Author Topic: Let's Lower Trading Fee  (Read 10153 times)

0 Members and 1 Guest are viewing this topic.

Offline tonyk

  • Hero Member
  • *****
  • Posts: 3308
    • View Profile
You better think over the partially filled order rules...

partially filled 1BTS worth of an order and  0.10  usd fee does not seem right.
Lack of arbitrage is the problem, isn't it. And this 'should' solves it.

Offline Empirical1.2

  • Hero Member
  • *****
  • Posts: 1366
    • View Profile
I think sharing this comment is good for discussion.

https://github.com/cryptonomex/graphene/issues/393
Quote
Market participants expect placing / moving orders to be cheap and filling orders to be expensive.

To prevent spam, all transactions require some kind of fee.
Placing orders consumes memory and cannot be trivially cheap so needs a high fee.
Canceling orders frees memory and has historically been free.

We can make placing orders free, filling orders (relatively) expensive, and canceling orders cheap with the following change.

User pays $0.25 to place an order
User pays $0.001 to cancel an order, and gets a refund of $0.25 * NETWORK_FEE_PERCENT
If the order fills then there is no refund of the $0.25
Under this plan the cost to place / update an order is $0.001 while the cost of having an order filled is $0.25.

The only remaining issue with the above is that part of the original fee ($0.20) goes to the referrer who may have withdrawn their cashback. To keep things simple, the only fee that gets refunded is the network fee.

If the user is a lifetime member then it looks like this:

User pays $0.25 to place an order and receives $0.20 cash back
User pays $0.001 to cancel an order, and gets a refund of $0.05 and $0.0008 cash back. (total cost $0.0002)
If the order is actually filled the user paid $0.05
If the user is an annual subscriber then it looks like this:

User pays $0.25 to place an order and receives $0.125 cash back
User pays $0.001 to cancel an order and receives $0.0005 cash back and a refund of $0.05 (total cost $0.0755)
If the order is actually filled the user paid $0.125
If the user is a basic member then it looks like this:

User pays $0.25 to place an order
User pays $0.001 to cancel an order and gets $0.05 back (total cost $0.20)
If the order is actually filled the user pays $0.25
The break even point for deciding whether to upgrade to a lifetime member is if the user would end up canceling 500 orders. If they end up canceling 10000 orders because they are a market maker then the average cost would be ($100 + 50000*.0002)/50000 => 0.0022 and the total spent would have been $110 dollars.

As long as the average order size is greater than $25 dollars, the market fees on the filled orders will be less than other exchanges that charge (0.2%) of volume. Anyone who does $55,000 of volume will pay over $110 in fees on centralized exchanges. If you are placing and canceling thousands of orders with a bot and doing less than $55,000 of volume then chances are your bot is not effective and is spamming the network.

I totally agree with bytemaster, but think the baseline (opening order costs $0.20 for basic member) is somewhat expensive. I would suggest $0.10 instead of $0.25 as a total trading fee. This increase the break-even point from 500 orders to 1250 orders, but will have less barriers against new users.

Yeah it looked good but still pretty expensive for a basic member. Unless their individual trade is > $250 it's cheaper to use a 0.1% exchange and if they might cancel their order once or twice like most of us then it's probably more like BTS is only cheaper if you're trade size is > $500.  So it would be nice if it could be made still cheaper for a basic member.

Also most trading sites reward people for high volume without charging upfront. Poker sites use it too. So if possible I would suggest automatically upgrading accounts to annual and lifetime once the network has earned say double from them than what the memberships costs.

So it's still cheaper to buy the memberships but if you don't and end up doing enough business we'll reward you with those cheaper rates as a thank you and to encourage your future business.



If you want to take the island burn the boats

Offline clayop

  • Hero Member
  • *****
  • Posts: 2033
    • View Profile
    • Bitshares Korea
  • BitShares: clayop
I think sharing this comment is good for discussion.

https://github.com/cryptonomex/graphene/issues/393
Quote
Market participants expect placing / moving orders to be cheap and filling orders to be expensive.

