The conceptualization of risk derived from Modern Portfolio Theory, developed by Harry Markowitz in 1952, is presented as the seminal theory insofar as project portfolio risk is a concern [28]. Modern Portfolio Theory has been used to establish a parallel between financial portfolios and project portfolios, specifically focusing on risk incorporation into the project portfolio selection process [17,19], thus adopting a general risk management approach [20]. These studies are focused on the risk associated with financial results, generally defining the risk as a unidimensional measure. In this regard, the risk is associated with variations in the portfolio’s financial results due to the variation derived from each project and, by considering the correlation between each pair of projects. In other words, total portfolio risk can be considered to be composed of independent project risk and interdependent project risk [19].

When looking for risk approaches oriented more towards project portfolio characteristics than general management, it has been posed that portfolio value cannot be measured only in monetary terms [13,29,30]. For instance, proposals oriented to integrate both monetary and non-monetary factors have been made [30]. To this end, a discrete comparison of alternatives has been explored [1,30], where attributes associated with the risk for each project are considered as criteria in the portfolio selection process. Other proposals have focused on estimating the risk as changes in the scope, time, or cost of projects due to some risk factor interdependency and project interdependency, or as changes in the cumulative cost of the projects within the portfolio [20,22,23,31]. Hence, a perspective on risk in which the ‘project portfolio risk’ is associated with the success of the projects within the portfolio is proposed here.

For this reason, these proposals have been developed under a project management approach [20], recognizing, in most cases, the multidimensional nature of project success in a way that tacitly considers the ‘project portfolio risk’ as a measure of impact on the success of the projects within the portfolio due to project interdependency and changes in project ‘risk factors’ during the project portfolio execution. For instance, considering interdependencies between projects, several studies have focused on how the risk is propagated across the projects within the portfolio [22,23]. However, portfolio-level ‘risk factors’ and the condition that the ‘project portfolio risk’ goes beyond the sum of the individual risk for each project within the portfolio, although highlighted, have not been explicitly incorporated into risk conceptualization under this approach.

Hofman et al. [5], Ghasemi et al. [9], and Bai et al. [12] present the identification and categorization of project-portfolio level ‘risk factors’ from a full spectrum of project portfolios. In addition, Ghasemi et al. [9] pose a risk assessment model oriented towards establishing the ‘project portfolio risk’ as the impact on the ‘project portfolio success’ as a unidimensional measure due to changes in project- and portfolio-level ‘risk factors’. Thus, extending the fact that project portfolios represent the bridge between projects and organizational strategy, an alternative approach for ‘project portfolio risk’ emerges which would allow for incorporating the specific portfolio level into the analysis, i.e., portfolio-level ‘risk factors’ and ‘project portfolio success’ as the expected portfolio measures that are impacted by the ‘risk factors’.

The adoption of this alternative approach oriented explicitly towards the portfolio level could help disclose the tacit role of the ‘project portfolio risk’ as the bridge between the project-management- and general-management-based risk approaches identified in the literature.

Risk categorization contributes to the effectiveness and quality of risk analysis, as well as leading to a risk-based decision process [4,5]. Hence, different categorization schemes for ‘risk factors’ associated with projects within a project portfolio have been identified in the literature. These categorization schemes have mainly aimed at classifying ‘risk factors’ by their sources. Alternative categorization schemes have recently been posed, for instance, the identification and categorization of critical risks to project portfolios over the life cycle [12].

According to the structured literature review conducted by Micán et al. [21], risk factor categories based on risk factor sources could be synthesized into four categories: (1) the Project Portfolio Management (PPM) level, considering aspects related to portfolio information management, conflicts at the portfolio level, and lack of portfolio management capabilities; (2) Project interactions, considering the resource interdependency between projects as ‘risk factors’ and the dependency of outputs between projects; (3) External conditions, in which are grouped ‘risk factors’ derived from external conditions to the parent organization; and (4) Organizational conditions, considering aspects related that organizational decisions or processes, including project management process and decisions related to the structure of the portfolio, which can influence the project portfolio results.

Alternatively, the conceptualization of systematic and non-systematic risks derived from Modern Portfolio Theory (MPT) has been adopted to conceptualize the ‘project portfolio risk’ sources [32,33]. This means that, to classify the ‘risk factors’ by the extent of their impact, the non-systematic risk is associated with individual risk contributions from each project risk factor, and systematic risk is derived from those ‘risk factors’ which have an impact on the whole project portfolio [17].

However, the representation of systematic risk has different interpretations; for example, Costa et al. [32] pose it as being exclusively associated with environmental factors, while Drake and Byrd [33] mention that, besides the factors of the environment, it includes the risk derived from the interdependency of ‘risk factors’ and project interactions as risk factor sources. In addition, although the conceptualization of systemic and non-systemic risk for project portfolios has been widely adopted for incorporating risk into the project portfolio selection process [17], it is scarcely considered for risk analysis in the project portfolio execution phase.

Five ‘project portfolio success’ dimensions are generally described in the literature [34,35]: (i) strategic fit, (ii) future preparedness, (iii) average project success, (iv) synergy exploitation, and (v) portfolio balance, where success dimensions i, iii, iv and v are based on the PPM goals defined by Cooper et al. [36] for portfolios of product development projects, which later were adopted for other types of project portfolios. 

In addition, the dimension related to economic success was incorporated in some works as a ‘project portfolio success’ dimension [34,35], while other publications regard this dimension as a business success dimension, and therefore a positive consequence of ‘project portfolio success’ more than a ‘project portfolio success’ dimension itself [37].