These forums have been archived and are now read-only.

The new forums are live and can be found at https://forums.eveonline.com/

Market Discussions

 
  • Topic is locked indefinitely.
 

Banks. We need them.

Author
Jake Andarius
Andarius Trading Corp.
#221 - 2012-03-20 03:53:19 UTC
I just remembered one other thing that I had been wondering earlier in regards to the main topic of the feasibility of a bank in EVE. How difficult is it from a management perspective to actually run a bank that gets tens to hundreds of billions in deposits, assuming a minimum deposit of one billion (or other numbers depending on player's knowledge)? How many different people are really required to run an operation of that size? And how much time do they have to spend managing their specific aspect of the bank? It seems like this management question is very pertinent to the overall discussion topic.
Darth Tickles
Doomheim
#222 - 2012-03-20 03:55:31 UTC
Jake Andarius wrote:
I just remembered one other thing that I had been wondering earlier in regards to the main topic of the feasibility of a bank in EVE. How difficult is it from a management perspective to actually run a bank that gets tens to hundreds of billions in deposits, assuming a minimum deposit of one billion (or other numbers depending on player's knowledge)? How many different people are really required to run an operation of that size? And how much time do they have to spend managing their specific aspect of the bank? It seems like this management question is very pertinent to the overall discussion topic.


Imagine it for yourself. "Innovate" how you would deal with deposits.
KrakizBad
Section 8.
#223 - 2012-03-20 04:12:13 UTC
Jake Andarius wrote:
KrakizBad wrote:
This is less "trust isn't dead" than "there are always new suckers."


I am confused about how that distinction is being made. If there are hundreds of "suckers" willing to invest billions of ISK in an EVE bank because they believe it will succeed, then trust is not dead. In the context of EVE banking, I think trust is dead when no banking venture can pull in even a few hundred billion anymore because there is enough cultural awareness about banks to just shut down investments. If Phaser is any indication, that simple does not seem to have happened yet. I think if anything is currently dead (or perhaps simply dormant), it is enthusiasm for MD players to try to attempt a bank again. That is not necessarily a bad thing considering there seems to be no really solid ideas for banking right now, but it is very different than saying investor trust is dead.

First of all, we have no evidence that Phaser took in anywhere near as much ISK as he claimed, and second, IIRC the dude who ran it made a point of avoiding any forum, where he would have instantly been called out. I'm not aware of any major investors who were taken by Phaser for any significant amount, surely not billions of ISK. Same as the latest similar scam, 'Berkshire & Strawman' or whatever.

The distinction is one of affirmation; while anyone can start a 'bank' and convince people to deposit ISK, especially by spamming local in eve where the vast, vast majority aren't even aware of MD, that's a very far cry from saying "yes, people who are informed of the history of banks in EVE will invest in a new one because they trust in the idea."
Hexxx
Sebiestor Tribe
Minmatar Republic
#224 - 2012-03-20 04:18:00 UTC
Jake Andarius wrote:
KrakizBad wrote:
This is less "trust isn't dead" than "there are always new suckers."


I am confused about how that distinction is being made. If there are hundreds of "suckers" willing to invest billions of ISK in an EVE bank because they believe it will succeed, then trust is not dead. In the context of EVE banking, I think trust is dead when no banking venture can pull in even a few hundred billion anymore because there is enough cultural awareness about banks to just shut down investments. If Phaser is any indication, that simple does not seem to have happened yet. I think if anything is currently dead (or perhaps simply dormant), it is enthusiasm for MD players to try to attempt a bank again. That is not necessarily a bad thing considering there seems to be no really solid ideas for banking right now, but it is very different than saying investor trust is dead.



That's the problem with saying Trust is Dead...

Banks are complicated enough and challenging enough and discouraged enough that you'll only see about one or two a year maximum. It takes quite a bit to get them set up and launch them with any credibility. I'd argue it takes more effort than an IPO in many cases. All modern banks that have grown to size (DBANK, EBANK, and Phaser) used software to handle everything automatically except for actioning withdraws....this isn't rocket science, but it does take a bit of work, which is another deterrent.

I'd argue that audits suffer more from cultural awareness of limitations then anything else...it's pretty well understood now that without MUCH better tools than we've seen to date, audits are going to be an exercise in sadomasochism.
Darth Tickles
Doomheim
#225 - 2012-03-20 04:24:25 UTC
Hexxx wrote:
That's the problem with saying Trust is Dead...