To prevent spam, all transactions require some kind of fee.
Placing orders consumes memory and cannot be trivially cheap so needs a high fee.
Canceling orders frees memory and has historically been free.

We can make placing orders free, filling orders (relatively) expensive, and canceling orders cheap with the following change.

User pays $0.25 to place an order
User pays $0.001 to cancel an order, and gets a refund of $0.25 * NETWORK_FEE_PERCENT
If the order fills then there is no refund of the $0.25
Under this plan the cost to place / update an order is $0.001 while the cost of having an order filled is $0.25.

The only remaining issue with the above is that part of the original fee ($0.20) goes to the referrer who may have withdrawn their cashback. To keep things simple, the only fee that gets refunded is the network fee.

If the user is a lifetime member then it looks like this:

User pays $0.25 to place an order and receives $0.20 cash back
User pays $0.001 to cancel an order, and gets a refund of $0.05 and $0.0008 cash back. (total cost $0.0002)
If the order is actually filled the user paid $0.05
If the user is an annual subscriber then it looks like this:

User pays $0.25 to place an order and receives $0.125 cash back
User pays $0.001 to cancel an order and receives $0.0005 cash back and a refund of $0.05 (total cost $0.0755)
If the order is actually filled the user paid $0.125
If the user is a basic member then it looks like this:

User pays $0.25 to place an order
User pays $0.001 to cancel an order and gets $0.05 back (total cost $0.20)
If the order is actually filled the user pays $0.25
The break even point for deciding whether to upgrade to a lifetime member is if the user would end up canceling 500 orders. If they end up canceling 10000 orders because they are a market maker then the average cost would be ($100 + 50000*.0002)/50000 => 0.0022 and the total spent would have been $110 dollars.

As long as the average order size is greater than $25 dollars, the market fees on the filled orders will be less than other exchanges that charge (0.2%) of volume. Anyone who does $55,000 of volume will pay over $110 in fees on centralized exchanges. If you are placing and canceling thousands of orders with a bot and doing less than $55,000 of volume then chances are your bot is not effective and is spamming the network.

I totally agree with bytemaster, but think the baseline (opening order costs $0.20 for basic member) is somewhat expensive. I would suggest $0.10 instead of $0.25 as a total trading fee. This increase the break-even point from 500 orders to 1250 orders, but will have less barriers against new users.
Bitshares Korea - http://www.bitshares.kr
Vote for me and see Korean Bitshares community grows
delegate-clayop

Offline Musewhale

  • Hero Member
  • *****
  • Posts: 2881
  • 丑,实在是太丑了 !
    • View Profile
I think BM's solution proposed in today's mumble session is satisfying both sides.
Keep or increase current trading fee ($0.03~0.05) but refund them when the order is cancelled. I think we should charge a small amount of fee for cancellation (1~5%) to prevent spam.

+1 +1 +1
MUSE witness:mygoodfriend     vote for me

Offline clayop

  • Hero Member
  • *****
  • Posts: 2033
    • View Profile
    • Bitshares Korea
  • BitShares: clayop
Bitshares Korea - http://www.bitshares.kr
Vote for me and see Korean Bitshares community grows
delegate-clayop

Offline Empirical1.2

  • Hero Member
  • *****
  • Posts: 1366
    • View Profile
Merchants are used to pay 35 cents + a % then charge them a touch less and their mind can handle it.
Traders are used to placing orders for free and paying big on filled trades, then approximate that closely.
Users are not used to seeing fees, but merchants are, display the fees to the receiver and not the sender.

  +5% Attempting to understand the needs & expectations of different users and adjusting the fee structure accordingly is a great improvement over this...

