Learning to Love Insider Trading? Let's Not

In today's WSJ, George Mason economics professor Donald Boudreaux proposed that we ought to be "learning to love insider trading":

Prohibitions on insider trading prevent the market from adjusting as quickly as possible to changes in the demand for, and supply of, corporate assets. The result is prices that lie.

And when prices lie, market participants are misled into behaving in ways that harm not only themselves but also the economy writ large. ...

In short, overall economic efficiency is reduced.

It's in the public interest, therefore, that prices adjust as quickly and as completely as possible to underlying economic realities—that prices adjust to convey to market participants as clearly as possible the true state of those realities.

As argued forcefully by Henry Manne in his 1966 book "Insider Trading and the Stock Market," prohibitions on insider trading prevent asset prices from adjusting in this way. Mr. Manne, dean emeritus at George Mason University School of Law, pointed out that when insiders trade on their nonpublic, nonproprietary information, they cause asset prices to reflect that information sooner than otherwise and therefore prompt other market participants to make better decisions.

Nope. Insider trading simply does not have the effects Boudreaux ascribes to it. In my book, Securities Law: Insider Trading (Turning Point Series), I explain that:

Basic economic theory tells us that the value of a share of stock is simply the present discounted value of the stream of dividends that will be paid on the stock in the future. Because the future is uncertain, however, the amount of future dividends, if any, cannot be known. In an efficient capital market, a security’s current price thus is simply the consensus guess of investors as to the issuing corporation’s future prospects. The “correct” price of a security is that which would be set by the market if all information relating to the security had been publicly disclosed. Because the market cannot value nonpublic information and because corporations (or outsiders) frequently possess material information that has not been made public, however, market prices often deviate from the “correct” price. Indeed, if it were not for this sort of mispricing, insider trading would not be profitable.

No one seriously disputes that both firms and society benefit from accurate pricing of securities. Accurate pricing benefits society by improving the economy’s allocation of capital investment and by decreasing the volatility of security prices. This dampening of price fluctuations decreases the likelihood of individual windfall gains and increases the attractiveness of investing in securities for risk-averse investors. The individual corporation also benefits from accurate pricing of its securities through reduced investor uncertainty and improved monitoring of management’s effectiveness.

Although U.S. securities laws purportedly encourage accurate pricing by requiring disclosure of corporate information, they do not require the disclosure of all material information. Where disclosure would interfere with legitimate business transactions, disclosure by the corporation is usually not required unless the firm is dealing in its own securities at the time.

When a firm withholds material information, its securities are no longer accurately priced by the market. In Texas Gulf Sulphur, when the ore deposit was discovered, TGS common stock sold for approximately eighteen dollars per share. By the time the discovery was disclosed, four months later, the price had risen to over thirty-one dollars per share. One month after disclosure, the stock was selling for approximately fifty-eight dollars per share. The difficulty, of course, is that TGS had gone to considerable expense to identify potential areas for mineral exploration and to conduct the initial search. Suppose TGS was required to disclose the ore strike as soon as the initial assay results came back. What would have happened? Landowners would have demanded a higher price for the mineral rights. Worse yet, competitors could have come into the area and bid against TGS for the mineral rights. In economic terms, these competitors would “free ride” on TGS’s efforts. TGS will not earn a profit on the ore deposit until it has extracted enough ore to pay for its exploration costs. Because competitors will not have to incur any of the search costs TGS had incurred to find the ore deposit, they will have a higher profit margin on any ore extracted. In turn, that will allow them to outbid TGS for the mineral rights. A securities law rule requiring immediate disclosure of the ore deposit (or any similar proprietary information) would discourage innovation and discovery by permitting this sort of free riding behavior—rational firms would not try to develop new mines if they knew competitors will be able to free ride on their efforts. In order to encourage innovation, the securities laws therefore generally permit corporations to delay disclosure of this sort of information for some period of time. As we have seen, however, the trade-off mandated by this policy is one of less accurate securities prices.

Manne essentially argued that insider trading is an effective compromise between the need for preserving incentives to produce information and the need for maintaining accurate securities prices. Manne offered the following example of this alleged effect: A firm’s stock currently sells at fifty dollars per share. The firm has discovered new information that, if publicly disclosed, would cause the stock to sell at sixty dollars. If insiders trade on this information, the price of the stock will gradually rise toward the correct price. Absent insider trading or leaks, the stock’s price will remain at fifty dollars until the information is publicly disclosed and then rapidly rise to the correct price of sixty dollars. Thus, insider trading acts as a replacement for public disclosure of the information, preserving market gains of correct pricing while permitting the corporation to retain the benefits of nondisclosure.

