Wait, who? The what assocation? Never heard of 'em. What are they selling?
The technology underpinning the UK's financial services sector needs urgent investment if regulators are to be able to spot potential abuses and prevent future financial crises, an influential report has claimed. Intellect, the trade association for the UK technology sector, said in its report (62-page/5MB PDF) that banks …
Guys should learn about "Austrian Business Cycle Theory", "Elimination of Moral Hazard Through Bailouts and Free Cash via Central Banking" and "Money not backed by a Physical Commodity is just another Wealth-Shifting Scam".
Until then, all this regulatory faggotry is just howling at the moon.
Are you aware faggot is also a tasty tasty meat/offal based dish in the uk made approriately enough by a company called Brains (other faggot producers exist)?
Thought not - and equally irrelevant to the matter at hand.
As someone working in the industry for 10 years including a stint in RBS, this person does not know what they're talking about. A correctly staffed and run 'legacy' system is often the most reliable thing in the house. The failure of financial IT comes from up top, trying to replace expensive skilled staff with multiple trained monkeys. Incidentally those are the same people who oversaw the rest of the financial crisis - what a surprise.
Remember that banks hire the same MBA's who were taught that there is no difference between that IT worker who's spent 5+ years maintaining and enhancing a system and the guy overseas in India who claims to have 5+ years in IT, except that the guy in India can be paid peanuts thereby adding more money to the bottom line.
Going to the Cloud to save costs?
I just saw a post on a Hadoop mailing list where an employee of one of those offshore consulting firms was begging for help on how to set up a 100 node Hadoop cluster. The point is that the poor bloke either exaggerated his skills or his employer exaggerated their competency in this area.
Clearly the message that you get what you paid for isn't being heard.
While we in the industry know the dangers of sourcing critical infrastructure to the lowest bidder, that doesn't mean that you can trust the global system integrators in providing the best and the brightest either. Most are hiring the same MBA mental midgets and are sourcing their labor from the same offshore pool...
Terrible as in it is that type of thinking that made the mess in banking IT. Also "As a long-term solution, the report suggests that banks move to a "central", cloud-based system for non-competitive elements of their infrastructure" looks like bullocks, smells like them and it is. Given that I suspect they have no idea what a mainframe is and how they are used in banking etc.
It's cluster mongs who write reports like this that cause the problem, they probably did a report few years back playing fashion-IT and said you need to outsource and get rid of contractors and the muggle managment who are useless went ahead and did that.
Has anybody heard of these muppets who wrote the report and have anything good to say about them?
"Banks are willing to spend money on cutting-edge technology that facilitates high frequency trading or reduces the time it takes to process a transaction in the capital markets ... allows them to deliver better customer services"
But isn't it the regulators that put the chinese walls in place that stop the capital markets part of the banks from being too closely tied to the customer / high-street side of the banks? The side that makes the most money, naturally, has the money to spend on improving it's systems. The other side also has infrastructure and numbers of accounts that so vast and widespread, that the sheer amount of work required to upgrade those systems that means such projects are very expensive and very time consuming.
Regulators suddenly saying "you need to do this" isn't going to get it done any faster. It just puts more pressure, and so more risk, on the under-funded part of the IT Dept.
Insufficient data and information sharing within banks prevented institutions from having an accurate view of their business as a whole, the report said.
Again, isn't this 'siloing of information' at the insistence of the regulators already?
Sounds like whoever wrote this report didn't really do any research into how things are currently run, and merely came up with their own ideas of what's wrong and what should be done to fix what they perceive as being the problem. This isn't a report, it's an opinion piece.
Yep, I was about to post similar welcoming this mob to the real World where the parts of the business that can make the most profit get invested in over making the customer feel all warm and fuzzy inside. It's a business out for profit and research has shown how much inertia there is in changing banks so why would they bother?
After the 'great' depression, There was an act passed in the US which limited what banks could do. This was the Glass-Steagal act of 1933.
It was overturned because the big banks could not compete with the financial trading firms which were not banks but were offering banking services like checking accounts tied to their customer's brokerage accounts.
This is what opened the door to all sorts of nonsense.
"Insufficient data and information sharing within banks prevented institutions from having an accurate view of their business as a whole, the report said."
That's funny. Banks have been running ALCO monthly meetings for the past 2 decades. ALCO == Asset/Liability Committee where senior bank execs determine their exposure to risks.
