U.S. stock market participation has increased remarkably over the second half of the 20th century. Starting from a low of 6% in 1952, stock market participation accelerated throughout the i980s and 1990s and reached 32% in 1989 and 49% in 1998.
Yet today, still half of the households do not own any stocks either direetly or indirectly. These observations raise two questions. First,, why is stock market participation so low? Second, why has it increased over time? These questions are important not only for understanding financial markets, but also for designing fiscal policies and social security systems.' This paper sheds light on these issues.
A simple explanation tor these observations relies on the existence of fixed stock market participation costs that have declined over time. Theoretical models commonly appeal to such costs to limit participation (e.g., Merton (1987). Hirshleiter (1988), Allen and Gale (1994). und Abel (2001)). In practice, participation costs encompass a wide range of costs, including trading costs, management fees, and lime and money spent keeping up with market developments. Empirically, wealth is by far the most signiticant determinant of household participation, lending support to a fixed cost explanation.
Furthermore, empirical studies show that relatively low participation costs can keep a large fraction of the population away from stocks. For example, Vissing-Jprgensen (2002) estimates that annual fixed costs of U.S.$260 (U.S.$50) can explain why three-quarters (one-half) of non stock holders do not participate. Whether unobservable information costs account tor a large part of these costs is difficult to assess. Nevertheless, there is evidence that they are a strong impediment to owning stocks. Using data from the 1978 Survey of Consumer Financial Decisions, King and Leape (1987) report that more than one-third of households that do nol own stocks or mutual funds say it is because they do not know enough about them.
Furthermore, participation costs have fallen in the last few decades, thereby attracting more households into the market. This is obvious with respect to trading costs and management fees. In the U.S., the cost of trading individual stocks (including commissions and bid-ask spreads) has dropped dramatically since commissions were deregulated in 1975 (Jones (2002)J. Mutual fund fees (including fund expenses, loads, and distribution costs) have also decreased steadily from 2.26% in 1980 to 1.35% in 1998 (Rea, Reid, and Lee (1999)). The picture is less clear for information costs. On one hand, the media have increased their coverage of the stock market since the 1970s and. more recently, the Internet has given investors instant access to information on compatiies. On the other hand, finding relevant, quality information in this ocean of facts and commentary is a daunting task.
In this paper. I offer a better understanding of how these costs affect stock ownership decisions. For this purpose, I split participation costs into two components. The information cost includes all activities undertaken by investors to improve their assessment of a stock's (or a fund's) performance. For example, investors may read the press, listen to radio and TV reports, search the Web, participate in seminars, subscribe to newsletters, join investment clubs, analyze company accounts, or hire financial advisors. The other component, the entry cosi. contains all other costs such as trading costs and management fees, time spent filing tax forms, opening brokerage accounts, understanding the principles of stocks, or simply becoming aware of their existence. Though they cannot by themselves limit participation, variable costs are bundled with the entry cost because they can act as powerful deterrents when combined with a small fixed cost.
Download
PDF Ebook Information vs. Entry Costs: What Explains U.S. Stock Market Evolution?
