Questioning the Need for Separate IT Risk Management Frameworks
Total Page:16
File Type:pdf, Size:1020Kb
Questioning the need for separate IT risk management frameworks Nicolas Racz1, Edgar Weippl1,Andreas Seufert2 1Institut fürSoftwaretechnik undInteraktive Systeme TU Wien Favoritenstr. 9-11/188 1040 Wien,Austria 2Institut fürBusiness Intelligence, Steinbeis Hochschule Berlin {racz, eweippl}@ifs.tuwien.ac.at; [email protected] Abstract: Thegrowing importance of enterprise risk management andthe resultingintegration efforts put theneed for separate IT risk management frameworks in question. In this research we analysecommonand distinct elements of theCOSO enterprise risk management andISACA RiskITframeworks. The analysis affirmsthe hypothesis that separate IT risk managementframeworks are redundant. 1 Motivation The alignment of IT with business objectives is an important part of contemporaryIT management. Ever sincethe creationofthe terms “enterprise risk management” (ERM) and “governance,risk, andcompliance” (GRC) the searchfor integration possibilities within these disciplines has been ongoing. Howeveras of todaydifferent frameworks are used forthe management of business risksand IT risks. The emergenceofhorizontal integration(across disciplines andacross departments) andvertical integration(across the organisationalhierarchyand across processlevels) has helped to realise that formerly separate approaches areoften redundant[Mi07],which provokes the authorstoestablish the hypothesis that a separate management of IT risksmight not be justified. 2 Status quo in research andpractice A quick scan of the ACM, SpringerLinkand EmeraldInsight databases showsthat in research as of todayenterprise risk management andITriskmanagement (IT-RM) have hardlyevercrossedpaths. FoleyusesERM processestomanagesecurityrisks [Fo09].In their high-level processmodelfor IT GRCmanagement, Racz et al. [RWS10]use COSO ERM[CO04] as risk management standard, assumingthat the selectionofaframework that does not focusonITwould facilitate integrationwithnon-ITGRC in future research. 245 In practice enterprise risk management andITriskmanagementare also treated as separate topics. With ISO 31000:2009[ISO09] (superseding AS/NZS 4360:2004 [AS04])and ISO/IEC27005:2008[ISO08] the InternationalOrganizationfor StandardizationtreatsERM andinformation security risk management (including IT- RM)intwo distinct standards.ISO 31000 does not even referenceISO/IEC27005.The alignmentofITwithbusinessinpractice is mainly done throughthe IT governance and management frameworks COBIT(Control Objectives forInformationand related Technology [ITG07])and ITIL (IT Infrastructure Library [OGC07]). These frameworks suggest enablingalignment through deriving IT goals from business goals. We canconcludethatwhile theconnectionofITrisks with business objectives is enforced at present, the merger of IT-RMwithERM on a processlevel is hardlylooked at. The frame of referencefor research of integrated GRC[Ra10]recommends identifying integration possibilities on the strategic, process, organisational and technology level. Strategically, throughthe alignment of IT goals with businessgoals, the integrationisalready ongoing. We suggest to take the next step andtoreview potential synergiesofERM andIT-RM on the processlevel.Following the claimof ERM to coverall risksofanenterprise, IT-RM should eitherbe completelycovered by ERM andtherefore be redundant; or it might enhancethe broaderERM throughdetailed considerationofITspecifics in the risk management process. 3 Methodology As a first step in evaluating ourhypothesiswe decidedtocarry outanexemplary comparison of an ERM frameworkwithanIT-RM framework. Of course a comparison of twoframeworksisnot representative,but we selectedwidely-used frameworks (see below) that therefore suffice to provide afirst indicationabout the hypotheses’ validity. The results should then be discussedwithother expertsat the Informatik2010 GRC workshop before takingfurther action. The methodology applied consists of four steps. First, we selected aframework forERM andone forIT-RM.Second, the frameworks’ commonalitieswere identified.Third,weanalysedthe references of the ERM frameworktoITriskand vice versa. Finallywe discussedand summedupthe results. In the selectionprocess foranERM framework we considered ISO 31000:2009 and COSO ERM, twowell-knownstandards forERM.Their processmodels are very similar.Onahigh level they only differ in theirwording.“Establishingthe context” in the ISO standard corresponds to the “internal environment” of COSO ERM,“risk evaluation” and“risk treatment” equal “riskresponse”and “control activities”,etc. Eventuallywe opted forCOSO ERM, as it is the successorofthe widely implemented COSO frameworkfor internal control[CO92], a de-factostandard explicitly acknowledgedinthe US Public CompanyAccounting Oversight BoardAuditing Standard No.5forfinancial reporting[PCA07].The standard is referenced in the Sarbanes OxleyAct of 2002, whichofall regulations passed in the newmillennium probablyhas thestrongest impact on risk management andinternal control systems. 