He's the only person who can view it, apart from any clown with a pair of polarising sunglasses. What a gimp. Better off putting the effort into something worthwhile. It's easier to just use your computer in a room with a closed door.
Trouble is, competence is not as common as you might like, and a competent person would probably get bored being a phone monkey, and move on to bigger, better things. The pay and work of answering support calls means less competent people will accumulate there.
For coding, I've not read of anyone that actually likes to work in an environment full of distractions. It sounds like maybe you don't mind it. In that case you're either very good at fitting programming problems + social interaction into your head at the same time, or you are working on some really simple problems.
Well the real world comes complete with distractions included. If you have a complex system to support and develop, you're going to need people who can deal with it. There will be support calls that need to be escalated to a developer. There will be changing priorities. There will be unreasonable customers. There will be urgent changes required on short timescales pushed on you by external parties over which you have no control. Team members will get sick, get pregnant, give birth and die and the show will have to go on.
Anyway, what about guys like me who are the only developer in the company, and just have to get on with it themselves? I don't feel that I need a team to help me figure things out. I seem to be getting on fine as-is.
Well I'd argue that you've never had to develop and support anything on the scale that I work with, or the pace of delivery required to maintain competitiveness. There's far more work than one person could handle, and even with the team at the size it is, the most productive developers (about a dozen of them) are over-stretched. But one thing's for sure, we'd be out of business if the guys and girls didn't work effectively as a team - a bunch of one-man armies just wouldn't produce the same result.
Developing as a distinct part of a group only has an advantage if others in the group actively take interest. I don't see any real truth to what you describe in my decade of professional experience so far.
The group doesn't have to actively take interest - they help in other ways anyway.
Simply having someone listen while one tries to explain a problem they're trying to solve is often a great help: having to express the problem verbally can make the solution obvious. (This would theoretically work even if the listener was inanimate, since the speaker comes to the solution on their own. Tried it with oversize teddy bears, but the developers don't go for it.)
Giving the group facilities, like a pool table and a table tennis table, will encourage them to take breaks together, and they'll inevitably talk. They may not think they're interested in what each other are working on, but ideas still cross-polinate. Sometimes offhand comments are what one needs to hear.
Some people react well to/b/tard-like abuse. With a diverse enough group, you'll end up with someone who can dish it out. The people who take it well will seek such a person out. Some people can and do raise their standards if told they're not good enough. (Don't try this as a management tactic - just keep a/b/tard or two around, and the people who need abuse will seek it out.)
After some time working together, people will identify complementary skills in each other, and get help when faced with obstacles. If the people who they seek assistance from aren't willing to spend some time helping, maybe they're not the kind of people you want around. (There's another kind of problem person - the one who's too eager to help, and in fact takes work off other developers and does it in its entirety, so the original assignee never learns.)
Seriously, if you haven't seen an effective team in a decade, you aren't working with people who are anywhere near best-of-best developers.
Most programmers seem to code much better when they are left to themselves for hours/days/weeks to just fully immerse themselves in the problem that they're trying to solve.
Bullshit - I'm a developer who's evolved into a guy who programs in between management and handling business relationships. The kinds of people who work better when left alone for extended periods of time just aren't that good. The best developers take advantage of being part of a group. Just talking about stuff with each other greatly improves what they produce. The interaction breeds better ideas, and having people to bounce ideas off, or even just banter with, keeps frustration levels down.
The act of programming itself is certainly not inherently social.
Yeah, that's true, but it neglect the fact that humans are inherently social.
The irony is a bunch of the things you listed really were the cause of serious heath problems. The negative impact of lead build-up in the body is well-known. Mercury in the ocean caused the Minamata disease. Most studies on non-Hodgkin lymphoma and leukemia show positive associations with pesticide exposure. Contraceptive pills increase the risk of venous thromboembolism. Including things where the "crazy self-promoting dickhead trying to get some publicity for himself with his stupid theory" turned out to be right doesn't support your point. Why didn't you include smoking in there as well? Hey, doctors used to endorse cigarettes after all.
Flash lets designers build interactive content. Macromedia/Adobe got/gets designers, and too many people don't realise that. You may not be interested in it, but there are some gems buried in NewGrounds along with all the crap. None of the other solutions are accessible to designer types the way Flash is. I predict one of two things will happen: Flash will die, and this kind of creative content will die with it until a new challenger appears; or more likely, Flash will just refuse to dies, and the geek elite just won't understand why.