Despite the anecdotal support for Manne’s position provided by Texas Gulf Sulphur and similar cases, empirical evidence on point remains scanty. Early market studies indicated insider trading had an insignificant effect on price in most cases. Subsequent studies suggested the market reacts fairly quickly when insiders buy securities, but the initial price effect is small when insiders sell. These studies are problematic, however, because they relied principally (or solely) on the transactions reports corporate officers, directors, and 10% shareholders are required to file under Section 16(a). Because insiders are unlikely to report transactions that violate Rule 10b-5, and because much illegal insider trading activity is known to involve persons not subject to the §16(a) reporting requirement, conclusions drawn from such studies may not tell us very much about the price and volume effects of illegal insider trading. Accordingly, it is significant that a more recent and widely-cited study of insider trading cases brought by the SEC during the 1980s found that the defendants’ insider trading led to quick price changes. That result supports Manne’s empirical claim, subject to the caveat that reliance on data obtained from SEC prosecutions arguably may not be conclusive as to the price effects of undetected insider trading due to selection bias, although the study in question admittedly made strenuous efforts to avoid any such bias.

Evaluating the efficient pricing thesis requires a brief digression into efficient capital market theory. Along with the portfolio theory and the theory of the firm, the efficient capital markets hypothesis has been one of the three economic theories most influential on corporate and securities law. In brief, the hypothesis asserts that in an efficient market current prices always and fully reflect all relevant information about the commodities being traded. In other words, in an efficient market, commodities are never overpriced or under-priced: the current price is an accurate reflection of the market’s consensus as to the commodity’s value. Of course, there is no real world condition like this, but the securities markets are widely believed to be close to this ideal.

The so-called semi-strong form of the hypothesis posits that current prices incorporate all publicly available information. The semi-strong form predicts that prices will change only if the information is new. If the information had been previously leaked, or anticipated, the price will not change. If correct, investors cannot expect to profit from studying available information because the market will have already incorporated the information accurately into the price.

The strong form of the hypothesis holds that prices incorporate all information, whether publicly available or not. The strong form makes no intuitive sense: how can the market, which after all is not some omnipotent supernatural being but simply the aggregate of all investors, value information it does not know. If the strong form were true, moreover, insider trading could not be a profitable trading strategy.

Empirical tests of the hypothesis have generally tended to confirm the semi-strong form, while disproving the strong form. To be sure, the validity of the efficient capital markets hypothesis is still hotly debated in academic circles. It is probably fair to say, however, that most scholars regard it as the best available description of how markets behave.

In an efficient market, how does insider trading affect stock prices? Although Manne’s assertion that insider trading moves stock prices in the “correct” direction—i.e., the direction the stock price would move if the information were announced--seems intuitively plausible, the anonymity of impersonal market transactions makes it far from obvious that insider trading will have any effect on prices. Accordingly, we need to look more closely at the way in which insider trading might work its magic on stock prices.

If you studied price theory in economics, your initial intuition may be that insider trading affects stock prices by changing the demand for the issuing corporation’s stock. Economics tells us that the price of a commodity is set by supply and demand forces. The equilibrium or market clearing price is that at which consumers are willing to buy all of the commodity offered by suppliers. If the supply remains constant, but demand goes up, the equilibrium price rises and vice-versa.

Suppose an insider buys stock on good news. The supply of stock remains constant (assuming the company is not in the midst of a stock offering or repurchase), but demand has increased, so a higher equilibrium price should result. All of which seems perfectly plausible, but for the inconvenient fact that a given security represents only a particular combination of expected return and systematic risk, for which there is a vast number of substitutes. The correct measure for the supply of securities thus is not simply the total of the firm’s outstanding securities, but the vastly larger number of securities with a similar combination of risk and return. Accordingly, the supply/demand effect of a relatively small number of insider trades should not have a significant price effect. Over the portion of the curve observed by individual traders, the demand curve should be flat rather than downward sloping.

Instead, if insider trading is to affect the price of securities it is through the derivatively informed trading mechanism of market efficiency. Derivatively informed trading affects market prices through a two-step mechanism. First, those individuals possessing material nonpublic information begin trading. Their trading has only a small effect on price. Some uninformed traders become aware of the insider trading through leakage or tipping of information or through observation of insider trades. Other traders gain insight by following the price fluctuations of the securities. Finally, the market reacts to the insiders’ trades and gradually moves toward the correct price. The problem is that while derivatively informed trading can affect price, it functions slowly and sporadically. Given the inefficiency of derivatively informed trading, the market efficiency justification for insider trading loses much of its force.