246 ForITriskmanagement we chose the ISACARiskITFramework because it complementsCOBIT,which is arguablythe most appropriate control andgovernance frameworkusedbymanyorganisations world-wide to ensure alignmentofITand business goals [RYC04]. The framework’simportance is expected to grow sincethe new COBITversion 5, which is currently in development, planstoconsolidate andintegrate the Risk IT framework[ISA10]. ISO/IEC27005:2008 wasalsoconsidered. As it includes all aspects of informationsecurity(includingnon-ITaspects), its scope surpasses the ISACAframework,which is limited to information technology.Inour opinion Risk IT is more detailed,and it drawsout the specifics of IT-RM more clearly. The identificationofthe frameworks’ commonalitiesinthe second phase of ourresearch wasdonethrough a mappingofthe described processesofISACA RiskITtothose of COSO ERM. The documentation of COSO ERM proved to be a hurdle. On the highest level the frameworkconsistsofseven processes andthe “internal environment” component. Unfortunatelythe processes are notbrokendown. Instead COSO just names the basic sub-components,suchas“risktolerance” or “inherent andresidual risk”. In ordertomap the processes of ISACARiskIT, we hadtogothrough the complete descriptionofthe COSO componentstofindifthe same processes were included. The qualitativeanalysisofreferencesfromERM to IT-RM andvice versa in the third research step wasfollowedbyadescriptive discussion andsummary of the insights gained in the researchprocess. 4 Results and discussion 4.1Mapping of ISACA RiskITtoCOSOERM Ourcomparisonofriskmanagement frameworks is based on the assumptionthat “risk” in ERM has the same characteristics as “risk” in IT-RM. In COSO ERM, risk is “the possibilitythat an event will occurand adverselyaffect the achievement of objectives; events with a potentiallypositive impact mayoffset negative impactsortheymay represent opportunities” [Co04].Throughout the framework“risk” then also refers to upside risk (opportunities).According to ISACARisk-IT,ITriskis“acomponent of the overall risk universe of the enterprise [...]. IT risk is businessrisk[...]. It consists of IT- relatedeventsand conditions that couldpotentiallyimpactthe business” [ISA09].The twoframeworksconsequentlyshare a common understandingofthe term“risk”. ISACARisk-IT consists of the threeprocessesriskgovernance, risk evaluation, andrisk response on level one, with threesub-processes each on level two. Level threecomprises 43 processes. COSO ERM on the otherhanddescribes 8highlevel processes with 41 sub-components.While theERM frameworkismoreprofound on the internal environment component andonriskaggregation, Risk IT is more specific when it comes to IT specific andcommunicationprocesses. Still,all but sevenofthe IT-RMprocesses caneasilybe mapped to COSO components (see appendix A). 247 Twoofthe exceptions deal with ERM integration: “RG2.2: Co-ordinate IT risk strategy andbusinessriskstrategy”,and “RG2.3: Adapt IT risk practices to enterprise risk practices”. They treatthe alignment of IT andbusinessrisks on astrategic andona processlevel; we will analysethemlater on in the sectionabout ERM references in the IT-RM framework. Threeother processesthat couldnot be mapped belong to the processgroup “RG3: Make risk-aware business decisions”: “RG3.1: Gain management buy-in forthe IT risk analysis approach”, “RG3.2: Approve IT risk analysis”,and “RG3.5: Prioritise IT risk response activities”.Management buy-in forriskanalysis approaches andtheir approval is notexplicitly mentionedinCOSO ERM, butit could seamlesslybe integrated with the “internal environment” component. Prioritisationof response activitiesisprobablysoself-evidentthat COSO ERM does not highlight it; in COSO the prioritisationcould be part of risk response.Furthermore the processes “RE2.4: Performapeer review of IT risk analysis” and “RE3.3: Understand IT capabilities” do not existinCOSO ERM. Peer reviewsare a control mechanismthat can be seamlesslyincludedin the ERM process. Understanding IT capabilities is an extremelygeneral “process” that is a prerequisite forany kind of IT activity, therefore suitable to be addedtothe “internal environment” component of COSO ERM. As we cansee,drawing from the standards IT risksmay be treated like anyother risk,as the IT-RMframework is completelyabsorbedinCOSO ERM, apart from the ERM integrationspecifics (RG2.2,RG2.3)analysedbelow. The ISACAframework does not explain whyanIT-specific risk management frameworkinthe hierarchical relationship to ERM wouldbe necessary. It even disposes of the