Sometimes, the analysts are able to

Sometimes, the analysts are able to motivate the established practice and the choices that thereby have become, ina sense, routine. In other cases, the analysts inherit choicesmade by former colleagues and are not informed of why theywere made. The reasons for certain choices are sometimesforgotten. Some choices are even made randomly. Focusingon ten companies in particular, as in the above, is an exampleof a choice somewhat randomly made.Teams with more staff often make attempts to systematise the analysis practice. They strive to make the analysispractice less dependent on the individual. This can beachieved by socialising the individual analysts. New analystswould, for example, conduct their first analyses togetherwith experienced colleagues. This way, shared ways of doingand reasoning can be established and make the analysisprocess less dependent on individual analysts. Another wayto systematise the practice is to turn processes into technology. One team received calculated scores of companies’financial attractiveness from their portfolio managers. Therewere also, for example, calculated scores of industries’attractiveness. These calculated scores could then guidethe analysts and the attention they devote to industriesand companies. The less attractive companies or industriescan even be removed from their ESG analysis. In less customised but also highly systemised approaches, the delimitation can instead be based on an index,such asthe MSCI World.According to such a definition of the epistemic object, the400 S. Du Rietzperformance of all companies on the MSCI World index shouldbe included.How the analyst team defines what the epistemic object isand consists of has consequences for other aspects of theirpractice. One such example is the choice of service provider.Though all respondents inevitably collect accounts that willbe used as information, the amount and type of accounts varyaccording to the time and energy the teams devote to it.Typically, asset owners rely on service providers and theirinformation or analysis products. These products consist of,for example, databases with companies and performancevariables or written analyses. Such products, in turn, arebased on a larger and more diverse set of collected accounts.Though price and deemed expertise are important factorswhen buying these services and contracting a service provider, a provider may also be chosen because its product fitshow the team has defined what the epistemic object is. Evenif a service provider offers what are considered high-qualityaccounts, the service needs to cover the entire epistemicobject, i.e., the performance of all the chosen companies.I wouldn’t claim that they are less good analysts, I don’tthink so. But all the surrounding things you could get andthe documentation you could get and the systems youcould get, the integrity of the product, was undoubtedlybetter at [service provider]. (. . .) There are other playerswe have had to deselect because it [the analysis] is onlyEuropean companies and not the entire world. Theiranalysis may be good, and the reports may be good,but they don’t deliver all the things we want, and thenit won’t work.Hence, defining what is an epistemic object helps reducethe amount of potential information, but there is a possibilitythat accounts covering all of it cannot be found. In the above,this was an issue for the analyst team when selecting serviceprovider. When you have defined what the epistemic objectis, you also need accounts to inform you of all of it.