Nah, if he was from Tasmania he's be too busy shagging his cousin to vote. You know you can marry your cousin in Tasmania. (Yes, I realise you can legally marry your cousin anywhere in Australia. Never let facts get in the way of giving Tasmanians shit, or Kiwis for that matter.)
You obviously haven't had to use both in anger. SystemTap is another "me too" project like so many things on Linux, where the only people saying it's as good are the people who haven't used the product it's an imitation of. Oh, and then there's the RMS type who will say it's "better because freedom has value" or something to that effect. Doesn't help you when you're actually trying to tune an application for performance.
Oh stop it already. The GPL is just as toxic. All these damn viral copyleft licenses are toxic. Issues like these project an image of "free software" and "open source" advocates as squabbling children. As long as this keeps up you are limiting scope for acceptance.
Yes. The colder air than usual in the stratosphere is caused by the fact that greenhouse gases insulate so much that less heat escape to space. Common sense actually. So yes, this phenomenon is a very good indication that the greenhouse effect is both real and increasing.
I am not a climate scientist. I am a former telecommunications engineer (I contributed to 802.11n standardisation process) who now works in market making (gotta follow the dollar when you have young mouths to feed). Explain to me in a way that I can understand why insulation that reduces heat escaping into space causes air in the stratosphere to be colder than usual. By my uninformed logic, I would expect air to be warmer if less heat is escaping.
Not trying to troll - just looking for an explanation that makes sense. I'm not uneducated, stupid, poor or unwilling to learn. I don't work in an industry that stands to lose if we have to cut back on carbon pollution. I'm just a curious guy who wants to know why the logic that appears to be contradicted is wrong.
As the saying goes, quantity has a quality all of its own. Its how the Soviets defeated the tank battalions of the German army - the German tanks were technologically advanced (power steering, active suspension systems etc etc - a leap ahead of other tanks of their days) but the Soviets produced their T-34s in vastly superior numbers, alongside the massive output of the US Sherman tanks...
That's not entirely fair - the T-34 was superior to German tanks in a number of ways. It was fast, difficult to spot because of its low profile, and it pioneered angled armour. It had what was possibly the best trade-off in performance characteristics of any WWII tank.
None of this is affected by the aforementioned delays, and none of it makes the gambling you describe in any way justified.
You're the one talking about predicting commodity prices in the future, aka speculation, aka gambling. Market making is not gambling. Market making is about determining fair prices for derivatives using mathematical models and quoting as close as possible to the fair price while still making enough money to run a business and pay employees (developers, QA, IT operations, analysts, traders) well enough that they don't go to the competition down the road. Delays affect how close to the fair price market makers can quote, and therefore what the transaction costs for the farmers, miners, manufacturers, etc. are when they go to take out their hedges. Delays don't affect speculators who are just out looking for a quick buck by using the market as a casino for all intents and purposes.
Sorry to burst your bubble, but one doesn't "predict future prices".
In a healthy market, one does -- in fact, that's the only thing that happens in a healthy market. The rest is bullshit.
No, commodity prices in the future are unpredictable. You don't know what world events will happen tomorrow. A war could start, a cyclone could destroy crops, a meteorite could wipe out Istanbul. All these things affect prices.
If you're growing wheat, you want some assurance that you'll be able to get half-decent money for it when it's ripe, so you hedge by buying put options (right to sell) with a half-decent strike price that expires around when the wheat's ready to sell. If the price of wheat drops below the strike price of your put options, you get to sell the wheat at the strike price, and your hedge has paid off. If the price of wheat ends up above the strike price of your option, you lost some money buying the hedge, but at least you can get decent money for your wheat on the open market.
If you're making flour, you hedge against undue increases in wheat prices by buying call options (right to buy), and the same logic applies in reverse. Same applies for other commodities - just substitute appropriate industries.
The premium that an option commands depends on the intrinsic value (difference between current price of commodity and option strike price), volatility (covering the risk that the price of the commodity will move unfavourably for the option position you're taking on), and the margins required for involved parties to make money.
I don't know anything about mathematics or finance, but this seems to assume that changing the rules would leave the premise unchanged -- i.e., that even with the delay, the prices would still swing $10 in 1 second. Is this a correct assumption?
