There is a long-standing debate across law, economics and finance regarding the justifications for a mandatory disclosure regime of the type exemplified by US securities law, and a related literature on the empirical question of whether mandatory securities regulation increases the value of firms (i.e. whether the benefits of regulation exceed the compliance costs). In our working paper The Costs and Benefits of Mandatory Securities Regulation: Evidence from Market Reactions to the JOBS Act of 2012 recently made publicly available on SSRN, we use a recent securities law reform to shed new light on this old question.
The Jumpstart Our Business Startups (“JOBS”) Act was passed by Congress in March 2012, and signed by the President on April 5, 2012. It relaxed disclosure and compliance obligations for a new category of firms defined by the Act, known as “emerging growth companies” (EGCs), that satisfied certain criteria (including, most prominently, generating less than $1 billion of revenue in its most recently completed fiscal year). The Act relaxed existing requirements for EGCs conducting initial public offerings (IPOs) on US equity markets, and also relaxed EGCs’ post-IPO disclosure and compliance obligations for a 5-year period. Perhaps most importantly, EGCs were permitted an exemption from auditor attestation of internal controls under Section 404(b) of the Sarbanes-Oxley Act of 2002, as well as exemption from certain future changes to accounting rules.
While the JOBS Act’s provisions were primarily prospective, the Act’s definition of an EGC involved a limited degree of retroactivity, by including firms that conducted IPOs after December 8, 2011 but prior to the enactment of the Act. Moreover, it was known from at least the beginning of March 2012 that the legislation (if passed) would include a retroactive December 8 cutoff. Thus, there is a group of firms that conducted IPOs after December 8, 2011 for which we can observe price data during the sequence of legislative events that occurred in March 2012, increasing the probability of the JOBS bill’s enactment. We use an event study approach to measure abnormal returns for these affected (“treatment”) firms around these legislative events. This provides a test of investors’ expectations about whether or not the value of the mandatory disclosure and compliance obligations that the JOBS bill relaxed exceed the associated compliance costs. However, as these EGCs are all newly traded on public markets and within a few months (at most) of their IPO, identifying a control group is a challenge. For the primary control group, we use firms that conducted IPOs from July 2011 to December 8, 2011 and that otherwise satisfied the EGC criteria.
Our empirical tests compare abnormal returns for the treatment firms with abnormal returns for the control firms. We collect data on IPOs conducted on the US market over the period from July 2011 to April 5, 2012, and merge this data with other relevant information from CRSP, Compustat, the SEC’s EDGAR database, and other sources. Our control group consists of 33 firms (with less than $1 billion in revenues that conducted IPOs prior to December 8, 2011). The treatment group of EGCs varies in size from 25 to 41, depending on the date; we have 27 treatment firms for our most important tests. We use a market model to compute cumulative abnormal returns (CAR) for the firms in our sample over an event window that spans the period from March 1, 2012 (when the retroactivity provision was introduced) to the Presidential signature on April 5, 2012, and for various shorter windows. We focus in particular on an event in the Senate on March 15, when the Senate Majority Leader signaled the importance of the bill by scheduling a vote and describing it as “a measure the Senate should consider expeditiously and pass in short order” suggesting that the possibility of a filibuster led by Democrats was unlikely. We then use a regression framework to test whether the CARs for the treatment firms are significantly different from those for the control group of firms (controlling for various firm-level variables).
Our central result is that the March 15 Senate event was associated with positive and statistically significant abnormal returns for treatment firms (i.e. EGCs), relative to control firms. This finding is robust to controlling for a variety of firm characteristics. The baseline results imply a positive abnormal return of between 3% and 4%. The implied increase in firm value is over $20 million for an EGC with the median market value in our sample. This is comparable in magnitude to, albeit larger than, estimates in the literature of the compliance costs associated with Section 404 of the Sarbanes-Oxley Act (under the JOBS Act, EGCs are eligible for a 5-year deferral of compliance with elements of Section 404(b) internal control disclosure, thus potentially saving a substantial fraction of these costs).
It should be emphasized that these findings are preliminary, and that we are engaged in ongoing research on this project in a number of directions. In particular, we are engaged in replicating the analysis using a number of alternative approaches to computing abnormal returns (such as the Fama-French model). We are also analyzing whether the results are affected by the industrial composition of the treatment and control groups of firms. Finally, we also aim to understand more thoroughly the political factors underlying the legislative process through which the JOBS Act was enacted, and how these considerations affect the interpretation of our results.
Notwithstanding these caveats, however, it is important to note that the JOBS Act’s partial retroactivity provides an unusual quasi-experimental setting in which to measure market expectations of the consequences of relaxing regulatory obligations for a subset of firms. This episode also appears to be unique, in relation to the prior literature, in that it involves a relaxation rather than a strengthening of regulatory obligations. Our results also shed light on the costs and benefits of mandatory securities regulation more generally.
The full paper is available for download here.