OK, I've been gone for while, doing other necessary, but not more enjoyable things than comp.lang.cobol, but we can dispense with the flood of "Welcome back!" can't we...
Flying under the radar on page B10 of the Oct 4 _Wall Street Journal_ is this superficially innocuous article, "Fed Looks to Update Payment System", which purports to introduce the Fed's new attempt at "modernization", a goal of a 24 hour a day / 7 days a week / 365 days a year real-time payment system.
Let us, for the moment, accept this literally as stated. I see as a necessary consequence of this the complete elimination of "batch." Can we at least agree upon this, and discuss the risks inherent in that?
An excerpt: "Tuesday's proposal would involve a new, *real-time* [emphasis added] settlement system where banks process transactions as soon as a customer needs money. It would be operated by the Federal Reserve banks in cities across the country and could be used by more than 11,000 U.S. financial institutions."
The article claims that "delays" in processing present real risks to the financial system in times of stress.
It is unclear to me if the "batch window" would continue to be allowed by major banks and clearinghouse hubs, to wit, "The Fed operates a payment system between banks, which runs alongside other systems such as bank-run automated clearinghouse and credit or debit cards."
It is also unclear to me how this new paradigm would accommodate such long standing regulatory practices such as "minimum cash capitalization" which necessitate things like banks borrowing "overnight funds" to meet those requirements. If the payment system operates 24/7, uh, just where *is" overnight located?
An descriptive article on the fed website can be found here:
https://www.federalreserve.gov/newsevents/pressreleases/other20181003a.htm
Who shall we appoint to be comp.lang.cobol spokesperson for Public Input? :-)
The article also has a link to a .pdf which is the Proposal.
Am I the only one to contend what a bad, very, very, very bad idea this is, as it subjects the payment system to "race conditions", with risks so much more severe than envisioned in the current system that has batch?
One is reminded of the scene in that very good movie _Margin Call_, a fictional portrayal of a "blended" Wall Street investment bank, in the throes of discovering the imminent melt-down of the "mortgage backed securities" fiasco which caused the steep recession of 2008.
In that movie, Jeremy Irons, the firm's Chairman, is explaining to Kevin Spacey, the Trade Floor Manager: "There are three ways to succeed in this business. Number one is cheat, but I don't cheat. Number two is to be smarter, and I'm not that smart. Number three is to be *first*! (Meaning he will be first to dump the securities before other firms realize they are worth pennies on the dollar)."
Spacey: "You're panicking!"
Irons: "It's not panicking if you're the one who starts it."
Opinions everyone?
Ken
snip
Ah, the best laid plans o'mice 'n men... I share the healthy skepticism.
In Canada, we have the Phoenix payroll disaster - an ongoing horror
movie of its own: an attempt to have 1 automated payroll system for all >Federal government employees.
Now in the running for the biggest debacle in computing history. In
fact, it is such a gong show that it might win not just first prize in
that dubious race, but second place as well...
There are many examples - alas, a lot of those examples are to be found
in the government or public service arena. Anyone believe that is a >coincidence? :)
OK, I've been gone for while, doing other necessary, but not more enjoyable things than comp.lang.cobol, but we can dispense with the flood of "Welcome back!" can't we...
Flying under the radar on page B10 of the Oct 4 _Wall Street Journal_ is this superficially innocuous article, "Fed Looks to Update Payment System", which purports to introduce the Fed's new attempt at "modernization", a goal of a 24 hour a day / 7 days a week / 365 days a year real-time payment system.
Let us, for the moment, accept this literally as stated. I see as a necessary consequence of this the complete elimination of "batch." Can we at least agree upon this, and discuss the risks inherent in that?
OK, I've been gone for while, doing other necessary, but not more
enjoyable things than comp.lang.cobol, but we can dispense with the
flood of "Welcome back!" can't we...
Flying under the radar on page B10 of the Oct 4 _Wall Street Journal_ is
this superficially innocuous article, "Fed Looks to Update Payment
System", which purports to introduce the Fed's new attempt at >"modernization", a goal of a 24 hour a day / 7 days a week / 365 days a
year real-time payment system.
Let us, for the moment, accept this literally as stated. I see as a
necessary consequence of this the complete elimination of "batch." Can
we at least agree upon this, and discuss the risks inherent in that?
An excerpt: "Tuesday's proposal would involve a new, *real-time*
[emphasis added] settlement system where banks process transactions as
soon as a customer needs money. It would be operated by the Federal
Reserve banks in cities across the country and could be used by more
than 11,000 U.S. financial institutions."