I'm assuming that the price swings would occur at the same rate. The rate at which price swings occur depends a lot on investor mood. When the world is spinning smoothly, prices move around relatively slowly; when there's talk of another GFC or Bin Laden gets shot, investors get jumpy and prices jump around like mad. If you're expecting big price swings, you quote wider just to play it safe.
The random transaction delay won't affect the rate that prices move at. Traders would still experience the same jitters, and even though view of the base price move would be delayed with respect to when the traders input orders, the rate at which the price can change wouldn't be affected. The killer for the market maker is the delay between seeing the new price and being able to move quotes out of the way to avoid trading quotes calculated off a stale base price.
Sorry to burst your bubble, but one doesn't "predict future prices". The future contract (i.e. contract to deliver at some future date) has a price, based on what people are currently prepared to bid/offer on it. The price of this contract is based on the value of the commodity. Options have intrinsic value from the difference between strike and current value of commodity or future, and implied volatility value. Market makers are the parties that dampen price swings and suck volatility out of the market, by quoting options based on fair price.
The "day traders" you speak of are speculators who want to gamble. They will never go away. In fact, they are far more prevalent in countries that have bans on casinos, like Korea and Hong Kong. There's a reason the KOSPI200 is the most actively traded index product in the world. You know the saying about a fool and their money? That's what happens to these speculators. They win less than half the time, because all their money gets eaten up in transaction costs.
Farmers, miners, manufacturers, investors etc. have needs to hedge against commodity price moves so they can plan activities that have long lead times. They do this by buying or writing options. They are the people who have legitimate reasons to want liquidity in the market. They are the people who would be screwed over by the wider quote spreads that would be necessitated if exchanges introduced random delays in transaction processing.
As an aside, some exchanges are implemented so poorly that they may as well have random delays in them. You don't see people flocking there for a better deal.
No, the price will be less fair, because the spreads will be wider. There will be more market makers (middlemen), because the potential profit on each trade will be higher. It won't cut the number of speculators - they just love to gamble. It won't cut demand from people with "legitimate" reasons to trade (the wheat farmer and exporter in my example), either, but they will be screwed over by the wider spreads.
You, sir, are absolutely correct. I must be too tired - good thing I'm not trading right now, or I'm be hitting myself or quoting below theo at this rate.
Adding delay will actually make investors worse off, because quoting will become less competitive. Let me explain this with a contrived and simplified example.
Let's say I'm a market maker quoting a derivative, a call option on wheat futures for example. I decide what I think the option is worth based on the current price of the future and my guess at volatility, and we come up with a fair price. Let's say that the fair price is $50. The fair price for the option will move when the price of the underlying contract moves. The proportion by which it moves is called delta. Let's say this option has delta of +0.5, so if the price of the future changes by $1, the fair price of the option moves by $0.50.
In order to make some money quoting it, I need to quote a spread - i.e. buy options for less than I sell them for. Let's say I want to quote a spread of 2% of the fair price, or $1 in this case. I drop our bid in at $49.50 and our offer at $50.50. You, as a wheat farmer or exporter, want to hedge yourself against fluctuations in wheat prices, so you're interested in trading these options. When I'm quoting a 2% spread, you can buy or sell options with a transaction cost of 1% of the fair price of the option, or $0.50 on a $50 option. That's not too bad.
But remember that pesky concept of delta? If the price of the wheat futures moves around, I need to move my quotes on the option. For example if the price of the future increases by $2, I need to move by quotes up by $1 on the delta +0.5 option, So if that were to happen, I'd be quoting at $50.50 bid and $51.50 offer - note that I'm still only taking a premium of about 1% of the fair price.
Hopefully you can see that if the price of the future moves around, I need to be able to keep up with it or I'll be screwed over when I try to hedge my options position. If the price of the future moves faster than I can move my quotes, I need to factor a safety margin into the spread I quote to cater for this.
Suppose you introduce a random delay of up to one second. That means I have to consider the worst case scenario. Maybe I think the price of the future might change by up to $10 in one second. Since this is a delta +0.5 option, I need to factor in a risk of a half of $10, or $5, into the spread I'm quoting, because the price of the future could move by that much before I can move my quotes.
So factoring in the $5 base move risk as well as my 2% spread that I'm trying to make money off, I'd be quoting $44.50 bid and $45.50 offer. Now your transaction cost has increased to $5.50 over the fair price per trade on the option, or 11%. It's not looking so attractive now, is it?
