In a recent speech the Governor of the Bank of Canada, David Dodge, has discussed how to promote the efficient operation of Canada’s financial market. Instead of approaching the topic in a detailed and technical way he has discussed the relevance of two qualities that are essential to the efficient operation of financial markets: trust and transparency. He has then defined trust as closely related to confidence and credibility, a kind of social capital—a shared asset that benefits everyone. No better speech could have been addressed after the damages done by Enron, the Asian Crises and the financial scandals that we have experienced in Europe and ultimately in Italy. The views of the Governor are well known by researchers in economics. The role of trust in economic exchange has been deeply studied. The most widely used definition of trust that economists like is that it is the subjective probability individuals attribute to the possibility of being cheated: it is partly based on objective characteristics of the financial system (the quality of investor protection, its enforcement, etc.) that determine the likelihood of frauds such as Enron and Parmalat. But trust reflects also the subjective characteristics of the person trusting, like his educational background rooted in past history and it is unlikely to fade away even with experience and data. These differences in trust seem to have economically important effects on trade, portfolio investments, and foreign direct investments. And they have been proved relevant also to explain the workers’ refusal to adhere to the statutory severance indemnity (TFR) reform recently approved in Italy.
Of course when we look at the effect of trust using macrodata we might doubt that the results obtained suffers from a reverse causality problem: We do not really know whether it is trade and capital movements that influence the trustworthiness of people or the other way round, that is trust that affect economic transactions. In a recent paper of mine ( http://papers.ssrn.com/sol3/papers.cfm?abstract_id=997934) the role of trust has been studied even at the micro level. In this paper we focus solely on generalized trust, and think of it as the subjective belief that an agent places on the probability of honest dealing. Given its subjective nature, an appropriate measurement of trust requires surveying of opinions. We then follow the literature and use the Eurobarometer data on bilateral trust among nations. This measure is based on the responses of citizens in one country about the trustworthiness of citizens from all other European countries, including their own. We then study the effect of trust on a particular type of investment: the venture capital (VC) investment. There are two main reasons for focussing on the role of trust on VC activity. First, the financing of a new company inherently involves limited information and high uncertainty. Moreover, there is a lot scope for opportunistic behavior. Under these circumstances it is reasonable to conjecture that subjective beliefs about trustworthiness matter. Second, the venture capital industry is tiny relative to the economy. According to the European Venture Capital Association, total investments in venture capital (excluding buyouts) accounted for less than 0.1% of European GDP in 2004. Venture capital activity is clearly irrelevant to the formation of trust among nations. This means that we have a setting where we need not worry about endogeneity problems: trust among nations can affect individual venture capital investments, but these investments do not have a reverse effect on trust among nations.
The data used consists of European venture capital investments for the period 1998-2001. This hand-collected dataset has several important strengths. It contains investors and companies from all across Europe, generating rich variation in investment patterns. It contains some information not commonly found elsewhere, such as the precise geographic location of every single company and investor, which allows us to calculate the exact distance between every investor-company pair. Most important, it contains some detailed information about contracts, including the structure of cash flow and control rights. Such contractual details cannot be obtained from any publicly or commercially available database.
One obvious challenge is to distinguish trust from other factors that might explain investment behavior. We do that by controlling for investor and even company characteristics, investor availability of information, the sharing of a common language and of the same legal origin. And again for differences in GDP, geographic distance, and for how a company fits with an investor’s industry and stage preferences. The question are essentially two: first of all whether trust affects investment decisions. We consider the sample of all potential deals and find that higher trust significantly increases the probability of making an investment. Then we ask whether contracts compensate for any lack of trust. We use the sample of realized deals to examine three issues. First we ask whether the investor brings other investors into the deal. We find that with more trust, an investor is more likely to invest on his own, and more likely to lead a syndicate rather than being a follower. The investor’s share in the total investment amount provided also increases with trust. This suggests that an investor’s willingness to take responsibility for the deal increases with trust. Second, we inquire about the financial securities used in the transaction: we focus on the extent to which an investor requires securities that provide downside protection. This means that the financial contract protects the investor in case of poor company performance, and can be achieved, for example, with debt or convertible preferred equity. We find greater use of downside protection with lower trust, suggesting that the security structure is used to alleviate investors lack of trust. Third, we examine the allocation of control rights. We find that the use of contingent control rights increases with trust. This result seems surprising at first, since contingent control rights are not used to assuage any lack of trust. Instead, it appears that high trust is required before two parties agree on contracts that involve the use of sophisticated control arrangements, such as making control contingent on performance.
These results should be carefully taken into consideration by policy makers and politicians that often distrust trust-based explanations of the functioning of financial markets. So do people that downplay the negative effect of the financial scandals on investment. And ultimately on economic growth.