Topic: Market making

A market maker is (roughly) a trader who is willing to continuously publish a single price, and trade (buy or sell) at that price. More precisely, the bid/ask spread remains, but is very narrow, with the same specialist on both sides.

In order to address the "thin market problem" (buyers and sellers don't want to go where there isn't a market, and there isn't a market because buyers and sellers don't go there), allow corporations or individuals to buy the right to be the specialist in a particular good at that station. The specialist would post a function (E.g. piecewise linear) from their current inventory to their price - e.g. at 1k titan, my price per titan would be 1000 nic (+-the small bidask gap), but at 1m titan, my price per titan would drop to 0.75 nic (+- the small bidask gap). This would require a "construct a piecewise linear function" GUI, but my guess is that these devs have written that GUI before.

If the specialist's inventory (of either nic or the good that they're specializing in) goes to zero, then they drop out of the specialist role, and the market for that good goes back to normal.

This would move players out from alpha, offer an alternative to petty 0.01 wars for traders who do not want to spend large amounts of time watching their bids, and create more liquid markets for players who just want to buy or sell their stuff and get back to the field.

Re: Market making

Artem Blue wrote:

A market maker is (roughly) a trader who is willing to continuously publish a single price, and trade (buy or sell) at that price. More precisely, the bid/ask spread remains, but is very narrow, with the same specialist on both sides.

In order to address the "thin market problem" (buyers and sellers don't want to go where there isn't a market, and there isn't a market because buyers and sellers don't go there), allow corporations or individuals to buy the right to be the specialist in a particular good at that station. The specialist would post a function (E.g. piecewise linear) from their current inventory to their price - e.g. at 1k titan, my price per titan would be 1000 nic (+-the small bidask gap), but at 1m titan, my price per titan would drop to 0.75 nic (+- the small bidask gap). This would require a "construct a piecewise linear function" GUI, but my guess is that these devs have written that GUI before.

If the specialist's inventory (of either nic or the good that they're specializing in) goes to zero, then they drop out of the specialist role, and the market for that good goes back to normal.

This would move players out from alpha, offer an alternative to petty 0.01 wars for traders who do not want to spend large amounts of time watching their bids, and create more liquid markets for players who just want to buy or sell their stuff and get back to the field.

What's the equivalent to what you're asking for... in a real world economy?

<you're really good at building computers and your prices are reasonable... you will be the PC supplier specialist>

I am no economist but this doesn't sound like a free market to me.

Maybe I am missing your point?

3 (edited by Lupus Aurelius 2010-12-14 19:38:53)

Re: Market making

Artem Blue wrote:

I'd rather the devs worked on player-built outposts, terraforming, new bots and more equipment, more special effects and more mechanics, more assignments and more content, not more bureaucracy UI to fit a niche need.

Change of sentiment?

At this stage of the market economy, margin trading really is not that big of an issue.  As far as the .01 NIC wars, the solution is relatively simple - allow for realtime ajustment of buy and sell orders, instead of having to cancel a buy/sell order and put up a new one.  Or if you are smart, with current mechanics, don't put all you product/NIC into one order, hold some back, and put on the market as someone undercuts your orders.

In the gods we trust, all others bring data!

4 (edited by Artem Blue 2010-12-14 21:15:50)

Re: Market making

Yes, this would be more bureaucracy UI. Yes, I was trolling you before: http://forums.perpetuum-online.com/topi … -features/

Complaining that the game doesn't support mutually untrusting cooperation better, when you chose to join the most explicitly treacherous, griefer corp in the game, seemed utterly ridiculous to me.

To be honest, your idea for better permissions is decent, though I think it doesn't go far enough - why not an API for corp management, and a bridge from in-game chat to IRC? Then player corps could invent whatever complex system of permissions and checks and counterbalances they want, and code that system into their corp's IRC bot - similar to what Sam Lantinga did for WoW's UI, allowing 3rd parties to script non-game-critical aspects.