Introducing delays won't hurt me as a market marker - I'll just increase my spreads to cover the risk, as will all the other market makers. It will definitely harm you as the person with a need to trade. Lower transaction latencies increases competition between market makers to quote tiny spreads, minimising the transaction costs for people who need to trade. Sure, the money is being distributed differently: instead of more market makers, each with a small slice of the pie, taking a big cut of each transaction, you now have fewer market makers taking a tiny cut of each transaction, competing to get a big enough slice of the pie to remain profitable.
Slashdot Mirror
He's the only person who can view it, apart from any clown with a pair of polarising sunglasses. What a gimp. Better off putting the effort into something worthwhile. It's easier to just use your computer in a room with a closed door.
Trouble is, competence is not as common as you might like, and a competent person would probably get bored being a phone monkey, and move on to bigger, better things. The pay and work of answering support calls means less competent people will accumulate there.
Well the real world comes complete with distractions included. If you have a complex system to support and develop, you're going to need people who can deal with it. There will be support calls that need to be escalated to a developer. There will be changing priorities. There will be unreasonable customers. There will be urgent changes required on short timescales pushed on you by external parties over which you have no control. Team members will get sick, get pregnant, give birth and die and the show will have to go on.
Well I'd argue that you've never had to develop and support anything on the scale that I work with, or the pace of delivery required to maintain competitiveness. There's far more work than one person could handle, and even with the team at the size it is, the most productive developers (about a dozen of them) are over-stretched. But one thing's for sure, we'd be out of business if the guys and girls didn't work effectively as a team - a bunch of one-man armies just wouldn't produce the same result.
The group doesn't have to actively take interest - they help in other ways anyway.
Simply having someone listen while one tries to explain a problem they're trying to solve is often a great help: having to express the problem verbally can make the solution obvious. (This would theoretically work even if the listener was inanimate, since the speaker comes to the solution on their own. Tried it with oversize teddy bears, but the developers don't go for it.)
Giving the group facilities, like a pool table and a table tennis table, will encourage them to take breaks together, and they'll inevitably talk. They may not think they're interested in what each other are working on, but ideas still cross-polinate. Sometimes offhand comments are what one needs to hear.
Some people react well to /b/tard-like abuse. With a diverse enough group, you'll end up with someone who can dish it out. The people who take it well will seek such a person out. Some people can and do raise their standards if told they're not good enough. (Don't try this as a management tactic - just keep a /b/tard or two around, and the people who need abuse will seek it out.)
After some time working together, people will identify complementary skills in each other, and get help when faced with obstacles. If the people who they seek assistance from aren't willing to spend some time helping, maybe they're not the kind of people you want around. (There's another kind of problem person - the one who's too eager to help, and in fact takes work off other developers and does it in its entirety, so the original assignee never learns.)
Seriously, if you haven't seen an effective team in a decade, you aren't working with people who are anywhere near best-of-best developers.
Bullshit - I'm a developer who's evolved into a guy who programs in between management and handling business relationships. The kinds of people who work better when left alone for extended periods of time just aren't that good. The best developers take advantage of being part of a group. Just talking about stuff with each other greatly improves what they produce. The interaction breeds better ideas, and having people to bounce ideas off, or even just banter with, keeps frustration levels down.
Yeah, that's true, but it neglect the fact that humans are inherently social.
The irony is a bunch of the things you listed really were the cause of serious heath problems. The negative impact of lead build-up in the body is well-known. Mercury in the ocean caused the Minamata disease. Most studies on non-Hodgkin lymphoma and leukemia show positive associations with pesticide exposure. Contraceptive pills increase the risk of venous thromboembolism. Including things where the "crazy self-promoting dickhead trying to get some publicity for himself with his stupid theory" turned out to be right doesn't support your point. Why didn't you include smoking in there as well? Hey, doctors used to endorse cigarettes after all.
If you are what you eat, does that mean I was a wet pussy last night?
Flash lets designers build interactive content. Macromedia/Adobe got/gets designers, and too many people don't realise that. You may not be interested in it, but there are some gems buried in NewGrounds along with all the crap. None of the other solutions are accessible to designer types the way Flash is. I predict one of two things will happen: Flash will die, and this kind of creative content will die with it until a new challenger appears; or more likely, Flash will just refuse to dies, and the geek elite just won't understand why.