The article claims that "delays" in processing present real risks to the >financial system in times of stress.
It is unclear to me if the "batch window" would continue to be allowed
by major banks and clearinghouse hubs, to wit, "The Fed operates a
payment system between banks, which runs alongside other systems such as >bank-run automated clearinghouse and credit or debit cards."
It is also unclear to me how this new paradigm would accommodate such
long standing regulatory practices such as "minimum cash capitalization" >which necessitate things like banks borrowing "overnight funds" to meet
those requirements. If the payment system operates 24/7, uh, just where
*is" overnight located?
An descriptive article on the fed website can be found here:
https://www.federalreserve.gov/newsevents/pressreleases/other20181003a.htm
Who shall we appoint to be comp.lang.cobol spokesperson for Public Input? :-)
The article also has a link to a .pdf which is the Proposal.
Am I the only one to contend what a bad, very, very, very bad idea this
is, as it subjects the payment system to "race conditions", with risks
so much more severe than envisioned in the current system that has
batch?
One is reminded of the scene in that very good movie _Margin Call_, a >fictional portrayal of a "blended" Wall Street investment bank, in the
throes of discovering the imminent melt-down of the "mortgage backed >securities" fiasco which caused the steep recession of 2008.
In that movie, Jeremy Irons, the firm's Chairman, is explaining to Kevin >Spacey, the Trade Floor Manager: "There are three ways to succeed in
this business. Number one is cheat, but I don't cheat. Number two is to
be smarter, and I'm not that smart. Number three is to be *first*!
(Meaning he will be first to dump the securities before other firms
realize they are worth pennies on the dollar)."
Spacey: "You're panicking!"
Irons: "It's not panicking if you're the one who starts it."
On 10/5/2018 6:10 AM, Ken wrote:
Opinions everyone?
Ah, the best laid plans o'mice 'n men... I share the healthy skepticism.
In Canada, we have the Phoenix payroll disaster - an ongoing horror
movie of its own: an attempt to have 1 automated payroll system for all >Federal government employees.
In article <a6b08917-8735-4824-beb5-c2383fbbe057@googlegroups.com>,
Ken wrote:
OK, I've been gone for while, doing other necessary, but not more
enjoyable things than comp.lang.cobol, but we can dispense with the
flood of "Welcome back!" can't we...
Welcome ba... oh, all right.
Flying under the radar on page B10 of the Oct 4 _Wall Street Journal_ is >this superficially innocuous article, "Fed Looks to Update Payment
System", which purports to introduce the Fed's new attempt at >"modernization", a goal of a 24 hour a day / 7 days a week / 365 days a >year real-time payment system.
Not quite, Mr Shafer... paragraph 3, sentence 2: 'The Board is not committing to any specific action and is seeking input on which, if any, actions the Federal Reserve should take.'
It is a meeting that seeks to generate more meetings.
Let us, for the moment, accept this literally as stated. I see as a >necessary consequence of this the complete elimination of "batch." Can
we at least agree upon this, and discuss the risks inherent in that?
Given that a rolling stone gathers no moss... then money that is always in motion has less time to sit and earn interest.
An excerpt: "Tuesday's proposal would involve a new, *real-time*
[emphasis added] settlement system where banks process transactions as
soon as a customer needs money. It would be operated by the Federal
Reserve banks in cities across the country and could be used by more
than 11,000 U.S. financial institutions."
The language here is so loose that it needs an antidiarrheal. A
transaction with a street-vendor might be seen as needing 'immediate
money' but a minor purchase - an automobile, a house, a parcel of land sufficiently large for a moderate-sized manufacturing facility - all take place with nary a coin nor banknote to be seen.
'One size fits all' does not apply to finance.
The article claims that "delays" in processing present real risks to the >financial system in times of stress.
Or... delays in processing prevent runaway panics. A small village can
make do with the ebb-and-flow of a river but a city of any size is well-served by a reservoir.
It is unclear to me if the "batch window" would continue to be allowed
by major banks and clearinghouse hubs, to wit, "The Fed operates a
payment system between banks, which runs alongside other systems such as >bank-run automated clearinghouse and credit or debit cards."
It is also unclear to me how this new paradigm would accommodate such
long standing regulatory practices such as "minimum cash capitalization" >which necessitate things like banks borrowing "overnight funds" to meet >those requirements. If the payment system operates 24/7, uh, just where >*is" overnight located?
The Royal Observatory in Greenwich.
An descriptive article on the fed website can be found here:
https://www.federalreserve.gov/newsevents/pressreleases/other20181003a.htm
Who shall we appoint to be comp.lang.cobol spokesperson for Public Input? :-)
I don't see a rate, or range of rates, associated with the position
offered.