Its not a question of not having the information. 15+ years ago, it took all night for a single high end workstation to run through a Monte Carlo simulation on interest rate risk against the entire bank's portfolio. Today a single server can do that in under an hour. Going to a Hadoop cluster, it could be done in less time, higher resolution and with more scenarios yielding more accurate data.
Sorry, but you want to blame something, focus on the Glass-Steagal act which would have prevented the greed and short term gains at your saving's risk. This isn't a technical problem, but a regulatory one. Trying to hype technology for a people issue? It will never end well.
Don't forget the spectacular failure of politicians and regulators.
You can see in Hansard, the leader of the Lib Dems at the time warning again and again that the level of debt built up in this country was dangerous, just to be shouted down by New Labour and the Tories.
Ian Michael Gumby. You are so implausibly sorrect. Collect £500. Advance to go. Do not go to jail.
In my opinion at least Gordon Brown's "Liberalise the banking community, to get us re-elected" plan, otherwise known as "Gordon's an economic genius" forced the US to force it to be rescinded in NY to become competitive, and then the Mansion House Gordon Brown moment "Lend anything you like to anyone for housing, because the americans are doing it, and they're getting re-elected" sealed the fate. Of course I could be wrong. http://www.investortoday.co.uk/news_features/Gordon-Brown--economic-genius-or-worst-Prime-Minister-ever-
"How else can the most unpopular Prime Minister of modern times have kept going for so long (even if he did, in his own words, "save the world" from financial ruin)?" -- It wasn't even him that came up with the idea. It was someone else.
Just 2 years later, the political spin machine orchestrated a hate campaign which forced Fred the shred, (see - even I'm doing it,) to leave the country, to hide the fact that the innumerate Messrs Balls, Miliband and the great high priest of mathematics, Gordon Brown, (massively qualified for a numbers job, with his 10 year PHd in history - how useful for a chancellor) had caused it all. It's always the same, Robin Hood complex destroys the world.
It's politicians, pure and simple. Though since I didn't see a single economist predicting the crash and then profitting from it, I don't really rate their talents either.
I've offered to sell the solution to the government to the world's economic crisis, so I can retire and live without their interference, but they don't want to buy it. They'd rather print 500 Billion at a time. That's PPE for you.
"It was overturned because the big banks could not compete with the financial trading firms which were not banks but were offering banking services like checking accounts tied to their customer's brokerage accounts."
and the guy that spearheaded this change is US gov legislation......
well he decamped into GoldmanSachs shortly thereafter.
obviously no conflict of interest or potential corruption there then!
What a keg of toss.
The best thing about legacy systems is they've usually had 99% of the bugs squished out of them 30 years ago. The next best thing is that the senior devs and platform SMEs have worked these systems since they were in short pants.
Who would be mad enough to update their banking platform every few years as new IT fads come and go?
A true legacy system is one that is written once and then maintained and has all the bugs squished out. Banking systems aren't that. That's because they keep having to be updated to handle new regulations, new types of accounts, new marketing methods.
It's also well and fine to have a legacy system, but as changes are made the code is twisted and massaged and ripped up and re-written. This leaves the original neat properly designed code a total mess. When it's written in a high level language then it's easier to stop the code becoming a pile of doo-doo. When it's in assembler, then all bets are off and the code is guaranteed to be a mess after a few decades worth of tweaking.
It doesn't require a new platform every few years, but it does require monitoring the computing environment and deciding to switch at the best time. For example, banks say that have to use assembler to get through all the days transactions overnight. Straight away you can spot the problem. Not real time. So how do other companies with millions of customers and millions of transactions (eg twitter) handle such large volumes. They don't use assembler is a clue.
A relative of mine used to work at Lloyds nee Halifax nee Leeds. The description of the assembler code was nightmareish, let alone the description of the programming practises. Imagine that when you are taught assembler at Lloyds that you have to use pen and coding sheets, no editor on a PC allowed. That's how bad it was (or is).
@The Axe, sorry but wrong.
Standard practice for financial services is to keep existing code and run a system alongside for new functionality/regulation, rather than make changes.
Banking isn't realtime because it never was realtime, in order to maintain the legacy systems which are proven and reliable, you have to maintain the legacy ways of doing things. Look at faster payments, it's practically real-time, but runs as a batch.
I know several banking assembler Devs quite well. They have core routines that do not change very frequently to do the real hard work. They don't just hack in there every time the business has a new requirement.