All of the ideas we're kicking around here are really "nice-to-have/someday/maybe" priority, far below any game-breaking exploits, stability/scalability, and keeping the content pipeline moving forward. We can live with the basics bin being raided once in a while (though we apparently should keep it smaller? I'm surprised Sednal thought it was worth taking to M2S, but if it has to be smaller, we keep it smaller). We can live with 0.01 price wars and relatively thin markets in the outposts, for years even, so long as the game continues getting better in the stability, scalability, and content directions.

5 (edited by Lupus Aurelius 2010-12-14 20:35:26)

Re: Market making

Artem Blue wrote:

Yes, this would be more bureaucracy UI. Yes, I was trolling you before - complaining that the game doesn't support mutually untrusting cooperation better, when you chose to join the most explicitly treacherous, griefer corp in the game, seemed utterly ridiculous to me.

What corp anyone choses to join, for whatever reason, has nothing to do with trying to help the game get better, nor the validity or logic of an arguement.  This forum is about features and requests, and I think it behooves all of us to keep this area non-political, and non-personal.  The appropriate area for the trolling is Corp Discussions, or possibly General Discussion.

As far as trusting each other, tbh, M2S members do, we have a communist/socialist structure.  But my posts in this specific forum section are concerning game mechanics, and that affects everyone who plays this game.  They have nothing to do with my corp, nor should anyone elses posts in this section, but instead should focus on providing Devs with data to consider for game improvement.

In the gods we trust, all others bring data!

Re: Market making

Sorry, you asked what in the real economy is similar - see http://en.wikipedia.org/wiki/Market_maker

NYSE has "specialists" which are slightly different, but the general concept of one entity that continuously posts simultaneous, nearby, bid and ask prices, and is willing to buy or sell at those prices to anyone (and changes their prices immediately after the exchange, because they have a different inventory) is called a market maker.

Re: Market making

Artem Blue wrote:

Sorry, you asked what in the real economy is similar - see http://en.wikipedia.org/wiki/Market_maker

NYSE has "specialists" which are slightly different, but the general concept of one entity that continuously posts simultaneous, nearby, bid and ask prices, and is willing to buy or sell at those prices to anyone (and changes their prices immediately after the exchange, because they have a different inventory) is called a market maker.


Forgive my ignorance... but isn't a large part of what the NYSE trades in is intangibles... not real goods that have real, 'reach out and touch it', value?

I just don't really see how this would be a benefit to a free market economy?

Again, I am fairly ignorant to most of this subject matter but would love to understand.

Re: Market making

Artem Blue wrote:

Sorry, you asked what in the real economy is similar - see http://en.wikipedia.org/wiki/Market_maker

NYSE has "specialists" which are slightly different, but the general concept of one entity that continuously posts simultaneous, nearby, bid and ask prices, and is willing to buy or sell at those prices to anyone (and changes their prices immediately after the exchange, because they have a different inventory) is called a market maker.

Interesting idea.

But wouldn't you think they need to bring in more Market information before a system like this would work?

Examples: Days Median, Moving Avg, day volume, High,Low, Avg. days price.

John 3:16 - Timothy 2:23

Re: Market making

Priority up to the devs, of course, and I don't have a strong opinion either way, but I can imagine a world where market information is not routinely published.

Market information is very valuable, and more and more real-world trading is moving to so-called "dark pools" of liquidity, so-called because they are unwilling to publish market information:
http://en.wikipedia.org/wiki/Dark_liquidity

This is a science-fiction game - all we need to do is to posit that the trend continues, and becomes the standard - where essentially all trading occurs without rich feeds of data from the market.

Re: Market making

I guess I still don't understand...

The way I am interpreting this is you are looking for people to be put into a position of power in the market and remain their simply because they were appointed this position of power?

Why do we need to create a game element to allow for this...why can't you do it yourself?


Am I still missing the point?

11 (edited by Gremrod 2010-12-14 21:18:54)

Re: Market making

GLiMPSE wrote:

Why do we need to create a game element to allow for this...why can't you do it yourself?

You're correct someone could control the market without this mechanic. It happens in other games.

TBH, if this is only being suggested because of the 0.01 nic bid wars. I don't think it is needed.

But if this was set up to become a game mechanic where corps can openly sell stock in their corp that would be cool.

I think more basic market features needed to be added in before something like the Market Maker idea is introduced if even at all.