This is mind blowing.  You can't judge the fees without the referral system working for several months. No other coin has a mechanism to pay merchants to go to the trouble of educating their user base to pay them in crypto.  We do.  That is unique and it is a chance (not a guarantee) that it we succeed where all other alts have failed.

Geo-pricing is just another level of confusion to add to the worlds most complicated product.  Lets not shall we.

Glad to see it.
If you want to take the island burn the boats

Offline clayop

  • Hero Member
  • *****
  • Posts: 2033
    • View Profile
    • Bitshares Korea
  • BitShares: clayop
I think BM's solution proposed in today's mumble session is satisfying both sides.
Keep or increase current trading fee ($0.03~0.05) but refund them when the order is cancelled. I think we should charge a small amount of fee for cancellation (1~5%) to prevent spam.
Bitshares Korea - http://www.bitshares.kr
Vote for me and see Korean Bitshares community grows
delegate-clayop

Offline luckybit

  • Hero Member
  • *****
  • Posts: 2921
    • View Profile
  • BitShares: Luckybit
Opinions from our trader (Ander). I support him.

https://bitsharestalk.org/index.php/topic,19394.msg248989.html#msg248989

Opinion frpm Bytemaster: https://bitsharestalk.org/index.php/topic,19394.msg249000.html#msg249000


By the way Max makes good sense on psychology but Bytemaster makes a strong case on resource scarcity. We have multiple constraints to consider in our design.

Max, a recent EU court ruling just said that crypto traders can trade tax free. No VAT tax. This would mean traders in the EU might be willing to pay higher fees and have the incentive to trade because it's tax free. If they can trade virtual stocks tax free that is a very strong selling point.

http://cointelegraph.com/news/115447/breaking-eu-court-rules-bitcoin-exchange-is-tax-free
« Last Edit: October 23, 2015, 04:41:58 am by luckybit »
https://metaexchange.info | Bitcoin<->Altcoin exchange | Instant | Safe | Low spreads

Offline mike623317

  • Hero Member
  • *****
  • Posts: 637
    • View Profile
Good point max.

Offline Stan

  • Hero Member
  • *****
  • Posts: 2908
  • You need to think BIGGER, Pinky...
    • View Profile
    • Cryptonomex
  • BitShares: Stan
Bitshares is a very complicated product.  the more places you can be exactly like what people are used to the better.

Merchants are used to pay 35 cents + a % then charge them a touch less and their mind can handle it.
Traders are used to placing orders for free and paying big on filled trades, then approximate that closely.
Users are not used to seeing fees, but merchants are, display the fees to the receiver and not the sender.

All of these step and many others reduce the barriers to entry for new people.  and not just by a little but by a lot.  When these things are not as the user expected, they begin asking questions... It is a rabbit hole that sucks balls for the sales rep trying to sell it.

As a sales person all i want to be able to say is to a merchant.
 So you hate banks and credit cards Mr Merchant?
 Yes.
Would you like to be able to cut them and keep the money you used to pay them?
Yes.
What if told you that not only could you keep the merchant fees you used to pay at this store, but if you introduce this to your customers and they use this technology at other stores you would get the merchant fees that merchant used to pay for the life of that customer forever and ever. In effect you become Visa, would that be exciting?
Yes.
Well here is what I need you to do. I need you to invest about 20 minutes learning how this works. Then we can go ahead and get you an account and you will be set up to receive payments can we do that now?
Yes.
In the next 20 minutes I explain the more I have to explain the more chance I lose the sale.  Make everything exactly what he is used to. And where it is different make it simpler.

To a trader.
Who do you trade with now?
XYZ.
I know a lot of traders and they many are doing quite well.  However I know many of them never quite feel safe after MtGox.
That is a risk we take, but its working out for now.
IF I could show you a new trading technology that looked, felt, and was similar to what you are using except it had no counterparty risk would you be interested?
Yes.
As it is a new technology the liquidity is still growing, but on the upside an experienced trader can fleece a lot of rookies there too.  but on the upside. they have a very cool Referal system and you can earn about $100 for every person you introduce to the platform, so I suspect it will grow pretty quick.  Do you want to take a look.
For sure.