Banking is so different to Twitter that I can't even refute your argument. It makes no sense :)
I can accept that the core routines don't change much. Money in, money out. What does change is all the interest rates, in different bands, in different accounts, etc. When that kind of stuff is hardcoded into assembler code rather than table driven or horrors in a simple database then you are asking for trouble when it comes to updating.
The only similarity is the number of transactions, everything else is so different. And no I'm not saying code the higher level non-core aspects of banking IT in php, even I know that's stupid. COBOL has it's place.
I totally disagree with your comments on legacy assembler in banks' IT: we use it for our real-time systems where response time for an online transaction is key, overnight batch processes chug away in COBOL.
Yes, 35-year-old legacy assembler needs to be managed to avoid it turning into a mess, but the quality of our codebase improved over the decades, even with a high rate of business driven change. The point at which things went wrong was when generic management took over and started outsourcing to bullshit merchants.
The reason for starting trainees with coding pencils and pads is to force them to think first, not just sit in front of an editor and keep tweaking something until it works, or cut&paste from someone else's work.
How do Twitter cope with their volumes? Well, they deal in the trivial and no-one expects them to exist in 40 years.
Good to hear that assembler is only used for core stuff at your place. Doesn't sound like Lloyds.
Any kind of code can be maintained so long as the original people are still there and new staff are slowly (very slowly) brought up to speed. But when the code is a mess, and the gurus does their best to make themselves indispensable by keeping the knowledge to themselves, then the moment they leave (or die) then the bank has problems. I agree that outsourcing to bullshit merchants is a bad thing, but when the code is mess that outsourcing is guaranteed to fail.
However your argument about pencil and paper is just soooo stupid. So how do they do they up front thinking? Upfront thinking is designing and that can either be done be scrawling on paper with a pencil or using some editor to do the hard work in flowcharting for you. Why not in an editor which allows them to easily correct mistakes and typos. And tweaking until something works means that they come across the mistakes from which they learn from. I learn more from making mistakes and bugs than I do if I was told how to do everything.
Copy/paste is wrong no matter what because it is so easy to miss something in the pasted code which needs to change and is then so hard to find afterwards because you assume it perfect code.
I didn't make my point about pencil & paper very well:
I wasn't advocating its use for day-to-day development, just as a training tool.
You would probably be horrified to see what passes as an editor for our devs: We'll have no truck with these new-fangled IDEs, green-screen terminal emulators rule! Don't press the 'transmit' button at the wrong position on the screen, or you might lose half your code!
re-engineering their core systems, but it's hard to justify the cost of doing this rapidly.
They'd rather gouge their own eyeballs out with a spoon than use the cloud though as they just don't have the level of control that they need. What they WILL do is move to message busses and Software as a Service, but that's a long-term project to complete.
True and can bet IBM or somebody will sell them - hey you see that CIC's transactions, well if we wrap that up with MQ and push down a SNA channel to a UNIX box we sell you can then process the transactions upon that and gradyuly move your mainframe bits of onto cheaper box's and do it will all work. WHilst a good appraoch and very sane given how many applications interwork on mainframes in such enviroments it still needs compitent people and by default upgrading a OS on said AIX box will stamp over the MQ transaction queue's due to there default location being non too ideal. This means they need people who know what there doing to make it work and not just people who appear to know what there doing. End result is overruns and deadlines missed and that cheap UNIX box option starts to become not as rosey afterall compared to just adding a CPU or some upgrade to the mainframe there trying to move away from.
Now todays trend is to use the word cloud and sell things on the back of that fashion statement, sadly nobody wil have the balls to say the mainframe is a BFC (Big Funken Cloud) already. Sadly the people who make these desicions read the big adverts on back of trade mags and website and not the content and facts behind them. But nothing new there.
The whole move away from the mainframe brigade been around since the 90's and still singing old tunes. SO your message busses and software as a service is already out there and happening and failing and falling for years and less often works out well.
About the only thing alot of companies have managed to do is to copy there database overnight to another cheaper system so the mungles can abuse it running reports without messing up the golden goose mainframe, which has been a blessing to many a operator I suspect.
Sorry that's a load of utter BS.
IBM will attempt to see the bank a net new banking model that they helped create for an Irish Bank to replace all core banking systems.
This way IBM's Global Services can rake in millions in terms of net new hardware, software and services dollars migrating the bank from their existing core systems to something that might work...
They think the "cloud" is real! What a bunch of Merchant Bankers. They don't realise that the "cloud" is just a large data centre with a gazillion servers in it.