Banks are complicated enough and challenging enough and discouraged enough that you'll only see about one or two a year maximum. It takes quite a bit to get them set up and launch them with any credibility. I'd argue it takes more effort than an IPO in many cases. All modern banks that have grown to size (DBANK, EBANK, and Phaser) used software to handle everything automatically except for actioning withdraws....this isn't rocket science, but it does take a bit of work, which is another deterrent.

I'd argue that audits suffer more from cultural awareness of limitations then anything else...it's pretty well understood now that without MUCH better tools than we've seen to date, audits are going to be an exercise in sadomasochism.


This is literally incoherent.

You are now discussing why automation is better than audits when scamming.

You have lost your mind in this thread, internet meltdown style.
Hexxx
Sebiestor Tribe
Minmatar Republic
#226 - 2012-03-20 04:25:33 UTC
Darth Tickles wrote:
Jake Andarius wrote:
I just remembered one other thing that I had been wondering earlier in regards to the main topic of the feasibility of a bank in EVE. How difficult is it from a management perspective to actually run a bank that gets tens to hundreds of billions in deposits, assuming a minimum deposit of one billion (or other numbers depending on player's knowledge)? How many different people are really required to run an operation of that size? And how much time do they have to spend managing their specific aspect of the bank? It seems like this management question is very pertinent to the overall discussion topic.


Imagine it for yourself. "Innovate" how you would deal with deposits.


I'm a bit surprised by this, but I'm not sure that the mechanics have been spelled out so I'll do that now.

Deposits - These are recognized through a deposit corporation owned by the Bank. All player transfer type journal API transactions to the corporation are treated as a deposit. Deposits by characters not linked to user profiles are queued and posted to accounts when character links are established. This should be done automatically by any EVE bank.

Withdraws - When a user profile requests a withdraw, the character it should be routed to is noted. This posts to the account and a teller has the request show up in their withdraw list. They action the withdraw manually and note it as complete, the system updates after this. This will always be done manually and represents the single largest "time sink" outside of generating revenue for the Bank.

How accounts are handled, interest recognized, etc is determined largely by how management wishes to treat it. As an example, with EBANK there were three account types, money could be moved between them within limits and transfered to other user profiles. We actually noticed a few people using EBANK for character sale payments.
Hexxx
Sebiestor Tribe
Minmatar Republic
#227 - 2012-03-20 04:27:09 UTC
Darth Tickles wrote:
[

This is literally incoherent.

You are now discussing why automation is better than audits when scamming.

You have lost your mind in this thread, internet meltdown style.


I appreciate how hard you're trying. Big smile
Jake Andarius
Andarius Trading Corp.
#228 - 2012-03-20 04:41:17 UTC  |  Edited by: Jake Andarius
Hexxx wrote:
Deposits - These are recognized through a deposit corporation owned by the Bank. All player transfer type journal API transactions to the corporation are treated as a deposit. Deposits by characters not linked to user profiles are queued and posted to accounts when character links are established. This should be done automatically by any EVE bank.

Withdraws - When a user profile requests a withdraw, the character it should be routed to is noted. This posts to the account and a teller has the request show up in their withdraw list. They action the withdraw manually and note it as complete, the system updates after this. This will always be done manually and represents the single largest "time sink" outside of generating revenue for the Bank.


I am assuming we are talking about a website database that pulls from Bank Character API that is uploaded to the website host server?

Do you mean you are surprised I was unaware of how EVE banks have handled this in the past? I custom built a google documents spreadsheet for my own needs that only requires me to update withdrawals and deposits (from less than ten investors) and automatically does the rest, so I have never delved into API usage with website database code and am a bit unfamiliar with that. Or do you simply mean you are surprised that someone has not addressed that in the thread yet?
Hexxx
Sebiestor Tribe
Minmatar Republic
#229 - 2012-03-20 04:48:32 UTC
Jake Andarius wrote:
Hexxx wrote:
Deposits - These are recognized through a deposit corporation owned by the Bank. All player transfer type journal API transactions to the corporation are treated as a deposit. Deposits by characters not linked to user profiles are queued and posted to accounts when character links are established. This should be done automatically by any EVE bank.

Withdraws - When a user profile requests a withdraw, the character it should be routed to is noted. This posts to the account and a teller has the request show up in their withdraw list. They action the withdraw manually and note it as complete, the system updates after this. This will always be done manually and represents the single largest "time sink" outside of generating revenue for the Bank.


I am assuming we are talking about a website database that pulls from Bank Character API that is uploaded to the website host server?

Do you mean you are surprised I was unaware of how EVE banks have handled this in the past? I custom built a google documents spreadsheet for my own needs that only requires me to update withdrawals and deposits (from less than ten investors) and automatically does the rest, so I have never delved into API usage with website database code and am a bit unfamiliar with that. Or do you simply mean you are surprised that someone has not addressed that in the thread yet?