Do either of these conversation sound remotely similar to anything any of us has experienced when selling bitshares?  I bet the answer is no.  And the reason why is it is so complicated.  Forget the crypto world. Lets look at the masses.  Give them what they are used to but a bit better an we will crush it.

Time to put in a worker proposal with bitmarket as Project Manager to get the necessary changes made?

Anything said on these forums does not constitute an intent to create a legal obligation or contract of any kind.   These are merely my opinions which I reserve the right to change at any time.

Offline bitmarket

  • Sr. Member
  • ****
  • Posts: 369
    • View Profile
    • BitShares TV
Bitshares is a very complicated product.  the more places you can be exactly like what people are used to the better.

Merchants are used to pay 35 cents + a % then charge them a touch less and their mind can handle it.
Traders are used to placing orders for free and paying big on filled trades, then approximate that closely.
Users are not used to seeing fees, but merchants are, display the fees to the receiver and not the sender.

All of these step and many others reduce the barriers to entry for new people.  and not just by a little but by a lot.  When these things are not as the user expected, they begin asking questions... It is a rabbit hole that sucks balls for the sales rep trying to sell it.

As a sales person all i want to be able to say is to a merchant.
 So you hate banks and credit cards Mr Merchant?
 Yes.
Would you like to be able to cut them and keep the money you used to pay them?
Yes.
What if told you that not only could you keep the merchant fees you used to pay at this store, but if you introduce this to your customers and they use this technology at other stores you would get the merchant fees that merchant used to pay for the life of that customer forever and ever. In effect you become Visa, would that be exciting?
Yes.
Well here is what I need you to do. I need you to invest about 20 minutes learning how this works. Then we can go ahead and get you an account and you will be set up to receive payments can we do that now?
Yes.
In the next 20 minutes I explain the more I have to explain the more chance I lose the sale.  Make everything exactly what he is used to. And where it is different make it simpler.

To a trader.
Who do you trade with now?
XYZ.
I know a lot of traders and they many are doing quite well.  However I know many of them never quite feel safe after MtGox.
That is a risk we take, but its working out for now.
IF I could show you a new trading technology that looked, felt, and was similar to what you are using except it had no counterparty risk would you be interested?
Yes.
As it is a new technology the liquidity is still growing, but on the upside an experienced trader can fleece a lot of rookies there too.  but on the upside. they have a very cool Referal system and you can earn about $100 for every person you introduce to the platform, so I suspect it will grow pretty quick.  Do you want to take a look.
For sure.

Do either of these conversation sound remotely similar to anything any of us has experienced when selling bitshares?  I bet the answer is no.  And the reason why is it is so complicated.  Forget the crypto world. Lets look at the masses.  Give them what they are used to but a bit better an we will crush it.

Host of BitShares.TV and Author of BitShares 101

Offline luckybit

  • Hero Member
  • *****
  • Posts: 2921
    • View Profile
  • BitShares: Luckybit
My approach is to attract more users first with lower fee and make profit later.

Like Microsoft... they intentionally leaved their product "unprotected" from piracy, the majority of user's had windows for free... And look them now!

Maybe we should copy Microsoft and put ads into the interface. I'm all for it.

http://betanews.com/2015/10/16/advertisements-in-the-windows-10-start-menu-are-a-good-thing/
https://metaexchange.info | Bitcoin<->Altcoin exchange | Instant | Safe | Low spreads

Offline clayop

  • Hero Member
  • *****
  • Posts: 2033
    • View Profile
    • Bitshares Korea
  • BitShares: clayop
Bitshares Korea - http://www.bitshares.kr
Vote for me and see Korean Bitshares community grows
delegate-clayop

Offline luckybit

  • Hero Member
  • *****
  • Posts: 2921
    • View Profile
  • BitShares: Luckybit
Here is an idea. Why not track the amount of revenue being generated per day by the fees and have the fees be somewhat dynamic based on the profitability and revenue of the DAC?

