How Efficient is Sufficient? Securities Litigation Post-Halliburton

The following post comes to us from Bradford Cornell at California Institute of Technology.

In its recent decision in Halliburton Co., et al. v Erica P. John Fund, Inc., the U.S. Supreme Court upheld the legal standard for reliance in Rule 10b-5 securities fraud class actions that it had established some 25 years ago in Basic, Inc. v. Levinson. This standard, known as the fraud-on-the market doctrine, created a rebuttable presumption that plaintiffs relied on the integrity of the market price if they can establish that the market for that security was efficient. Defendants can rebut this presumption in several ways, including showing that the market for the security was not efficient or that the security’s price was not affected by the misrepresentations at issue. In delivering its ruling, the Halliburton Court noted that market efficiency is not a binary, yes-or-no proposition but is instead a matter of degree, pointing out that “a public, material misrepresentation might not affect a stock’s price even in a generally efficient market.” (Halliburton, 573 U.S. ___ at 10.)

In How Efficient is Sufficient? Securities Litigation Post-Halliburton, I first discuss Fama’s classic definition of the semi-strong form of market efficiency, which holds that a market is efficient if a security’s price reflects all publicly available information related to the value of that security. I then examine the upper and lower bounds of market efficiency because real-world markets are never either perfectly efficient nor are they markedly inefficient. Markets efficiency arises because sophisticated investors perform research and trade accordingly, thereby incorporating new information into security prices. These investors need to earn a sufficient return to cover the cost of their efforts, but if prices continually reflect all publicly available information, they will not be compensated. Consequently, there must be sufficient inefficiency that sophisticated investors have an incentive to bear research costs. On the opposite end of the spectrum, in a completely inefficient market would be no correlation between price and value. In such a market, a cheap plastic Bic pen would be equally likely to sell for more or less than a gold Montblanc. Of course, no real-world market could stay that way because astute buyers and sellers would quickly enter, take advantage of the massive discrepancies between price and value, and place trades that eliminate those discrepancies.

Next, I discuss how efficiency can be measured. As yet, there is no way to scientifically measure the degree of efficiency between the bounds. However, finance theory implies that the degree of efficiency should be related to the costs of gathering and processing value-related information and the associated benefits. These benefits are presumably related to the size positions an investor can take without unfavorably moving prices. Following this logic, in the Cammer v. Bloom and Krogman v. Sterritt cases, courts suggested looking at several factors in an attempt to measure the degree of efficiency. Most of these factors relate to the costs and benefits associated with gathering, processing, and profiting from information. The only direct measure of efficiency is the speed with which the security reacts to the disclosure of new information, which is one on the few hallmarks of efficiency that can be measured precisely.

The ability to quantify reaction time led to a distinction between informational and fundamental efficiency. A market is informationally efficient if prices respond to public announcements of new information sufficiently quickly that investors cannot make abnormal returns if they try to trade on that information. This type of efficiency relates only to reaction speed rather than whether that reaction is rational or accurate. Hundreds of studies confirm that developed securities markets are highly informationally efficient. By contrast, a fundamentally efficient market is one in which the market price always equals the present value of cash flows discounted at the appropriate cost of capital. Because fundamental efficiency incorporate informational efficiency, a market that is fundamentally efficient is also by definition informationally efficient. Testing for fundamental efficiency is problematic because it requires a model of asset pricing to allow comparison of fundamental model prices with market prices. Unfortunately, there is great dispute as to the appropriate asset pricing model to use. As a result, there is no unambiguous method to test for the degree of fundamental efficiency.

Finally, I explore the implications of the Supreme Court’s recognition that market efficiency is relative, not binary. One implication is that it may be appropriate to use a different measure of efficiency to determine reliance from that used as a basis for estimating damages in a securities litigation context. Most investors have high costs of gathering and processing information; thus, it makes sense in most instances for those investor to presume that market price is the best measure of fair value. For purposes of assessing reliance, this implies that courts would want to adopt a relatively low standard of efficiency.

However, measuring damages by reference to price movements is different. Such measurement requires that the market be sufficiently fundamentally efficient that the stock price accurately reflects both the omitted or misstated information and the effect of its subsequent revelation. This is complicated because there is a selection bias that arises when damages are measured after the alleged misstatements or omissions have been disclosed to the market. Plaintiffs can simply wait to file suits after observing big market drops. But conditional on the fact that a large price decline has occurred, there is a greatly increased likelihood that the drop is due, at least in part, to factors unrelated to fundamental valuation and will thus overstate damages unless the market is close to full fundamental efficiency. To minimize selection bias impact, it makes sense for courts to require a higher standard of efficiency for measuring damages than that employed for assessing reliance.

But this example is not likely to be the only implication of courts recognizing that market efficiency is relative rather than binary. Once it is recognizing that market efficiency is a continuous variable the door is open to using different benchmarks depending on the facts of individual cases.

The full paper is available for download here.

Both comments and trackbacks are currently closed.