The article also has a link to a .pdf which is the Proposal.
Am I the only one to contend what a bad, very, very, very bad idea this
is, as it subjects the payment system to "race conditions", with risks
so much more severe than envisioned in the current system that has
batch?
Mr Trembley might have some interesting views about a large business
that's always open.
One is reminded of the scene in that very good movie _Margin Call_, a >fictional portrayal of a "blended" Wall Street investment bank, in the >throes of discovering the imminent melt-down of the "mortgage backed >securities" fiasco which caused the steep recession of 2008.
In that movie, Jeremy Irons, the firm's Chairman, is explaining to Kevin >Spacey, the Trade Floor Manager: "There are three ways to succeed in
this business. Number one is cheat, but I don't cheat. Number two is to
be smarter, and I'm not that smart. Number three is to be *first*!
(Meaning he will be first to dump the securities before other firms
realize they are worth pennies on the dollar)."
Spacey: "You're panicking!"
Irons: "It's not panicking if you're the one who starts it."
A while back there was no Federal Reserve and banks would issue their own notes. Banks, like any other businesses - governments included - sometimes cannot meet the promises they've made for money and the bank is then, literally, broken... 'bank-rupt'.
(the language tells so many things if one only listens!)
Anyhow... a bank in, say, Rochester, New York would fail and its paper
would become worthless. Notice of the bankruptcy would be printed and sent by barge to Albany, the state capitol... about 250 miles, give or take a few.
Meanwhile... riders on horseback would race the barges, hoping to reach Albany and sell off the notes before anyone found out they were worthless.
This was in the mid-19th century. Knowing first can make for a tidy
profit.
DD
*You make me laugh* all the while financial calamity is "imminent" !
OK, I've been gone for while, doing other necessary, but not more enjoyable things than comp.lang.cobol, but we can dispense with the flood of "Welcome back!" can't we...
Flying under the radar on page B10 of the Oct 4 _Wall Street Journal_ is this superficially innocuous article, "Fed Looks to Update Payment System", which purports to introduce the Fed's new attempt at "modernization", a goal of a 24 hour a day / 7 days a week / 365 days a year real-time payment system.
Let us, for the moment, accept this literally as stated. I see as a necessary consequence of this the complete elimination of "batch." Can we at least agree upon this, and discuss the risks inherent in that?
An excerpt: "Tuesday's proposal would involve a new, *real-time* [emphasis added] settlement system where banks process transactions as soon as a customer needs money. It would be operated by the Federal Reserve banks in cities across the country and could be used by more than 11,000 U.S. financial institutions."
The article claims that "delays" in processing present real risks to the financial system in times of stress.
It is unclear to me if the "batch window" would continue to be allowed by major banks and clearinghouse hubs, to wit, "The Fed operates a payment system between banks, which runs alongside other systems such as bank-run automated clearinghouse and credit or debit cards."
It is also unclear to me how this new paradigm would accommodate such long standing regulatory practices such as "minimum cash capitalization" which necessitate things like banks borrowing "overnight funds" to meet those requirements. If the payment system operates 24/7, uh, just where *is" overnight located?
An descriptive article on the fed website can be found here:
https://www.federalreserve.gov/newsevents/pressreleases/other20181003a.htm
Who shall we appoint to be comp.lang.cobol spokesperson for Public Input? :-)
The article also has a link to a .pdf which is the Proposal.
Am I the only one to contend what a bad, very, very, very bad idea this is, as it subjects the payment system to "race conditions", with risks so much more severe than envisioned in the current system that has batch?
One is reminded of the scene in that very good movie _Margin Call_, a fictional portrayal of a "blended" Wall Street investment bank, in the throes of discovering the imminent melt-down of the "mortgage backed securities" fiasco which caused the steep recession of 2008.
In that movie, Jeremy Irons, the firm's Chairman, is explaining to Kevin Spacey, the Trade Floor Manager: "There are three ways to succeed in this business. Number one is cheat, but I don't cheat. Number two is to be smarter, and I'm not that smart. Number three is to be *first*! (Meaning he will be first to dump the securities before other firms realize they are worth pennies on the dollar)."
Spacey: "You're panicking!"
Irons: "It's not panicking if you're the one who starts it."
Opinions everyone?
