Investments driven by theory, loyalty or peer pressure?

The fact that people invest on the basis of other factors than purely economic and financial is well known. In fact, there is a whole field of finance out there called as Behavioural Finance. But this is an interesting question. How do you invest your 401k? This paper sheds some very interesting light on this issue. I quote their conclusions here:

This paper addressed the issue of diversification in retirement funds. Empirically, employees tend to hold a dangerously high level of company stock as a proportion of their retirement investments. Instead of developing a theory of the optimal level of diversification, we take the alternative route of trying to find rational justifications for the observed levels of holdings.
Models of purely individualistic behavior usually deliver results that are incompatible with the patterns observed in the data. Here we developed models of non-individualistic behavior. Particularly, we addressed issues such as peer competition and preferences for reputation. In both cases, the main ingredient is the observation that people in general care about their peers. Usually, the closer a peer is, the more influence he would have. So, we took coworkers in a company as the relevant peer group.
The models of social behavior developed were able to shed some light into the stylized facts collected in the introduction. We were able to show that, in a social environment, behavior would considerably depart from what might have been optimal in an individualistic framework. Moreover, the departures were in the desired directions. By allowing for multiple assets we were also able to generate a set of model specific predictions. More specifically, the loyalty model predicts that employees will try to compensate their extra risk by buying more of stocks that are negatively correlated with their company stock and less of positively correlated ones. In the peer group effects model no such conclusion applies. Employees in this case hold the same amount of other stocks that they would in the absence of any peer group effects. These conclusions allow for tests of the empirical strength of each model.

Marcelo Pinheiro, Loyalty, peer group effects, and 401(k), The Quarterly Review of Economics and FinanceVolume 48, Issue 1, , February 2008, Pages 94-122.()
Abstract: In this paper we analyze investors' behavior in the presence of social interactions. Two main models are explored: one based on peer group effects and the other on loyalty. The motivation for this study is the observed behavior of employees when choosing asset allocation for their 401(k) funds. Employees are typically overinvested in own company stock, while theory predicts a strong diversification motive. We argue that the proposed models justify this behavior, even in an environment with full rationality. We also contrast the implications of each model concerning other risky assets.Keywords: Social interactions; Loyalty; Peer group effects; 401(k); Short-selling

All this to be taken with a grain of piquant salt!!!