= Subscriber content; or subscribe now to access all American Banker content.

Wanted at Banks: Young Tech Pros with Old-Tech Smarts

One item is rising to the top of many bank chief information officers' list of worries: the concern that one or a few key people who understand core legacy systems will retire or leave. When that happens, updating the system to meet a new or modified regulation or offer a new product becomes almost impossible.

This "key-person risk" is compounded by the fact that banks are having trouble finding talented young techies who want to work in a bank and a shortage of people with mainframe and COBOL skills. The mainframe was supposed to have been be replaced by farms of smaller commodity servers and cloud computing by now, but it still endures at many banks.

"It's really hard to get new graduates interested in those old technologies," says Bruce Livesay, the CIO at First Horizon National Corp., the $26 billion-asset parent of First Tennessee Bank in Memphis.

Key-person risk is a side effect of the U.S.'s aging population and the fact that many banks' legacy systems were built 40 or more years ago by people who are now headed for retirement.

It's not just a private-sector problem, says Bob Olson, a vice president at Unisys. One of Unisys' government clients "has someone who runs a chunk of their technology who's on oxygen, he's 70 years old, he knows the keys to the kingdom, he knows where everything is, it's all sitting in his head. They send out a police car to pick him up every morning and bring him in to work in a vault-like room."

According to IBM, 92 of the top 100 banks use mainframes and 71% of all Fortune 500 companies have their core businesses on a mainframe.

And yet few students are graduating college with mainframe skills, partly because schools are not teaching them, says Livesay. Those that do teach them tend to be technical schools and community colleges rather than major universities.

To try to jump-start mainframe enthusiasm among students, IBM has been running a "Master the Mainframe" contest for the past three years that's meant to "develop the skills of a new generation of mainframe experts." More than 68,000 students have gone through the program so far, the company says. IBM also runs a System z Academic Initiative program in which it provides materials for a 15-week IT course that's offered by 90 U.S. colleges.

Key-person risk is also a concern with core systems. In some cases, only one or two people understand the core banking software the bank runs on and the older programming language in which it is written, such as COBOL.

First Horizon's answer is to bring in young techies and train them on the older technologies.

"You don't want to be in a situation where you have that key-person risk," Livesay says.

Even so, finding young, talented tech workers who want to work for a bank isn't easy, he says. "It's difficult right now to attract even .net and Java programmers into the banking space," Livesay says. "The banking industry has gotten so much negative publicity through the past several years, it has made it more difficult to recruit people. We're seeing fewer people feeling motivated to get into banking."

Banks lack the allure that tech giants like Google and Facebook naturally have for college graduates.

"I don't think you can compete with Google, Twitter or Facebook on the coolness factor. It's just not doable," says Inder Koul, CIO at the $38 billion-asset First Niagara Financial Group in Buffalo. "It would be difficult for me to try to be a Google because I cannot match that."

However, he ticks off several advantages his bank does have in the competition for IT talent, including proximity to technical schools in Rochester, Syracuse and Buffalo and a chief executive who has articulated a strong commitment to investing in technology. CEO Gary Crosby has said that the bank will spend to $250 million over the next three or four years upgrading old technologies and investing in new digital banking platforms.

Koul adds that his bank has avoided key-person risk by making sure it has no situations in which only one person knows what's going on in an area.

First Niagara is also trying to help IT staff see a direct alignment with the business, whether it's infrastructure, business services, applications, or digitalization. "That makes it much easier for the rank-and-file tech talent to see how their work impacts others," Koul says.


(7) Comments



Comments (7)
I think the article is right on the money about the subject matter experts. In an industry where the server tier and front-end programmers are a commodity, there is no incentive for the mainframe programmer to cross-train or document. Similarly, there is an over-supply of mainframe programmers in the economy and pay is worse than even Java server side programmers; so, again a large disincentive for SME mainframe programmers to reveal their secrets. What can truly be said is they and those benefitting from mainframe were self conscious enough not to eliminate themselves. Will they change? Nope. Mainframe issue will only exacerbate. Especially in industries which view technical expertise as worth less than janitorial expertise.
Posted by timjowersab2 | Wednesday, June 10 2015 at 2:00PM ET
I just attended the IBM SHARE conference in Pittsburgh and spoke first-hand with both clients and IBMers alike involved in the zNextGen project, IBM Academic Initiative, and Master the Mainframe contest.

While all of these efforts are doing a fantastic job, they are just a drop in a very large bucket of the upcoming shortfall of mainframe talent in the workforce.

If banks are going to be able to staff these open positions, I believe there will need to be both an increased number of colleges offering these skills plus some banks will need to invest more developing internal training programs to provide these niche skills.

As I disclaimer I should mention that I am a partner in a training firm that specializes in providing zSeries & iSeries mainframe courses for new-hires and working professionals. We stand ready to help!
Posted by ProTechVPBusDev | Sunday, August 10 2014 at 1:57PM ET
Great article Penny. You've highlighted many resourcing issues that big businesses face with core business systems. It's certainly true that fewer universities today are teaching application development languages such as COBOL, and PL/I. For many organizations with significant application investments using these languages, it presents both a risk and a challenge.