I had just assumed it was fairly well understood, but now that I think about it, it really hasn't been discussed very often at all.

So basically you have a batch job setup on an hourly basis to query the deposit corporation API for player transfers and credit new deposits to the appropriate account. No button pushing, just a background service doing an API query on the hour every hour. It takes some work to build it but the benefit is that the whole thing runs like clockwork and you can even tell the user when the next deposit check is.

The interest payments worked the same way, but on a daily basis. Whatever you had in your account at the end of the day had a small amount of interest credited to it....each and every day without fail. I think people liked that. They could log in every day and every day they had a little bit more in their account balance than the day before - it's a psychological thing I think.
Jake Andarius
Andarius Trading Corp.
#230 - 2012-03-20 05:03:07 UTC  |  Edited by: Jake Andarius
Hexxx wrote:
So basically you have a batch job setup on an hourly basis to query the deposit corporation API for player transfers and credit new deposits to the appropriate account. No button pushing, just a background service doing an API query on the hour every hour. It takes some work to build it but the benefit is that the whole thing runs like clockwork and you can even tell the user when the next deposit check is.


Automating as many tasks as possible definitely seems like the best way to handle banks of a large size. However, I am curious about one thing: if the database were tampered with by someone with admin access, would that destroy depositor's records? I could see that being a huge security risk if the depositor's records were not stored in a way that tampering with them could not be easily reversed.

Hexxx wrote:
The interest payments worked the same way, but on a daily basis. Whatever you had in your account at the end of the day had a small amount of interest credited to it....each and every day without fail. I think people liked that. They could log in every day and every day they had a little bit more in their account balance than the day before - it's a psychological thing I think.


I agree with this entirely. I also do daily interest rates (for private investors, not public currently) and give all investors a private sheet for their own account. Investors definitely appreciate getting to see interest accumulate daily on their account. It builds trust between the bank and investor that you are actively keeping records of their account for them.

Edit: One other related question: how are withdrawals requested by the investors?
Hexxx
Sebiestor Tribe
Minmatar Republic
#231 - 2012-03-20 05:15:41 UTC
Jake Andarius wrote:
Hexxx wrote:
So basically you have a batch job setup on an hourly basis to query the deposit corporation API for player transfers and credit new deposits to the appropriate account. No button pushing, just a background service doing an API query on the hour every hour. It takes some work to build it but the benefit is that the whole thing runs like clockwork and you can even tell the user when the next deposit check is.


Automating as many tasks as possible definitely seems like the best way to handle banks of a large size. However, I am curious about one thing: if the database were tampered with by someone with admin access, would that destroy depositor's records? I could see that being a huge security risk if the depositor's records were not stored in a way that tampering with them could not be easily reversed.


Absolutely, but unless you're writing everything to two separate databases or making full backups every 5 minutes and retaining all of them....it's possible. You manage the risk by restricting who has direct access to the database. During EBANK's run, no more than two people had access to the database at any given time.

Jake Andarius wrote:

Hexxx wrote:
The interest payments worked the same way, but on a daily basis. Whatever you had in your account at the end of the day had a small amount of interest credited to it....each and every day without fail. I think people liked that. They could log in every day and every day they had a little bit more in their account balance than the day before - it's a psychological thing I think.


I agree with this entirely. I also do daily interest rates (for private investors, not public currently) and give all investors a private sheet for their own account. Investors definitely appreciate getting to see interest accumulate daily on their account. It builds trust between the bank and investor that you are actively keeping records of their account for them.

Edit: One other related question: how are withdrawals requested by the investors?


They would go to a withdraw page or see a withdraw option by their account summary, simply type in the amount to withdraw and designate the character it be routed to and then submit. If the amount is valid it goes to the withdraw queue and if it's not, an error message would display. Easy as pie (for the customer).


Jake Andarius
Andarius Trading Corp.
#232 - 2012-03-20 05:27:58 UTC
Hexxx wrote:
Absolutely, but unless you're writing everything to two separate databases or making full backups every 5 minutes and retaining all of them....it's possible. You manage the risk by restricting who has direct access to the database. During EBANK's run, no more than two people had access to the database at any given time.


Shocked Just my immediate impression from that is that it should be unacceptable for banks to store their records in a way that could even theoretically be destroyed by one person with no way to restore them. From my perspective, that would warrant database vulnerabilities being a huge consideration on the list of things that are currently preventing a bank from being successful in EVE.
Trading Gives MeWood
State War Academy
Caldari State
#233 - 2012-03-20 06:19:56 UTC
I said it way earlier in the post and I will say it again. Until there is some mechanism to clawback loan losses, there is no point to develop a banking system. I.E. until the bank can recover assets that it loans out through some method of foreclosure and loans are based solely on trust, there is no real incentive to loan or receive a loan.