Isn't taswegia just evidence that kiwis can swim? (To answer the question, I meant "giving shit to taswegieans and kiwis.")
Nah, if he was from Tasmania he's be too busy shagging his cousin to vote. You know you can marry your cousin in Tasmania. (Yes, I realise you can legally marry your cousin anywhere in Australia. Never let facts get in the way of giving Tasmanians shit, or Kiwis for that matter.)
You obviously haven't had to use both in anger. SystemTap is another "me too" project like so many things on Linux, where the only people saying it's as good are the people who haven't used the product it's an imitation of. Oh, and then there's the RMS type who will say it's "better because freedom has value" or something to that effect. Doesn't help you when you're actually trying to tune an application for performance.
Oh stop it already. The GPL is just as toxic. All these damn viral copyleft licenses are toxic. Issues like these project an image of "free software" and "open source" advocates as squabbling children. As long as this keeps up you are limiting scope for acceptance.
I think we have a winner, and thankfully it wasn't a car analogy!
I am not a climate scientist. I am a former telecommunications engineer (I contributed to 802.11n standardisation process) who now works in market making (gotta follow the dollar when you have young mouths to feed). Explain to me in a way that I can understand why insulation that reduces heat escaping into space causes air in the stratosphere to be colder than usual. By my uninformed logic, I would expect air to be warmer if less heat is escaping.
Not trying to troll - just looking for an explanation that makes sense. I'm not uneducated, stupid, poor or unwilling to learn. I don't work in an industry that stands to lose if we have to cut back on carbon pollution. I'm just a curious guy who wants to know why the logic that appears to be contradicted is wrong.
That's not entirely fair - the T-34 was superior to German tanks in a number of ways. It was fast, difficult to spot because of its low profile, and it pioneered angled armour. It had what was possibly the best trade-off in performance characteristics of any WWII tank.
I can't remember the last time I charged my phone over the Bluetooth port. Oh wait, it doesn't charge over Bluetooth.
420 4eva bra! let's get baked!
You're the one talking about predicting commodity prices in the future, aka speculation, aka gambling. Market making is not gambling. Market making is about determining fair prices for derivatives using mathematical models and quoting as close as possible to the fair price while still making enough money to run a business and pay employees (developers, QA, IT operations, analysts, traders) well enough that they don't go to the competition down the road. Delays affect how close to the fair price market makers can quote, and therefore what the transaction costs for the farmers, miners, manufacturers, etc. are when they go to take out their hedges. Delays don't affect speculators who are just out looking for a quick buck by using the market as a casino for all intents and purposes.
No, commodity prices in the future are unpredictable. You don't know what world events will happen tomorrow. A war could start, a cyclone could destroy crops, a meteorite could wipe out Istanbul. All these things affect prices.
If you're growing wheat, you want some assurance that you'll be able to get half-decent money for it when it's ripe, so you hedge by buying put options (right to sell) with a half-decent strike price that expires around when the wheat's ready to sell. If the price of wheat drops below the strike price of your put options, you get to sell the wheat at the strike price, and your hedge has paid off. If the price of wheat ends up above the strike price of your option, you lost some money buying the hedge, but at least you can get decent money for your wheat on the open market.
If you're making flour, you hedge against undue increases in wheat prices by buying call options (right to buy), and the same logic applies in reverse. Same applies for other commodities - just substitute appropriate industries.
The premium that an option commands depends on the intrinsic value (difference between current price of commodity and option strike price), volatility (covering the risk that the price of the commodity will move unfavourably for the option position you're taking on), and the margins required for involved parties to make money.
I'm assuming that the price swings would occur at the same rate. The rate at which price swings occur depends a lot on investor mood. When the world is spinning smoothly, prices move around relatively slowly; when there's talk of another GFC or Bin Laden gets shot, investors get jumpy and prices jump around like mad. If you're expecting big price swings, you quote wider just to play it safe.
The random transaction delay won't affect the rate that prices move at. Traders would still experience the same jitters, and even though view of the base price move would be delayed with respect to when the traders input orders, the rate at which the price can change wouldn't be affected. The killer for the market maker is the delay between seeing the new price and being able to move quotes out of the way to avoid trading quotes calculated off a stale base price.