In other words, EXACTLY WHAT THEY HAVE ALREADY!
I'm speechless. I imagine the banking IT guys are reading this and pissing in their pants!
"Cloud" is someone else's data centre packed with general purpose machines and configured in a fairly dynamic way.
"Private Cloud" is the same kind of thing but in your own data centre.
What the banks have are data centres split up by security zone inside of which are a mix of general purpose and highly specialised machines. They might, for example, have boxes with accelerator hardware to perform HTTPS connections and encryption. They will almost certainly have Oracle RAC database clusters for fault tolerant, high performance database work. There will probably be a mix of mini and mainframe boxes scattered about for assorted legacy systems etc.
Heathen! Thou art not of the Body!
No software shall ever be required to be "fit for purpose"! So it has been since the beginning of time, so shall it ever be!
Renounce thy sins! Repent and be forgiven!
Else we shall banish you from the community forever.
Intellect's been around for 10-12 years. I've been to some of their workshops. You have to remember that they are an organisation of IT vendors and related suppliers (lots of lawyers for some reason) so they are usually going to recommend something that involves spending lots of money. They tend to exist as a way for government departments to find out how much their latest bright idea is really going to cost (I remember a workshop where the Dept of Schools and Families wanted to give every kid a laptop and were sitting there agape as a guy from HP was totting up exactly how much this meant in support costs, spares and loads of stuff they hadn't thought about). There is a tendency for those meetings to get hijacked by salesmen and entrepreneurs with axes to grind though, it looks like that might have happened here.
When compared to the amount of money to be made from speculation, the amount of money to be made from retail banking is miniscule.
Of course bankers appear to have a hard time figuring into that calculation the amount of money that can be lost from speculation - et, voila, we have our present crisis.
Back to the first line: This is why banks will invest in systems to aid in their specualtion but not in retail banking systems.
Next week I am also releasing a 60 page document. It will explain that banks have failed to invest in the one key asset that will in future prevent the global banking system from collapsing.
They are called employees, but most senior managers probably refer to them as excessive operational expenses.
I can understand (and would expect) that an IT industry trade association would suggest that the solution involves buying more hardware and software.
The management of the banks are (or should be) in the best position to determine if their systems are fit for purpose. The "purpose" is to support business processes, provide "customer service" and address requirements imposed by regulation. Whether these requirements are met by 30 year old mainframe based systems or brand new kit is not really the issue.
Regulators should not be telling businesses how to configure their infrastructure. [If they do, then I can't wait for the reaction from the water companies when they are told to replace the Victorian pipes and sewers]
(Old joke warning) Q. How can you tell when a salesman is lying? A. His lips are moving!
The regulators job is to ensure that the banks are fit and proper institutions to hold a banking license.
Part of that assesment in the 21st century should be are they actually capable of delivering the banking services.
It is frankly unaceptable for a quarter of the UK banking system to be offline because one bank can't adequately staff, train, equip and have appropriate BC/DR plans in place, and this really should be included as a factor, like keeping appropriate records, when considering should the bank keep it's license.
Whilst the regulator should not say what equipment should be used, they should be setting the standards expected, and auditing to ensure they are met.
I agree that the public have a right to expect not to suffer inconvenience as a result of a bank's failure to plan/train/etc, but I think that banks are having to steer a difficulty course:
The expectations have changed dramatically since a lot of these systems were put in place. We've gone from a culture of half-day closing on Wednesday to always-on 24*7 online banking. There will never be a convenient time to pause and put in a proper modern solution, even if you could buy an off-the-shelf retail banking system which wasn't vapourware, what would the regression test plan for its implementation look like?
The only course is gradual evolution of the legacy systems. The barrier to this remains senior management unable to understand that cheaper != better when it comes to developers.
... what is due to a deficient human mindset. Again.
What needs to be fixed is the short-term-focused and greedy logic of the current market system.
Of course that's a lot more difficult, and it doesn't make many people wealthier, while making those in power poorer, so what's the chance of this happening?
The legacy systems are falling over purely because many of the staff who are capable of successfully maintaining and extending them have been replaced by cheaper sub-standard offshore staff. I suspect that if a fraction of the monies that went to pay fat banker bonuses has been spent on keeping the original staff and strengthening their teams with indigenous well-trained staff, we wouldn't be in the mess we are in now. Another example of bankers misappropriating funds and stuffing their own pockets at the expense of service to customers.