There are technology solutions, however, that help bridge this gap. One such solution is that provided by modern COBOL. COBOL is now a .NET language or a language built for the Java Virtual Machine. This capability allows organizations to take existing COBOL applications to new platforms including Cloud, Mobile, and the Web. Modern COBOL technologies also allow IT shops to better maintain and enhance this code in the future. COBOL can now sit, side-by-side, with C#, VB.NET, or Java, all managed within a single development environment - Visual Studio or Eclipse. This modernization of COBOL allows organizations to more efficiently manage existing COBOL apps, but also attract the next generation of talent with modern IDE and language skills.

Micro Focus has also been heavily invested in the Academic scene through its Academic Program. Over 300 global universities are teaching COBOL in the classroom today and that number is growing every day. Universities that join this program receive free access to modern COBOL development software, COBOL courseware and documentation, as well as education services. More universities are bringing COBOL back into the classroom. For students, as well, this program offers free access to modern COBOL development tools for at-home, personal use. Full program information as well as a current listing of academic partner universities can be found at: http://academic.microfocus.com

I agree that there is more work to be done in improving the perception of COBOL, particularly amongst the next generation of developers. The integration of modern development tools, with the COBOL language, and active community participation with universities will help business organizations prepare and develop the next generation of talent to support these business critical applications.

Ed Airey, Product Marketing Director | Micro Focus
Posted by edairey | Friday, August 01 2014 at 5:13PM ET
I am fairly new to the banking industry, as well as the professional mainframe community, but it seems to me that mainframe technology isn't going away any time soon. I was fortunate enough to have attended a university that still includes mainframe courses in their curriculum, and those courses were by far the most educational and rewarding courses that I took as an undergrad.
I had great teachers for the mainframe courses I took, and I think that they did a great job at preparing me for what I would encounter in a real-world position. I also understand that there is a ton of information that has to be transferred from the existing personnel to the generation that will follow. I have been in the fortunate position where my predecessors have done nothing but prepare me for what is to come based mostly on what they have seen in their many years of experience, and that is an invaluable resource.
I worked hard to get where I am, and I am thankful for where I ended up every day. I certainly don't represent every mainframe programmer, but I just wanted to put my opinion out there.
One last thing I wanted to say is that although the article says that the Master the Mainframe competition has only been around for 3 years, I have competed in it personally for 5 years, and I have a friend who participated in it about 8 years ago, if not longer ago.
Posted by Banksauce | Wednesday, July 30 2014 at 12:09AM ET
I call Bullshit. Banks have the one thing to make jobs attractive. MONEY! You also have the power to create MONEY out of thin air. The choice is yours! If the job ain't cool, then you need to provide a tool for your COBOL engineers to be cool. The reason Banks are getting negative publicity is because it takes "spending" money to generate more money. Why are such intelligent bankers so damn stupid at the same time?
Posted by w1sdom | Monday, July 28 2014 at 9:28AM ET
I agree that there is a real technical problem with mainframe applications written in assembler. For those written in COBOL I think it is more a problem of application business knowledge, than one of mainframe and COBOL skills. I manage complex mainframe core banking systems, and what I have found is that these systems are very robust, with decades of complex business logic built into them. Contrary to popular belief, most of these systems are highly parameterized, with business change not requiring a programmer. What this means is that programmers are not needed as often, but when a programmer is needed, it is not to perform simple coding. It is to resolve critical financial and regulatory issues, or to make complex architectural changes. All of these require a deep understanding of the application's business logic, which is driven by parameters, not hard code. In short, maintenance of these systems has little use for "programmers". What is needed are "application architects".

This means that training new staff, to be truly effective, is often a years-long process, a difficult business case to build. Retaining the staff long enough for them to become productive is also an issue. Finally, these types of changes don't lend themselves to training assignments. The combination of these challenges results in continuing to focus on using the current experts, instead of growing the new ones. I believe this is why we have mainframe programmers begging for work, while at the same time legacy systems are starving for replacement talent.

Teaching a person to program on the mainframe can be solved in months, and is an important first step. Taking that programmer to a point that they can independently support a core banking application is the real problem to be solved.
Posted by gcgeil | Wednesday, July 16 2014 at 2:31PM ET
At first glance, the article makes sense; decades-old legacy applications run the core business and would be prohibitively expensive to replace, so they remain and need the people to run them. But then where are the job openings and help-wanted ads for computer professionals with those skills?

I say this from experience. After over 40 years of building and maintaining applications on the standard IBM mainframe and Burroughs (now Unisys) mid-range systems, my employer is moving off the mainframe platform and it's time for me to move on. I'm 62 years young and as far as I know quite capable of working for many more years. But the papers and Internet job search sites have very few openings for mainframe DB2 and IMS application developers, even in the New York City area which must be one of the most overbanked regions in the world. Nearly all the openings are for .NET, C#,Java, Hadoop and so on.

If the banks really need people to keep that old code working, they should let the job market know.
Posted by dmar112@freeshell.org | Wednesday, July 16 2014 at 11:43AM ET
Add Your Comments:
Not Registered?
You must be registered to post a comment. Click here to register.
Already registered? Log in here
Please note you must now log in with your email address and password.