John 3:16 - Timothy 2:23

Re: Market making

Artem Blue wrote:

Priority up to the devs, of course, and I don't have a strong opinion either way, but I can imagine a world where market information is not routinely published.

Market information is very valuable, and more and more real-world trading is moving to so-called "dark pools" of liquidity, so-called because they are unwilling to publish market information:
http://en.wikipedia.org/wiki/Dark_liquidity

This is a science-fiction game - all we need to do is to posit that the trend continues, and becomes the standard - where essentially all trading occurs without rich feeds of data from the market.

yeah, well the sellers and buyers are Dark in game currently. Transactions don't tell who bought from you or sold to you. I like that....

John 3:16 - Timothy 2:23

Re: Market making

The market specialist is not a position of infinite power. They're limited by the requirement to post a single price and both buy and sell at that price (plus or minus the small bid/ask gap). So if they're ripping of miners by paying ridiculously small amounts for titan ore, then they're also providing refiners a great deal. If they're ripping refiners off, then they're providing miners a great deal.

They are also required to specify how they would change that price as their inventory changes - so if someone buys 10 units, then the market specialist's inventory goes down 10 units, and presumably they've set their pricing function to go up if that happens.

This supplies liquidity (the ability to exchange immediately), which is pleasant for those of us who don't want to be playing the market all day. It is profitable for the market specialist in a couple ways. They profit by an amount equal to the bid/ask gap on every time a buy follows a sell, or a sell follows a buy. The market specialist usually gets better (more prompt, more in-depth) information by virtue of their role, which may be very valuable if they are also large players in the market - large suppliers or consumers of that good.

For someone who just wants to set things up and make a steady profit, the curve ratchets the 0.01 nic (or however many you've set it up to ratchet) for you, so that you don't have to be watching the market to adjust it. It's not quite the same because it ratchets in response to a purchase or sale, not another trader's bid, but the effect is the same - less staring at trading screens.

Re: Market making

Artem Blue wrote:

For someone who just wants to set things up and make a steady profit, the curve ratchets the 0.01 nic (or however many you've set it up to ratchet) for you, so that you don't have to be watching the market to adjust it. It's not quite the same because it ratchets in response to a purchase or sale, not another trader's bid, but the effect is the same - less staring at trading screens.

Then why play? Any good game mechanic should make the player engage in playing the game.

Isn't looking at trading screens what traders do?

Not all real world models are good for games.

John 3:16 - Timothy 2:23

Re: Market making

Gremrod wrote:
Artem Blue wrote:

For someone who just wants to set things up and make a steady profit, the curve ratchets the 0.01 nic (or however many you've set it up to ratchet) for you, so that you don't have to be watching the market to adjust it. It's not quite the same because it ratchets in response to a purchase or sale, not another trader's bid, but the effect is the same - less staring at trading screens.

Then why play? Any good game mechanic should make the player engage in playing the game.

Isn't looking at trading screens what traders do?

Not all real world models are good for games.

This would create an artificial market environment, also this would not be agile enough at times.........massive battle, 400 assault mechs pop in an hour, demand for mechs and mats go through the roof......There is no way that i would want to be chained to this system or anything like it.
There are times you could end up reverse, where you are forced to buy at a price higher than your sell price.
This would no longer be a free market, and would lose things that many traders are traders for.
Controlling the market would be similar to controlling PvP. Removing options and trying to make things "more fair" only makes it more generic and restricts creative tactics in both arenas.
Dumbing things down is a big bane in MMO's, it rewards the lazy and makes the people who want to work work more to keep ahead.

Who the $#@% is stEvE?

Re: Market making

Zyhntil, I think you've misunderstood the mechanism; this would make prices more responsive, not less, and the point is to address thin markets, not make things more fair, more generic, or dumb things down (far from it).

Currently, if a seller leaves (to sleep or to go to work), their prices stop moving. A market maker specifies what their price would be, if their inventory was different. If there is a rumor of a massive battle, and speculators buy up mats and mechs, then if the seller posted one large sell order, then the seller is disadvantaged - if they were online, they could adjust their price, but as is, they cannot. Therefore, sellers spread their orders out, selling some at one price, selling some at a higher price, some at a still higher price, trying to capture some of the value of a (possible) price excursion.

