Dave founded the PCI Knowledge Base, a research community that shares information to help organizations achieve PCI compliance, after a 14-year tenure as an analyst at Gartner Inc. He was a wealth of information for my articles on PCI and provided his expert advice in an article we published this summer. Dave was always ready to help and never shied away from expressing his opinions. I really enjoyed working with him and will miss him, as will the industry.]]>
And so it goes with this week’s post. Here are some nuggets that I’ve gathered over time:
Policy and procedure: I was talking to a client today about password reset lengths. Turns out for one of their products they changed the password frequency to expire after 1,000 days. Their logic was that it was low risk because the application didn’t store NPPI and the security was really only necessary to ensure proper segregation of duties. So I asked them if they had a password policy (they did) and if so were they in compliance with the policy (they weren’t). After a momentary silence, their quiet reply was “good point.” Being the auditor that I am I couldn’t help point out that the worst thing any institution could do was to deviate from a documented policy or procedure, regardless of the reason. Once an examiner discovers something like that, they figure it’s an indication of related issues and wind up digging a bit deeper. Document what it is you do and than make sure you’re doing it; while it may seem simple enough, you’d be surprised how many companies fail on that point.
Pandemic planning: There’s still heightened concern regarding the swine flu and my industry continues to beat the drum about needing to have a pandemic response plan in place. While it’s a valid point, I’ve been polling my clients over the past few months regarding their first hand experiences with the flu epidemic. Only a few have been confronted with any legitimate outbreaks and none of them have experienced an absentee rate that required unusual planning or intervention. While I’m not advocating that a pandemic response plan is superfluous, I am questioning my peers who are pushing this as a top of the list agenda item. For my money I’d rather spend time making sure that a properly vetted and tested business continuity plan is in place and spend less time and effort getting caught up in the hype.
SOX: Banks that are required to be SOX compliant need to take some time to make sure that they’re thinking things through. GLBA is a fairly rigorous and encompassing regulation and extends deeply into a financial institution’s infrastructure. To a certain extent, it serves to drive a bank’s general controls framework, be it informal or otherwise, and as a byproduct goes a long way towards establishing controls typically associated with SOX. So when I encounter clients who are tackling SOX as if though it’s its own separate set of requirements I throw up the caution flag and try and force a reset. While it may be true that larger institutions need to extend significantly from GLBA to controls around financial reporting within the infrastructure, that would only represent a subset. Before doing anything different, the bank should bring in someone who has experience working with both SOX and GLBA to identify the (many) commonalities and produce a consolidated framework so that efficiencies are both identified and realized.
Year-end activities: In my last post I discussed how there’s an uptick in services work this time of year when many banks and credit unions remember that they still need to conduct a wide range of audits and assessments in support of GLBA/NCUA regulations. If you spend some time reading through FFIEC guidance (seriously, it’s not nearly as dry and boring as you might think) there are multiple references to “your most recent audit or assessment.” For those of you who think that the need to conduct this work is suggested rather than required, consider how it looks to your examiner(s) when they discover that your most recent risk assessment was either conducted several years ago or not at all. Do you really think it reflects well on your institution that you haven’t taken a serious look at the myriad risk factors swirling about your infrastructure for any considerable length of time? In a day and age when new threats emerge almost daily if not hourly how can you justify neglecting such a critical task? The examiners expect a current set of reports not only because it’s required but also because it’s a clear indication of solid management and oversight activities.
And on a final note, I’d like to share this link to the FDIC website. You’ll find a video message from Chairman Bair on the current state of both the FDIC and the banking industry. It’s really more of a “happy recap” (with all due respect to Mets fans) of similar messages she’s released over the last year. But I think it’s worth your time (about four minutes total) to hear it for yourself and gain a sense of calm about the security of your own deposits. And for those of you who might think I’m keeping to some sort of schedule regarding Sheila Bair references, as long as she keeps doing the right things I’m going to keep bringing her name up.
But this leads to another interesting quirk about how the examiners often operate. Generally speaking, if the reports are available, they don’t dig much (if at all) beyond the reports contents. And so the information security and IT components of many exams become more about inventorying recent reports and not much else. We see evidence of this all of the time when we conduct a first-year audit or assessment and discover gaps or issues that have been in place for years and which the exams never picked up on.
I’ve written in the past about how surprisingly few institutions maintain a current business continuity plan and even fewer properly test that plan. But what surprises me more is that these conditions have existed for years spanning many exam cycles. How is that even possible?
I’ll tell you how: There’s a documented plan that is provided upon request and by and large the examiner conducting the fieldwork checks off that they received it and voila, you have a non-issue. And because the people in the field are typically given too few hours to cover too much landscape, they don’t have enough time to dig in deeper. Sometimes it happens where an examiner happens to actually open the document and vet it for key details – every now and again we come across a DoR or an MoU where the absence of a recent business impact analysis was tagged – but that happens almost never.