Just my two cents.
Jake Andarius
Andarius Trading Corp.
#234 - 2012-03-20 06:54:20 UTC
Hexxx wrote:
Absolutely, but unless you're writing everything to two separate databases or making full backups every 5 minutes and retaining all of them....it's possible. You manage the risk by restricting who has direct access to the database. During EBANK's run, no more than two people had access to the database at any given time.


Thinking about it a bit more, it seems like so long as multiple copies of the API retrievals are on backup, someone could entirely destroy the database and it would not destroy the actual bank records. That is, from the API record, everything else could be restored. Would that be correct?
Vaerah Vahrokha
Vahrokh Consulting
#235 - 2012-03-20 07:44:38 UTC
Jake Andarius wrote:
Hexxx wrote:
Absolutely, but unless you're writing everything to two separate databases or making full backups every 5 minutes and retaining all of them....it's possible. You manage the risk by restricting who has direct access to the database. During EBANK's run, no more than two people had access to the database at any given time.


Thinking about it a bit more, it seems like so long as multiple copies of the API retrievals are on backup, someone could entirely destroy the database and it would not destroy the actual bank records. That is, from the API record, everything else could be restored. Would that be correct?


Past records shown that the guy doing that, contextually sucked money off. You can restore a database as much as you want but it won't magically return the missing billions in the bank.
Adunh Slavy
#236 - 2012-03-20 08:31:40 UTC
Jake Andarius wrote:

This was posted much earlier and I meant to respond to it. In a technical sense, you are correct. The shear act of money creation is necessarily inflation. That is, if there is more money added to the supply, then the cost of goods will go up ever so slightly in response. However, I am curious if you were implying something else when you wrote this, such as: if a fiat debt system is in place, then inflation must always progressively get higher over time. In that case, I would entirely disagree. For example, if labor and production of goods exceeds the amount to which the US Treasury is issuing Treasury Notes (which will be bought through private bank intermediaries before the Federal Reserve purchases them by creating money), then the price of goods would actually go down (that is, deflation would occur).


Yes, if productivity increases match monetary inflation, then prices should remain relatively stable. This assumes a perfect world with no sudden changes and perfectly rational actors.

It is the monetary inflation I was addressing, the nature of the money, debt, ensures that there is never enough money to pay for all of the debt. To service the debt the money supply has to expand, though not always expressed as Note and Coin, which incurs yet more debt; so it goes on and on and on. The rate of increase is an exponential growth function. Sooner or later it is going to explode beyond what productivity can match.

Oh, as another side note, they are not treasury notes, they are Fed Reserve Notes, read the fine print, on the bill it self. :) Last time we had treasury notes was Executive Order 11110 I believe.

Necessity is the plea for every infringement of human freedom. It is the argument of tyrants; it is the creed of slaves.  - William Pitt

Kara Roideater
#237 - 2012-03-20 09:30:15 UTC
Hexxx wrote:

Absolutely, but unless you're writing everything to two separate databases or making full backups every 5 minutes and retaining all of them....it's possible. You manage the risk by restricting who has direct access to the database. During EBANK's run, no more than two people had access to the database at any given time.




This is one aspect of the tech risk I alluded to earlier in the thread. Banks require this sort of system but history strongly suggests that eve players aren't good enough/reliable enough at programming to pull off the task. DBANK failed precisely because of a database problem and the subsequent decision of the boss that the workload involved in recovering was too high. So he kept all the cash. EBANK found out that (one of?) their programmer had been manipulating the database to siphon funds into his own account. With a bank you have very low returns due to the size and structure of the business model but considerable additional risks, such as this tech risk, over a smaller business. Any sensible investor would want greater security for a lower return but banks offer the worst of both worlds.

Btw, stop being mean to Darth. Consistency isn't his strong suit as it can play merry hell with the flow of his rhetoric. Clearly one or the other has to be sacrificed.