If the rumor is subsequently disproven, and speculators dump their mats and mechs back onto the market, then the price would rapidly fall. If the price fell down below the original price, then the offline trader could have buy orders waiting - but if the price doesn't fall that far, the offline trader cannot capture any of that action.

A market maker specifies a function from their inventory to a price (and the bid-ask gap is always fixed at something like 1% or 0.1% or 0.01% apart). When the rumor comes, the function doesn't know about the rumor, (AI-complete problem) - but it does react to the first trade. Inventory drops, and the function's price rises. As the crowd of speculators buy up the mats and mechs, the function raises the price further and further. Then when the rumor is disproven, the function buys those goods back, at the new (now falling) price, and makes a small profit on every trade, equal to the bid/ask gap.

Of course, if the specialist was online, they could adjust their pricing curve in response to the rumor even before the first speculator comes and executes a trade, and gain an advantage over a specialist that is offline - and if they make mistakes in such adjustments, like believing the rumor too much and keeping their prices high for too long, they can lose money. The adjustments to the price function would have to be limited by timeouts and fees, just like adjustments to bids now. This would not reduce incentives for traders to watch the real world and the markets closely, and continuously decide what the correct price ought to be, at all.

Re: Market making

Artem Blue wrote:

Zyhntil, I think you've misunderstood the mechanism; this would make prices more responsive, not less, and the point is to address thin markets, not make things more fair, more generic, or dumb things down (far from it).

Currently, if a seller leaves (to sleep or to go to work), their prices stop moving. A market maker specifies what their price would be, if their inventory was different. If there is a rumor of a massive battle, and speculators buy up mats and mechs, then if the seller posted one large sell order, then the seller is disadvantaged - if they were online, they could adjust their price, but as is, they cannot. Therefore, sellers spread their orders out, selling some at one price, selling some at a higher price, some at a still higher price, trying to capture some of the value of a (possible) price excursion.

If the rumor is subsequently disproven, and speculators dump their mats and mechs back onto the market, then the price would rapidly fall. If the price fell down below the original price, then the offline trader could have buy orders waiting - but if the price doesn't fall that far, the offline trader cannot capture any of that action.

A market maker specifies a function from their inventory to a price (and the bid-ask gap is always fixed at something like 1% or 0.1% or 0.01% apart). When the rumor comes, the function doesn't know about the rumor, (AI-complete problem) - but it does react to the first trade. Inventory drops, and the function's price rises. As the crowd of speculators buy up the mats and mechs, the function raises the price further and further. Then when the rumor is disproven, the function buys those goods back, at the new (now falling) price, and makes a small profit on every trade, equal to the bid/ask gap.

Of course, if the specialist was online, they could adjust their pricing curve in response to the rumor even before the first speculator comes and executes a trade, and gain an advantage over a specialist that is offline - and if they make mistakes in such adjustments, like believing the rumor too much and keeping their prices high for too long, they can lose money. The adjustments to the price function would have to be limited by timeouts and fees, just like adjustments to bids now. This would not reduce incentives for traders to watch the real world and the markets closely, and continuously decide what the correct price ought to be, at all.

The more I read about the idea, the more I like it.

But I am still not sure its a fit for being in a game. Seems to powerful for a passive feature.

What I like is how it could be easily extended for corps to sell shares.

John 3:16 - Timothy 2:23

18 (edited by Thierry 2010-12-15 23:20:27)

Re: Market making

Being a liquidity provider in Perpetuum is actually profitable as of now. Put EPs in finance and place buy and sell limit orders. You have to manage your inventory manually, but it's not that time consuming if you place big enough orders to start with.

Adverse selection (or the winner's curse) that you describe is something you have to live with, there would be no fun otherwise, but there are few events that make the prices jump in Perpetuum's static world.

(just to clarify, some kind of automated trading is not a bad idea, but I'm pretty sure there are far more interesting things to spend development time on first)

PS: one development that is definitely needed is better graphs, with the ability of easily read historical prices!