Financial Pricing of Software Dev...
0 74 0 -74 5 9 / 10 / $ 2 6 . 0 0 �� 2 0 10 I E E E September/October 2010 I E E E S O F T WA R E 65 ������How���much���would���the���project���cost���change���for��� one���unit���deviation���of���each���risk���factor? ������How���much���should���a���project���manager���be���willing��� to���spend���to���control���each���risk���alone? Finance��� practitioners��� face��� similar��� questions��� about��� economic��� risk��� factors,��� such��� as��� the��� follow- ing:���If���inflation���is���one���half���a���percent���higher���than���ex- pected,���what���would���be���the���expected���change���in���re- turns���on���inflation-sensitive���assets?���To���answer���such��� questions,���finance���relies���on���risk-pricing���information��� comprising���asset���returns������sensitivity���to���a���risk���factor��� and���the���premium return���expected���per���unit���expo- sure���to���that���factor.���The���analogy���to���software���devel- opment���is���straightforward,���where���project���costs���and��� premium��� cost��� replace��� asset��� returns��� and��� premium��� return.��� A��� crucial��� advantage��� of��� risk-pricing��� infor- mation���is���that���the���risk���parameters���are���standardized��� across���projects.���The���risk���premium���for���each���risk���fac- tor���is���universal���to���all���projects,���and���the���sensitivity��� to���each���risk���factor���is���common���to���groups���of���similar��� projects��� (systems��� software,��� scientific��� applications,��� and���so���on).���Hence,���these���two���risk���parameters���can��� serve���as���economic���risk���benchmark���measures. In��� this��� article,��� we��� review��� fundamental��� risk��� and���risk���measurement���concepts,���and���we���present���a��� method���for���pricing���software���development���risks���that��� estimates���these���two���parameters���for���each���risk���factor. Application Areas for Risk Pricing Risk-pricing��� information��� is��� useful��� in��� several��� ar- eas���for��� example,��� the��� assessment��� of��� risk��� for��� a��� portfolio���of���projects.���This���assessment���is���based���on��� a���basic���principle���from���insurance:������The���only���way���to��� successfully���manage���project���risk���is���to���manage���risk��� across���your���organization���s���entire���project���portfolio,������ because���insuring���any���single���project���against���all���its��� risks���would���make���its���price���prohibitive.1���Given���risk- pricing���information,���comprising���each���project���s���sen- sitivities���to���different���risk���factors���and���the���premium��� cost���per���factor,���we���can���calculate���the���portfolio���risk��� as��� a��� function��� of��� the��� sum��� of��� smaller��� project-level��� risks���and���the���projects������costs: RP Ci i j factor projecti = ��� ��� ��� ������ ������ ������ ��� ��� ��� ��� ��� ��� ��� �� ��j ,j where���RP���is���the���portfolio���risk,���bi,j���is���the���sensitivity���of��� P ricing��� individual��� risk��� factors��� could��� benefit��� software��� development��� decision- making���because���not���all���software���development���risk���factors���are���equally���critical.��� For���example,���suppose���the���analyst���capability���and���platform���experience���levels��� assigned���to���a���specific���project���deviate���up���or���down���from���initial���expectations.��� Risk-pricing���information���permits���answering���such���questions���as���the���following: This method for pricing software development risk estimates two parameters for each risk factor: extra cost incurred per unit exposure and project sensitivity. Michel Benaroch, Syracuse University Ajit Appari, Dartmouth College Financial Pricing of Software Development Risk Factors software development risk