rigged from a data-centric approach

#143

Post by tjgq »

ska wrote: post #83
asgard1024 wrote: post #64

Earlier quoted context omitted.

> If you want a reasonably efficient market, you need some participants to have close to perfect information. Is this proven somewhere or you just assume the optimal strategy for markets is continuous? I mean, it's not clear that the optimal strategy for "slightly imperfect markets" is at all close to the optimal strategy for markets with perfect information. And I actually doubt it can be proven, in the general case…

Not only is it not clear, intuition from other areas of optimization would suggest it's unlikely to be true. I've asked a couple of economists about this, but didn't get a satisfying answer. To be fair, it wasn't their area at all - and I may just have misunderstood what they were saying.

I'm curious: can you name an optimization problem you would get that intuition from?

Re: A detailed exposé on how the market is rigged from a data-centric approach

#144

Post by VikingCoder »

harryh wrote: post #142
VikingCoder wrote: post #141

Earlier quoted context omitted.

Best Buy runs a flier with a price. You walk in, and Best Buy tells you they're not selling that product at that price... but they do have it in inventory at a higher price. You drive up to a gas station, listing one price on its sign. By the time you pay, they've changed the price. You see a house for sale. You make an offer, at that exact amount. There are no higher competing offers, but they want to back out. You…

1) That's not because Best Buy isn't allowed to change prices. There's even fine print on their marketing that says the prices on the flyer are marketing only and may or may not match pricing when you get to the store. They just choose not to most of the time for marketing reasons. 2) Actually, at least in New York there are no regulations that discuss the large roadsign pricing signs at gas stations. The law just sa…

There's still bait-and-switch lawsuits... so advertising has some legal binding. I guess all advertisers know how to write in enough small print to get around it now, huh?

And I guess securities don't have any "this offer good for [x seconds]" on them. That's totally counter-intuitive to laymen, and speaking as a professional layman, I'm pretty sure I'm getting screwed because of it.

Electronic labels - you should be able to reserve a price for a time. Like, you want that shaving cream for $1.45, then you should be able to scan your Walmart member card at the electronic label to reserve that price. Should be valid for an hour, or a day, for up to X number of them.

...because the nightmare of picking it off the shelf, and the price is different when you check-out is just awful for consumers.

...and just to ramble a bit more... Sometimes it's cheaper for me to DRIVE to City A, fly back to my town, and then fly to my real destination than it is to book a flight from my town to my real destination. That's crazy, and I'm totally getting screwed, and I hope laws are enacted to stop that.

Re: A detailed exposé on how the market is rigged from a data-centric approach

#145

Post by harryh »

VikingCoder wrote: post #144
harryh wrote: post #142

Earlier quoted context omitted.

1) That's not because Best Buy isn't allowed to change prices. There's even fine print on their marketing that says the prices on the flyer are marketing only and may or may not match pricing when you get to the store. They just choose not to most of the time for marketing reasons. 2) Actually, at least in New York there are no regulations that discuss the large roadsign pricing signs at gas stations. The law just sa…

There's still bait-and-switch lawsuits... so advertising has some legal binding. I guess all advertisers know how to write in enough small print to get around it now, huh? And I guess securities don't have any "this offer good for [x seconds]" on them. That's totally counter-intuitive to laymen, and speaking as a professional layman, I'm pretty sure I'm getting screwed because of it. Electronic labels - you should be…

Ya, I get the whole "as a professional layman, I'm pretty sure I'm getting screwed because of it" when it comes to HFTs. But if you look at the actual data it's blindingly obvious that the exact opposite is happening. Trading costs have come down enormously!

Iit would be nice if folks looked at the data instead of falling back on their irrational gut logic.

Re: A detailed exposé on how the market is rigged from a data-centric approach

#146

Post by asgard1024 »

tjgq wrote: post #143
ska wrote: post #83

Earlier quoted context omitted.

Not only is it not clear, intuition from other areas of optimization would suggest it's unlikely to be true. I've asked a couple of economists about this, but didn't get a satisfying answer. To be fair, it wasn't their area at all - and I may just have misunderstood what they were saying.

I'm curious: can you name an optimization problem you would get that intuition from?