@Hexxx and Jake - I have very grave worries about any business model that starts from the premise that it is only workable if it targets the ignorant and exploits the stupid. If you can only get over the very high risk/ very low return issue by targetting people that don't understand how to carry out a basic risk assessment then you start from a position whereby you implicitly represent the security of the investment as something other than it is. Bankers willing to go down that dishonest road just so they can have a bank to play with are precisely the kind of bankers that led to all the previous problems with banks.
Hexxx
Sebiestor Tribe
Minmatar Republic
#238 - 2012-03-20 09:36:23 UTC
Vaerah Vahrokha wrote:
Jake Andarius wrote:
Hexxx wrote:
Absolutely, but unless you're writing everything to two separate databases or making full backups every 5 minutes and retaining all of them....it's possible. You manage the risk by restricting who has direct access to the database. During EBANK's run, no more than two people had access to the database at any given time.


Thinking about it a bit more, it seems like so long as multiple copies of the API retrievals are on backup, someone could entirely destroy the database and it would not destroy the actual bank records. That is, from the API record, everything else could be restored. Would that be correct?


Past records shown that the guy doing that, contextually sucked money off. You can restore a database as much as you want but it won't magically return the missing billions in the bank.


You could also have table logging implemented, however....again, the admin could modify that too. This is a problem in real life as well, but in real life we have proper Segregation of Duties, Change Control, and other mitigating detective controls such as logging via a separate tool to a separate database managed by a separate team. It's not impossible to replicate these in EVE....but pragmatically speaking, I wouldn't expect to see that.

And VV is right....all we can do is detect fraud. If we get really good at detecting it, we may be able to stop it before money is moved but it's difficult. This is one of the soft spots of Banking using a web system.
Hexxx
Sebiestor Tribe
Minmatar Republic
#239 - 2012-03-20 09:46:35 UTC
Kara Roideater wrote:
Hexxx wrote:

Absolutely, but unless you're writing everything to two separate databases or making full backups every 5 minutes and retaining all of them....it's possible. You manage the risk by restricting who has direct access to the database. During EBANK's run, no more than two people had access to the database at any given time.




This is one aspect of the tech risk I alluded to earlier in the thread. Banks require this sort of system but history strongly suggests that eve players aren't good enough/reliable enough at programming to pull off the task. DBANK failed precisely because of a database problem and the subsequent decision of the boss that the workload involved in recovering was too high. So he kept all the cash. EBANK found out that (one of?) their programmer had been manipulating the database to siphon funds into his own account. With a bank you have very low returns due to the size and structure of the business model but considerable additional risks, such as this tech risk, over a smaller business. Any sensible investor would want greater security for a lower return but banks offer the worst of both worlds.


One of the very important points in talking about risk management is the difference between possible and probable. Right now, today, as we speak, it is possible that your Bank in RL could experience fraud. Collusion is all it takes. It could be caught eventually, but not right away. However, this has a very very low probability. Not being dismissive of your point, I just wanted to touch on that. In EBANK's case, the database fraud was not financially significant (that is to say, not a significant amount when compared to the total finances). That doesn't mean it's ok, it simply means that the amount wasn't terribly damaging to the overall finances.


Kara Roideater wrote:

Btw, stop being mean to Darth. Consistency isn't his strong suit as it can play merry hell with the flow of his rhetoric. Clearly one or the other has to be sacrificed.

@Hexxx and Jake - I have very grave worries about any business model that starts from the premise that it is only workable if it targets the ignorant and exploits the stupid. If you can only get over the very high risk/ very low return issue by targetting people that don't understand how to carry out a basic risk assessment then you start from a position whereby you implicitly represent the security of the investment as something other than it is. Bankers willing to go down that dishonest road just so they can have a bank to play with are precisely the kind of bankers that led to all the previous problems with banks.


I believe that was a bit of facetiousness on my part at least. EBANK started with MD first and continued to get customers from MD, but I'm not sure that MD has 6,000 active participants. I would also suggest that people outside of MD are not ignorant and stupid. I am VERY interested however in presenting a realistic picture of risk to customers and investors.

We MUST do better in this regard. I cited some of the past losses as part of this effort to be transparent. This is less about "All Banks are scams" and more about "Total Bank losses over a 5 year period equal 4.4 Trillion".
Hexxx
Sebiestor Tribe
Minmatar Republic
#240 - 2012-03-20 09:50:06 UTC
Jake Andarius wrote:
Hexxx wrote:
Absolutely, but unless you're writing everything to two separate databases or making full backups every 5 minutes and retaining all of them....it's possible. You manage the risk by restricting who has direct access to the database. During EBANK's run, no more than two people had access to the database at any given time.


Thinking about it a bit more, it seems like so long as multiple copies of the API retrievals are on backup, someone could entirely destroy the database and it would not destroy the actual bank records. That is, from the API record, everything else could be restored. Would that be correct?


You would miss payments submitted to loans and transfers between user profiles. But yes....API records can be used as a consistency check, however the work here to make this happen would be very very significant.