Ken
Check your calendar, and then recognize that change is inevitable andChange is inevitable, and so is the motivation of the greedy to "work the system." The problem with "real-time" updates is that it is not instantaneously real-time across the globe. There will be those who will take advantage of the latency, no matter how small. Evidence of this is in the stock market, where certain powerful concerns spent large sums of money to obtain a few microseconds advantage in trading by installing high-bandwidth systems available to them, but not to all. "Batch" allows some breathing room to sort that out.
the best you can do is embrace it and help make sure it goes well.
Pete.
On Saturday, October 6, 2018 at 8:18:46 PM UTC-4, pete dashwood wrote: <<stuff deleted<<
Check your calendar, and then recognize that change is inevitable and
the best you can do is embrace it and help make sure it goes well.
Pete.
Change is inevitable, and so is the motivation of the greedy to "work the system." The problem with "real-time" updates is that it is not instantaneously real-time across the globe. There will be those who will take advantage of the latency, no matter how small. Evidence of this is in the stock market, where certain powerful concerns spent large sums of money to obtain a few microseconds advantage in trading by installing high-bandwidth systems available to them, but not to all. "Batch" allows some breathing room to sort that out.
As far as your directive for me to "check my calendar," there is nothing wrong with being a technological conservative, and insisting that risks be identified and managed. And I find the haughtiness in that directive unpersuasive. (Though I know you mean well, and I have known you for a while, so you are advantaged with a "special dispensation." :-) )
I have kept an eye on the financial excesses the past few decades, and the ruins left in their wake. The financial news reporter analyst "tells" (as in poker "tells") that preceded those crises always reflected a significant uptick in words such as "creative financial instruments", "progress", and "efficiency."
I see these same "tells" in these proposals by the Federal Reserve.
Ken
On 8/10/2018 12:49 AM, Ken wrote:
On Saturday, October 6, 2018 at 8:18:46 PM UTC-4, pete dashwood wrote:the system." The problem with "real-time" updates is that it is not >instantaneously real-time across the globe. There will be those who will
<<stuff deleted<<
Check your calendar, and then recognize that change is inevitable and
the best you can do is embrace it and help make sure it goes well.
Pete.
Change is inevitable, and so is the motivation of the greedy to "work
take advantage of the latency, no matter how small. Evidence of this is
in the stock market, where certain powerful concerns spent large sums of >money to obtain a few microseconds advantage in trading by installing >high-bandwidth systems available to them, but not to all. "Batch" allows
some breathing room to sort that out.
nothing wrong with being a technological conservative, and insisting
As far as your directive for me to "check my calendar," there is
that risks be identified and managed. And I find the haughtiness in that >directive unpersuasive. (Though I know you mean well, and I have known
you for a while, so you are advantaged with a "special dispensation."
:-) )
the ruins left in their wake. The financial news reporter analyst
I have kept an eye on the financial excesses the past few decades, and
"tells" (as in poker "tells") that preceded those crises always
reflected a significant uptick in words such as "creative financial >instruments", "progress", and "efficiency."
OK, There is no point in us debating this as no minds will be changed
I see these same "tells" in these proposals by the Federal Reserve.
Ken
(or even adjusted... :-))
life.Thanks for this post, Doc.
Now... if make clay tiles and a small loan would allow you to purchase
molds and forms sufficient to increase your output 25%/month... that's
Maybe not yours, Mr Dashwood, and maybe not Mr Shafer's... but there might be others whose opinions are not so ossified.Gee, thanks, Doc ... for making me feel OLD. :-)
Mr Shafer speaks of decades of experience which has lead him to aThree banking story anecdotes. Maybe not directly related to batch vs. realtime, then again, maybe so.
different conclusion. That might - at least as an attempt at professional courtesy - merit a bit more consideration than a flippant 'look at your calendar'.
On Sunday, October 7, 2018 at 11:50:29 PM UTC-4, docd...@panix.com wrote:A bit of a followup to the Manhattan bank story...
On 8/10/2018 4:50 PM, docdwarf@panix.com wrote:
life.
Now... if make clay tiles and a small loan would allow you to purchase
molds and forms sufficient to increase your output 25%/month... that's >Thanks for this post, Doc.
You just reminded me why I've almost stopped coming here.
On Sunday, October 7, 2018 at 11:50:29 PM UTC-4, docd...@panix.com wrote:
Maybe not yours, Mr Dashwood, and maybe not Mr Shafer's... but there might >> be others whose opinions are not so ossified.
Gee, thanks, Doc ... for making me feel OLD. :-)
My sales rep, who brought me the replacement check, asked me, "Just what
did you do and say at that branch? Their manager said he was a Vietnam >veteran and if you ever came back to his branch he would KILL you!"
I laughed, but since this involved money, took his threat seriously.