Sorry to burst your bubble, but one doesn't "predict future prices". The future contract (i.e. contract to deliver at some future date) has a price, based on what people are currently prepared to bid/offer on it. The price of this contract is based on the value of the commodity. Options have intrinsic value from the difference between strike and current value of commodity or future, and implied volatility value. Market makers are the parties that dampen price swings and suck volatility out of the market, by quoting options based on fair price.
The "day traders" you speak of are speculators who want to gamble. They will never go away. In fact, they are far more prevalent in countries that have bans on casinos, like Korea and Hong Kong. There's a reason the KOSPI200 is the most actively traded index product in the world. You know the saying about a fool and their money? That's what happens to these speculators. They win less than half the time, because all their money gets eaten up in transaction costs.
Farmers, miners, manufacturers, investors etc. have needs to hedge against commodity price moves so they can plan activities that have long lead times. They do this by buying or writing options. They are the people who have legitimate reasons to want liquidity in the market. They are the people who would be screwed over by the wider quote spreads that would be necessitated if exchanges introduced random delays in transaction processing.
As an aside, some exchanges are implemented so poorly that they may as well have random delays in them. You don't see people flocking there for a better deal.
No, the price will be less fair, because the spreads will be wider. There will be more market makers (middlemen), because the potential profit on each trade will be higher. It won't cut the number of speculators - they just love to gamble. It won't cut demand from people with "legitimate" reasons to trade (the wheat farmer and exporter in my example), either, but they will be screwed over by the wider spreads.
My head hurts too, so that makes too of us. Rough day at work. Sure could do with a blowjob right now.
You, sir, are absolutely correct. I must be too tired - good thing I'm not trading right now, or I'm be hitting myself or quoting below theo at this rate.
Adding delay will actually make investors worse off, because quoting will become less competitive. Let me explain this with a contrived and simplified example.
Let's say I'm a market maker quoting a derivative, a call option on wheat futures for example. I decide what I think the option is worth based on the current price of the future and my guess at volatility, and we come up with a fair price. Let's say that the fair price is $50. The fair price for the option will move when the price of the underlying contract moves. The proportion by which it moves is called delta. Let's say this option has delta of +0.5, so if the price of the future changes by $1, the fair price of the option moves by $0.50.
In order to make some money quoting it, I need to quote a spread - i.e. buy options for less than I sell them for. Let's say I want to quote a spread of 2% of the fair price, or $1 in this case. I drop our bid in at $49.50 and our offer at $50.50. You, as a wheat farmer or exporter, want to hedge yourself against fluctuations in wheat prices, so you're interested in trading these options. When I'm quoting a 2% spread, you can buy or sell options with a transaction cost of 1% of the fair price of the option, or $0.50 on a $50 option. That's not too bad.
But remember that pesky concept of delta? If the price of the wheat futures moves around, I need to move my quotes on the option. For example if the price of the future increases by $2, I need to move by quotes up by $1 on the delta +0.5 option, So if that were to happen, I'd be quoting at $50.50 bid and $51.50 offer - note that I'm still only taking a premium of about 1% of the fair price.
Hopefully you can see that if the price of the future moves around, I need to be able to keep up with it or I'll be screwed over when I try to hedge my options position. If the price of the future moves faster than I can move my quotes, I need to factor a safety margin into the spread I quote to cater for this.
Suppose you introduce a random delay of up to one second. That means I have to consider the worst case scenario. Maybe I think the price of the future might change by up to $10 in one second. Since this is a delta +0.5 option, I need to factor in a risk of a half of $10, or $5, into the spread I'm quoting, because the price of the future could move by that much before I can move my quotes.
So factoring in the $5 base move risk as well as my 2% spread that I'm trying to make money off, I'd be quoting $44.50 bid and $45.50 offer. Now your transaction cost has increased to $5.50 over the fair price per trade on the option, or 11%. It's not looking so attractive now, is it?
Introducing delays won't hurt me as a market marker - I'll just increase my spreads to cover the risk, as will all the other market makers. It will definitely harm you as the person with a need to trade. Lower transaction latencies increases competition between market makers to quote tiny spreads, minimising the transaction costs for people who need to trade. Sure, the money is being distributed differently: instead of more market makers, each with a small slice of the pie, taking a big cut of each transaction, you now have fewer market makers taking a tiny cut of each transaction, competing to get a big enough slice of the pie to remain profitable.