I am not a game theorist, but take Centipede game for instance. If you know the exact number of rounds in advance, the optimal strategy is markedly different than if you don't know it. And I think there are many weird behaviors like that in iterated games, where optimal solution for infinite time horizon is not the same (or doesn't even exist) as the limit of optimal solutions for finite horizons approaching infinity.

Re: A detailed exposé on how the market is rigged from a data-centric approach

#147

Post by noxn »

throwaway283719 wrote: post #132
noxn wrote: post #121

Earlier quoted context omitted.

"Fresh Apples here! Only the best apples for 2 dollars!" - "I would like one, please." - "Thatll be 2.50, sir." - "What? I thought you just said 2?" - "Demand has just gone up."

I'm glad you brought this up, because this kind of thing happens precisely never on a financial exchange. If you go to a store, the store owner sees you take your apples up to the counter, and so he can theoretically change his price before you get there (although in practice, if he ever did that he would soon be out of business). On a financial exchange, the market maker doesn't even find out that you wanted to buy…

I accept that your explanation makes sense, but the fact that this can happen automagically in an intransparent way when someone just wants to buy a number of shares at a quoted price just feels wrong. And reading a few other articles on Nanex also give me weird image of what happens in stocks in general.

Then again, I have no idea about these things and should probably shut up.

Re: A detailed exposé on how the market is rigged from a data-centric approach

#148

Post by noxn »

kasey_junk wrote: post #123
noxn wrote: post #121

Earlier quoted context omitted.

"Fresh Apples here! Only the best apples for 2 dollars!" - "I would like one, please." - "Thatll be 2.50, sir." - "What? I thought you just said 2?" - "Demand has just gone up."

This is the flawed analogy that keeps being propagated that is completely untrue. A much better one is to think of a string of gas stations running down the highway. They all have an advertised price. A tanker truck arrives at the first gas station and buys all the gas at the advertised price. It then goes down the road and buys all the next stations gas at the same advertised price. Then the manager of the 2nd gas s…

Fair enough.

Re: A detailed exposé on how the market is rigged from a data-centric approach

#149

Post by throwaway283719 »

noxn wrote: post #147
throwaway283719 wrote: post #132

Earlier quoted context omitted.

I'm glad you brought this up, because this kind of thing happens precisely never on a financial exchange. If you go to a store, the store owner sees you take your apples up to the counter, and so he can theoretically change his price before you get there (although in practice, if he ever did that he would soon be out of business). On a financial exchange, the market maker doesn't even find out that you wanted to buy…

I accept that your explanation makes sense, but the fact that this can happen automagically in an intransparent way when someone just wants to buy a number of shares at a quoted price just feels wrong. And reading a few other articles on Nanex also give me weird image of what happens in stocks in general. Then again, I have no idea about these things and should probably shut up.

Just remember that when you say

  "someone wants to buy a number of shares at a quoted price"
what you really mean is

  "a giant investment bank or hedge fund with some privileged
   information about a stock wants to buy so many shares that
   they actually need to go to multiple exchanges to satisfy 
   their demand"
and you'll be all set ;)

Re: A detailed exposé on how the market is rigged from a data-centric approach

#150

Post by TophWells »

harryh wrote: post #122
noxn wrote: post #121

Earlier quoted context omitted.

"Fresh Apples here! Only the best apples for 2 dollars!" - "I would like one, please." - "Thatll be 2.50, sir." - "What? I thought you just said 2?" - "Demand has just gone up."

That's wrong. It's more like: "Fresh Apples here! Only the best apples for 2 dollars!" - "I would like every single apple you have, please. Also I'm buying all the apples from the guy across the street too." - "Thatll be 2.05 each, sir." - "What? I thought you just said 2?" - "Demand has just gone up."

"Fresh Apples here! Only the best apples for 2 dollars!" "I would like 1000 apples, please." "Sir, I only have 600 apples in the shop. I can give you those now, and I'll ask the guy across the street if he has any." "How much will that cost?" "I think he's also selling them at 2 dollars each. Let me just look - oh, nope, he's seen us talking and changed his sign. Well. Have these 600 at the quoted price, and if you want more you'll have to pay the new price."
Post reply on HN