From that time on, I had only to go to Mr. Friendly Bank account rep,
who would call up Mr. Vietnam Vet who wanted to kill me, and ask him,
"Has check number 2233 cleared?" Upon a confirming yes, they
chit-chatted pleasantries for thirty seconds, then he hang up and told
me the good news.
"How was he?" I inquired. "Very friendly," Mr. Nice Banker said.
Empirical Conclusion important to this discussion: Bankers treat other >bankers differently than how they treat you and me.
So what does it mean for a check to clear?
Story 3 - This also recent. Was talking with a counselor friend of mine,
who related how his savings account at a major U.S. bank (you would
recognize its name), had been systematically pilfered with a series of >smallish transactions (ATM?), originating from another country, right
before his, and the bank service rep's, very eyes. He exclaimed, "I'm
the account owner! I'm right here! I'm being robbed! Do something!" The
bank service rep said there was nothing they could do, except let the
account go to ZERO, which it did, and then he could apply to be "made
whole." And that would take "a while." It mattered not at all to the
Bank that his mortgage was an automatic withdrawal from that account,
and that the money would not be there for a "while." In fact, it was a
month before he was made whole.
No, that was not it. What I eventually figured out is that he was saying
that if the music stops, and there is as big ol' mess o' bad debt out
there, and different institutions are holding different amounts of it,
with transactions made at differing times, then THERE WAS NO PRECEDENT,
AND ESPECIALLY, NO LEGAL PRECEDENT TO DETERMINE WHO IS LEFT HOLDING THE
BAG. It is that very thing which resulted in the credit lockup that
sent us into a near depression. Nobody would invest until they were sure
they would get paid by who owed them, and those who owed would not pay
until they were sure the money had "cleared" by who owed THEM, etc. etc.
etc.
All of that will be compounded a hundred times worse when we are talking >about money itself.
For today, just what *is* money? It is no longer (and has not been for a
long time) convertible into gold or silver. Money is simply the
*confidence* that we have that some other party will accept it as
"payment" for what we want. When that confidence is eroded, like what >happened with the mortgaged backed securities, the system locks up.
Batch allows for closing down the poker game. The chips are cashed in. >Everyone goes home and sleeps it off. And then getting up the next day
and playing again.
Sleep is a natural phenomenon of living things. So is breathing.
In article <a6b08917-8735-4824-beb5-c2383fbbe057@googlegroups.com>,
Ken <klshafer@att.net> wrote:
OK, I've been gone for while, doing other necessary, but not more
enjoyable things than comp.lang.cobol, but we can dispense with the
flood of "Welcome back!" can't we...
Welcome ba... oh, all right.
Flying under the radar on page B10 of the Oct 4 _Wall Street Journal_ is
this superficially innocuous article, "Fed Looks to Update Payment
System", which purports to introduce the Fed's new attempt at
"modernization", a goal of a 24 hour a day / 7 days a week / 365 days a
year real-time payment system.
Not quite, Mr Shafer... paragraph 3, sentence 2: 'The Board is not
committing to any specific action and is seeking input on which, if any, actions the Federal Reserve should take.'
It is a meeting that seeks to generate more meetings.
Let us, for the moment, accept this literally as stated. I see as a
necessary consequence of this the complete elimination of "batch." Can
we at least agree upon this, and discuss the risks inherent in that?
Given that a rolling stone gathers no moss... then money that is always in motion has less time to sit and earn interest.
An excerpt: "Tuesday's proposal would involve a new, *real-time*
[emphasis added] settlement system where banks process transactions as
soon as a customer needs money. It would be operated by the Federal
Reserve banks in cities across the country and could be used by more
than 11,000 U.S. financial institutions."
The language here is so loose that it needs an antidiarrheal. A
transaction with a street-vendor might be seen as needing 'immediate
money' but a minor purchase - an automobile, a house, a parcel of land sufficiently large for a moderate-sized manufacturing facility - all take place with nary a coin nor banknote to be seen.
'One size fits all' does not apply to finance.
The article claims that "delays" in processing present real risks to the
financial system in times of stress.
Or... delays in processing prevent runaway panics. A small village can
make do with the ebb-and-flow of a river but a city of any size is well-served by a reservoir.
It is unclear to me if the "batch window" would continue to be allowed
by major banks and clearinghouse hubs, to wit, "The Fed operates a
payment system between banks, which runs alongside other systems such as
bank-run automated clearinghouse and credit or debit cards."
It is also unclear to me how this new paradigm would accommodate such
long standing regulatory practices such as "minimum cash capitalization"
which necessitate things like banks borrowing "overnight funds" to meet
those requirements. If the payment system operates 24/7, uh, just where
*is" overnight located?