I’m fond of advising clients that you conduct much of the required compliance work for one of two reasons: You do it because it’s the right way to manage your institution and reduce your risk or because you have to. Because of the approach taken by examiners, way too many institutions lean towards the latter and simply want to have a report available to hand over when asked. But is this really the right way to run a financial institution?
And when you consider that the value of the report is largely defined by its contents and the competency of the practitioners conducting the fieldwork behind it, isn’t there an increased likelihood that there are important issues that go undetected? If all you do is pay for the report (often issued by the firm submitting the lowest bid) and all the examiners do is check off that the report was available and issued during the appropriate time frame, is there any real value in even bothering with this process?
I’m a bit biased regarding the value of reports. My firm is on a constant hunt for real risk and not just simply working our way down a checklist to kick out a document and collect our money. We tend to examine our clients infrastructure as if though we have our own money deposited with them and tie what we see straight back to GLBA and NCUA requirements. The value in this approach is that we produce a report that the board of directors can relate to, not just the IT folks.
But again, if no one really even cares about the content of the report and only that it exists, why bother doing a good job?
Maybe our industry needs to adopt an approach similar to the PCI folks. Maybe the FDIC and NCUA should issue certifications to practitioners validating them as properly trained and educated experts with regards to GLBA. There would still be a variance from firm to firm to a certain degree, but at least there would be a recognized standard and an increased likelihood that if an examiner is going to rely on the competency and completeness of a report there’s some justification behind that decision.
Something’s going to have to change though and hopefully sometime soon. Because using the “last minute” logic is flawed and only serves to reinforce my own bad habits.
Sometimes I think COBIT is used much the same way.
For those of you who aren’t familiar with COBIT, it’s a framework that has revolutionized the world of governance and compliance for the better. It was the only beacon in the vast, dark ocean of SOX insanity a few years back, providing much needed guidance for corporate America to follow and continues to serve as the best source when designing controls within the infrastructure. It’s comprehensive, well organized and when understood and applied properly, it can be very effective.
But it’s not akin to the Bible and it’s definitely not an IT audit framework or program.
And yet I often hear fellow practitioners dropping COBIT references like it somehow validates them as legitimate members of the IT audit club (which by the way is called ISACA and only requires an annual membership fee).
Just this week, I heard that someone discussed conducting a COBIT-based audit when asked about their approach to conducting an IT general controls (ITGC) audit. Two weeks ago, my partner asked me about an RFP we received in which the institution wanted to know if we based our ITGC audit on COBIT or any other recognized framework. It’s gotten to the point where the term “COBIT-based” has become ubiquitous within the IT audit domain. Years ago during the aforementioned SOX insanity, there was a running joke with a client in which every sentence was laced with a SOX reference (e.g. Good SOX morning, Happy SOX New Year, etc.). Now it seems as if though COBIT has replaced SOX in that regard.
Um, has anyone actually read the framework? I mean actually sitting down and reading it from executive summary through to ME4 (the last of the control objective areas in the PDF). And how many people have actually tried to implement COBIT as it’s intended to be used? It’s a mountain of information that requires a ton of analysis and customization prior to being implemented. And it’s not intended for organizations both big and small. For many of the community banks and similarly sized credit unions that I commonly work with, it’s simply overkill.
But again, it’s not an audit framework and it’s not an audit program. And it’s entirely possible to build out an IT controls framework and never once rely upon COBIT to do so.
By the way, for those of you who aren’t familiar with the IT Governance Institute (ITGI), it’s a research think tank that exists to be the leading reference on IT governance for the global business community. In the time since COBIT made its inroads into corporate America and the audit vernacular, ITGI has amped it up a notch. Now they also publish Val IT and more recently Risk IT.
So now I’m bracing for the onslaught of risk assessments that are “Risk IT” based. But I never had a problem conducting a risk assessment before this standard existed and I doubt I’ll crack it open when conducting one in the near future. Did we really need this? And how will this drive the audit and compliance industry?
Frameworks have a place in this world, don’t get me wrong. But it’s like when I bought my Roto Zip hand saw a few years back; I walked around my house looking for things I could use it for rather than simply using it when it made sense. COBIT is awesome and it’s helped provide clarity in many, many ways. But it isn’t the official book of record for audit and compliance within IT; it’s just another tool in the toolbox. I realize that on the planet of ISACA that’s akin to blasphemy, but I offer no apologies. I refuse to build an audit program for a community bank that’s supported by two IT resources based on the 200 plus control objectives in COBIT.
And on that note I bid you a good COBIT day.]]>