Good idea. But we should talk about the starting point. Remember that the current trading fee is based on the dev teams arbitrary decision. We should listen to many traders and bot makers as well.
I basically agree with you that we must have profitability. My approach is to attract more users first with lower fee and make profit later.

Our service is obviously superior that centralized exchanges. But what traders really care is money (profit surely), which is a function of liquidity and trading costs.

1)  There might not be a later. The DAC should take profit whenever profit is available and should not wait until a later which might not come.

We need to determine a metric or set of metrics which can track. We can track profitability from these metrics. For example the daily burn rate would tell us something, top 10 most popular transaction types would tell us something, the most popular asset of the day with the amount of revenue it generated for the network would tell us something. Without these statistics we don't have quality data from which to make decisions and are literally shooting in the dark. So if you want the most effective possible fee structure you need the highest quality data possible so as to produce and over time optimize the fee structure.


2) Traders should be able to profit but should not profit at the expense of the Bitshares DAC. The relationship between traders and the DAC has to be mutually beneficial, symbiotic. The DAC also has to profit and maybe the DAC can attract people by giving people ways to make money by doing stuff other than trading. The referral program is a critical example of this and many of the people making money from referrals will be the people who will now have money to trade with, and I doubt they'd complain about the fees.

Professional crypto-traders aren't the heart of an economy. They simply make money through trading money they already have. Make ways for people to earn fee vouchers, rewards, points, credits, etc. Let the exchanges have token lotteries, be creative about how people can make a profit other than trading, and the fees no longer are an issue. The fees only are an issue to people who want to make bots but who don't want to do anything other than trade, which is kind of strange because if they have a lot of money to start with then the fees aren't a big deal and lifetime membership is cheap, but if they don't then they must want to trade fractions of a penny?

The fee should adjust dynamically. For example it may be possible to use a genetic algorithm to have the fees adapt to the usage patterns of the participants. The dynamic fee structure could then be trained from the statistics and evolve to fit the demographics without violating any privacy. As long as the statistics are anonymous, you could overlay this on top of Bitshares eventually to all who choose to participate in providing the statistics, which would be information like how many of a certain kind of trades are being done, or which kind of transactions generate the most or least revenue for the DAC.

Features which are valued a lot by participants could train the genetic algorithm but of course this would take a bit of effort to implement. There are simpler ways to create the dynamically adjusting smart fee mechanism.


An easier version would be to simply use a metric, and to adjust the fee on a daily basis. On hours where profitability is high for the DAC the fees could dynamically adjust. Bytemaster may be onto something with that but in general if the fee difficulty adjusts according to the actual usage patterns of Bitshares then you'd have real time adjustable fees. During the times of great profitability the bots could jump in on the low fee hours, but during times of low profitability, the fees should increase just as the difficulty in Bitcoin increases when a lot of people mine it.

The concept of difficulty applied to fees might be a way to make a sort of dynamic smart fee mechanism so that fees adjust themselves according to usage. Usage statistics could be added later so that in the backend the Bitshares DAC could know exactly how we like to use it, and then evolve.

https://metaexchange.info | Bitcoin<->Altcoin exchange | Instant | Safe | Low spreads

Offline clayop

  • Hero Member
  • *****
  • Posts: 2033
    • View Profile
    • Bitshares Korea
  • BitShares: clayop
My approach is to attract more users first with lower fee and make profit later.

Like Microsoft... they intentionally leaved their product "unprotected" from piracy, the majority of user's had windows for free... And look them now!

Yes.... Be evil LOL
Bitshares Korea - http://www.bitshares.kr
Vote for me and see Korean Bitshares community grows
delegate-clayop