The Royal Observatory in Greenwich.
An descriptive article on the fed website can be found here:
https://www.federalreserve.gov/newsevents/pressreleases/other20181003a.htm >>
Who shall we appoint to be comp.lang.cobol spokesperson for Public Input? :-)
I don't see a rate, or range of rates, associated with the position
offered.
The article also has a link to a .pdf which is the Proposal.
Am I the only one to contend what a bad, very, very, very bad idea this
is, as it subjects the payment system to "race conditions", with risks
so much more severe than envisioned in the current system that has
batch?
Mr Trembley might have some interesting views about a large business
that's always open.
On 10/5/2018 3:34 PM, docdwarf@panix.com wrote:
In article <a6b08917-8735-4824-beb5-c2383fbbe057@googlegroups.com>,
Ken <klshafer@att.net> wrote:
OK, I've been gone for while, doing other necessary, but not more
enjoyable things than comp.lang.cobol, but we can dispense with the
flood of "Welcome back!" can't we...
Welcome ba... oh, all right.
Flying under the radar on page B10 of the Oct 4 _Wall Street Journal_ is >>> this superficially innocuous article, "Fed Looks to Update Payment
System", which purports to introduce the Fed's new attempt at
"modernization", a goal of a 24 hour a day / 7 days a week / 365 days a
year real-time payment system.
Not quite, Mr Shafer... paragraph 3, sentence 2: 'The Board is not
committing to any specific action and is seeking input on which, if any,
actions the Federal Reserve should take.'
It is a meeting that seeks to generate more meetings.
Let us, for the moment, accept this literally as stated. I see as a
necessary consequence of this the complete elimination of "batch." Can
we at least agree upon this, and discuss the risks inherent in that?
Given that a rolling stone gathers no moss... then money that is
always in
motion has less time to sit and earn interest.
An excerpt: "Tuesday's proposal would involve a new, *real-time*
[emphasis added] settlement system where banks process transactions as
soon as a customer needs money. It would be operated by the Federal
Reserve banks in cities across the country and could be used by more
than 11,000 U.S. financial institutions."
The language here is so loose that it needs an antidiarrheal. A
transaction with a street-vendor might be seen as needing 'immediate
money' but a minor purchase - an automobile, a house, a parcel of land
sufficiently large for a moderate-sized manufacturing facility - all take
place with nary a coin nor banknote to be seen.
'One size fits all' does not apply to finance.
The article claims that "delays" in processing present real risks to the >>> financial system in times of stress.
Or... delays in processing prevent runaway panics. A small village can
make do with the ebb-and-flow of a river but a city of any size is
well-served by a reservoir.
It is unclear to me if the "batch window" would continue to be allowed
by major banks and clearinghouse hubs, to wit, "The Fed operates a
payment system between banks, which runs alongside other systems such as >>> bank-run automated clearinghouse and credit or debit cards."
It is also unclear to me how this new paradigm would accommodate such
long standing regulatory practices such as "minimum cash capitalization" >>> which necessitate things like banks borrowing "overnight funds" to meet
those requirements. If the payment system operates 24/7, uh, just where
*is" overnight located?
The Royal Observatory in Greenwich.
An descriptive article on the fed website can be found here:
https://www.federalreserve.gov/newsevents/pressreleases/other20181003a.htm >>>
Who shall we appoint to be comp.lang.cobol spokesperson for Public
Input? :-)
I don't see a rate, or range of rates, associated with the position
offered.
The article also has a link to a .pdf which is the Proposal.
Am I the only one to contend what a bad, very, very, very bad idea this
is, as it subjects the payment system to "race conditions", with risks
so much more severe than envisioned in the current system that has
batch?
Mr Trembley might have some interesting views about a large business
that's always open.
I suppose I should wake up.
I've been retired for over two years, but I am peripherally aware of a Clearing and Settlement system that runs 24 hours a day, 365.25 days a
year, and which processes hundreds of millions of transactions per day
for billions of dollars. To the best of my knowledge it is a batch processing system, even though it has many features of a real-time
system. And it is written entirely in COBOL.
If you're interested in the technical details of systems that are truly real-time, you should probably look at debit card processing instead.
One of the challenges of high-availability systems is how to stay
available while installing hardware upgrades, software upgrades, code
fixes, and database conversions.
On 10/10/2018 03:11 AM, Arnold Trembley wrote:
On 10/5/2018 3:34 PM, docdwarf@panix.com wrote:
In article <a6b08917-8735-4824-beb5-c2383fbbe057@googlegroups.com>,
Ken <klshafer@att.net> wrote:
OK, I've been gone for while, doing other necessary, but not more
enjoyable things than comp.lang.cobol, but we can dispense with the
flood of "Welcome back!" can't we...
Welcome ba... oh, all right.
Flying under the radar on page B10 of the Oct 4 _Wall Street Journal_ is >>> this superficially innocuous article, "Fed Looks to Update Payment
System", which purports to introduce the Fed's new attempt at
"modernization", a goal of a 24 hour a day / 7 days a week / 365 days a >>> year real-time payment system.
Not quite, Mr Shafer... paragraph 3, sentence 2: 'The Board is not
committing to any specific action and is seeking input on which, if any, >> actions the Federal Reserve should take.'
It is a meeting that seeks to generate more meetings.
Let us, for the moment, accept this literally as stated. I see as a
necessary consequence of this the complete elimination of "batch." Can >>> we at least agree upon this, and discuss the risks inherent in that?
Given that a rolling stone gathers no moss... then money that is
always in
motion has less time to sit and earn interest.
An excerpt: "Tuesday's proposal would involve a new, *real-time*
[emphasis added] settlement system where banks process transactions as >>> soon as a customer needs money. It would be operated by the Federal
Reserve banks in cities across the country and could be used by more
than 11,000 U.S. financial institutions."
The language here is so loose that it needs an antidiarrheal. A
transaction with a street-vendor might be seen as needing 'immediate
money' but a minor purchase - an automobile, a house, a parcel of land
sufficiently large for a moderate-sized manufacturing facility - all take >> place with nary a coin nor banknote to be seen.
'One size fits all' does not apply to finance.
The article claims that "delays" in processing present real risks to the >>> financial system in times of stress.
Or... delays in processing prevent runaway panics. A small village can
make do with the ebb-and-flow of a river but a city of any size is
well-served by a reservoir.
It is unclear to me if the "batch window" would continue to be allowed >>> by major banks and clearinghouse hubs, to wit, "The Fed operates a
payment system between banks, which runs alongside other systems such as >>> bank-run automated clearinghouse and credit or debit cards."
It is also unclear to me how this new paradigm would accommodate such
long standing regulatory practices such as "minimum cash capitalization" >>> which necessitate things like banks borrowing "overnight funds" to meet >>> those requirements. If the payment system operates 24/7, uh, just where >>> *is" overnight located?
The Royal Observatory in Greenwich.
An descriptive article on the fed website can be found here:
https://www.federalreserve.gov/newsevents/pressreleases/other20181003a.htm
Who shall we appoint to be comp.lang.cobol spokesperson for Public
Input? :-)
I don't see a rate, or range of rates, associated with the position
offered.
The article also has a link to a .pdf which is the Proposal.
Am I the only one to contend what a bad, very, very, very bad idea this >>> is, as it subjects the payment system to "race conditions", with risks >>> so much more severe than envisioned in the current system that has
batch?
Mr Trembley might have some interesting views about a large business
that's always open.
I suppose I should wake up.
I've been retired for over two years, but I am peripherally aware of a Clearing and Settlement system that runs 24 hours a day, 365.25 days a year, and which processes hundreds of millions of transactions per day
for billions of dollars. To the best of my knowledge it is a batch processing system, even though it has many features of a real-time system. And it is written entirely in COBOL.
Impossible. COBOL is dead.
If you're interested in the technical details of systems that are truly real-time, you should probably look at debit card processing instead.
Hardly real-time. It works even when the ATM is off-line. And
certainly doesn't require real-time response. I have yet to use
an ATM that responded in anything resembling speed.
Sorry to tell ya Bill.One of the challenges of high-availability systems is how to stay available while installing hardware upgrades, software upgrades, code fixes, and database conversions.
Easily done. Look at OpenVMS Clusters. Take one machine out
of the cluster, upgrade it, put it back in the cluster, repeat
until all machines have been upgraded. Oh wait, OpenVMS is as
dead as COBOL. :-)
bill
On Wednesday, 10 October 2018 22:05:58 UTC+10, Bill Gunshannon wrote:
On 10/10/2018 03:11 AM, Arnold Trembley wrote:
On 10/5/2018 3:34 PM, docdwarf@panix.com wrote:
In article <a6b08917-8735-4824-beb5-c2383fbbe057@googlegroups.com>,
Ken <klshafer@att.net> wrote:
OK, I've been gone for while, doing other necessary, but not more
enjoyable things than comp.lang.cobol, but we can dispense with the
flood of "Welcome back!" can't we...
Welcome ba... oh, all right.
Flying under the radar on page B10 of the Oct 4 _Wall Street Journal_ is >>>>> this superficially innocuous article, "Fed Looks to Update Payment
System", which purports to introduce the Fed's new attempt at
"modernization", a goal of a 24 hour a day / 7 days a week / 365 days a >>>>> year real-time payment system.
Not quite, Mr Shafer... paragraph 3, sentence 2: 'The Board is not
committing to any specific action and is seeking input on which, if any, >>>> actions the Federal Reserve should take.'
It is a meeting that seeks to generate more meetings.
Let us, for the moment, accept this literally as stated. I see as a
necessary consequence of this the complete elimination of "batch." Can >>>>> we at least agree upon this, and discuss the risks inherent in that?
Given that a rolling stone gathers no moss... then money that is
always in
motion has less time to sit and earn interest.
An excerpt: "Tuesday's proposal would involve a new, *real-time*
[emphasis added] settlement system where banks process transactions as >>>>> soon as a customer needs money. It would be operated by the Federal
Reserve banks in cities across the country and could be used by more >>>>> than 11,000 U.S. financial institutions."
The language here is so loose that it needs an antidiarrheal. A
transaction with a street-vendor might be seen as needing 'immediate
money' but a minor purchase - an automobile, a house, a parcel of land >>>> sufficiently large for a moderate-sized manufacturing facility - all take >>>> place with nary a coin nor banknote to be seen.
'One size fits all' does not apply to finance.
The article claims that "delays" in processing present real risks to the >>>>> financial system in times of stress.
Or... delays in processing prevent runaway panics. A small village can >>>> make do with the ebb-and-flow of a river but a city of any size is
well-served by a reservoir.
It is unclear to me if the "batch window" would continue to be allowed >>>>> by major banks and clearinghouse hubs, to wit, "The Fed operates a
payment system between banks, which runs alongside other systems such as >>>>> bank-run automated clearinghouse and credit or debit cards."
It is also unclear to me how this new paradigm would accommodate such >>>>> long standing regulatory practices such as "minimum cash capitalization" >>>>> which necessitate things like banks borrowing "overnight funds" to meet >>>>> those requirements. If the payment system operates 24/7, uh, just where >>>>> *is" overnight located?
The Royal Observatory in Greenwich.
An descriptive article on the fed website can be found here:
https://www.federalreserve.gov/newsevents/pressreleases/other20181003a.htm
Who shall we appoint to be comp.lang.cobol spokesperson for Public
Input? :-)
I don't see a rate, or range of rates, associated with the position
offered.
The article also has a link to a .pdf which is the Proposal.
Am I the only one to contend what a bad, very, very, very bad idea this >>>>> is, as it subjects the payment system to "race conditions", with risks >>>>> so much more severe than envisioned in the current system that has >>>>> batch?
Mr Trembley might have some interesting views about a large business
that's always open.
I suppose I should wake up.
I've been retired for over two years, but I am peripherally aware of a
Clearing and Settlement system that runs 24 hours a day, 365.25 days a
year, and which processes hundreds of millions of transactions per day
for billions of dollars. To the best of my knowledge it is a batch
processing system, even though it has many features of a real-time
system. And it is written entirely in COBOL.
Impossible. COBOL is dead.
If you're interested in the technical details of systems that are truly
real-time, you should probably look at debit card processing instead.
Hardly real-time. It works even when the ATM is off-line. And
certainly doesn't require real-time response. I have yet to use
an ATM that responded in anything resembling speed.
One of the challenges of high-availability systems is how to stay
available while installing hardware upgrades, software upgrades, code
fixes, and database conversions.
Easily done. Look at OpenVMS Clusters. Take one machine out
of the cluster, upgrade it, put it back in the cluster, repeat
until all machines have been upgraded. Oh wait, OpenVMS is as
dead as COBOL. :-)
bill
Sorry to tell ya Bill.
I wrote a real time bank-side of a share trading system in Cobol. It cost my client $25,000 compared to someone elses $1m for another bank. It is real-time response to share trades. It also has a batch component. Overnight systems synchronize with TCP/IP messages that contain many entries per message hence it is batch. My Cobol system was also faster than other providers.
| Sysop: | DaiTengu |
|---|---|
| Location: | Appleton, WI |
| Users: | 991 |
| Nodes: | 10 (0 / 10) |
| Uptime: | 81:20:39 |
| Calls: | 12,949 |
| Calls today: | 3 |
| Files: | 186,574 |
| Messages: | 3,264,666 |