Comments

[Anon]

An interesting opposing view. I think the core difference of opinion b/w Profs. Boudreau and Bainbridge come down to whether individual stocks are close substitutes -- i.e., Prof. Bainbridge suggests that I might be just as happy with stock in ACME, Inc. as Zenith, Inc.

Maybe I'm an odd duck of an investor, but my preference in stock is very idiosyncratic. I don't buy a share unless I understand the business very well and feel comfortable with that specific company's business model and ability to execute. For example, I don't see DO and RIG (two oil drillers that are often discussed as competitors), as being close substitutes.

I suppose Prof. Bainbridge would say that my voice is drowned out by the millions of day traders, but I don't think so. They're a constant baseline of noise --albeit liquidity-providing noise -- spread across the spectrum of stocks. Just as many sales as buys (tautology intended). The marginal difference in demand over a long enough run (i.e., weeks, not seconds or minutes) are professional investors and individuals like me. Or so I think.

[save_the_rustbelt]

I believe Don B. argues that cheating individual investors is the price to be paid for making capital allocation more efficient and effective. Wow.

Given that insider trading has been illegal for quite some time, the theories are being presented as facts, when in facts the theories are just theories.

(If we legalize insider trading, should all stocks trades require a consumer warning? Instantaneous disclosure by the insider or company? If numerous insiders conspire to manipulate trading patterns for self-enrichment, should that be a crime?)

Have manipulations by hedge funds become a bigger problem, making insider trading a null issue?

[Bill Stepp]

Given the inefficiency of derivatively informed trading, the market efficiency justification for insider trading loses much of its force.

Even if this is true, it has some effect. There is also the natural rights argument for insider trading, and the Spoonerian fact that the State is a band of robbers. (He called it a secret band of robbers.)
Also, your very first statement is problematic:
a stock is valued on its discounted expected cash flow, not on its expected stream of dividends. Dividends, after all, are paid out of cash flow, and usually amount to a small portion of it. And how could a compnay that
never paid a dividend be valued (e.g. Berkshire Hathaway), or that only started paying one after being in business for years (e.g. Microsoft)?

[Henry G. Manne]

Steve,

Your argument with Boudreaux seems wrong to me on at least three scores, though in all fairness I should acknowledge that most of the errors you make are quite common in these discussions, and economists are almost as guilty as law professors in getting this wrong.

First, your major argument for insider trading is, I think, grossly exaggerated in its significance. This is the argument that IT allows companies to keep justifiably proprietary information secret while letting the market price adjust values to the new but undisclosed information. This is so, but the circumstances in which this argument comes into play, as compared to garden variety insider trading cases, is tiny indeed. While this argument has some throw-away merit, it can hardly serve as the central defense of insider trading. That clearly must relate to the pricing efficiency notion.

However, before I get to that I want to make another criticism that I think is crucial to this entire debate and which you seem to ignore. Empirical data in this field is almost worthless! I know that is a surprising statement to apply to an area where we have perhaps the best data known to empirical economic science. But the data in all the studies of, for example, how quickly insiders' actions are reflected in stock prices,is woefully inadequate to the task. That is so for one simple and uncontracicted reason: all data was taken from the market after the regulation of IT by the SEC was well underway. Thus, mechanisms that undoubtedly existed (Cf. the "clearinghouse function of brokers" described in my 1966 book)in an unfettered IT regime had ostensibly disappeared by the 1970s, and the dissemination of information had become far, far less efficient than it was prior to 1961 (Cady Roberts). This would undoubtedly have had considerable influence on the results of empirical tests for rapidity of dissemination. Yet most if not all the empirical studies to which you refer involve this time period and thus tell us next to nothing about the speed with which the system worked pre-1961. It is also not surprising that these studies reach a variety of conflicting results. That is the nature of the use of inappropriate empirical data.

Finally to get to your most meaty (and erroneous) argument I turn to your position that insider trading does not really help guarantee an efficient market. This you say is so for two reasons, the first being that the elasticity of demand for any one company's stock is near zero since it is always substitutable (in diversified portfolios)for the very large number of shares that have the same Beta. Thus an insider's transactions are generally too small to affect the price.

The second argument, which follows from the first, is that derivatlively informed trading is the kind that must necessarily account for price changes when there is IT. And since derivative trading is necessarily less "perfect" than trading with real direct information, the price will generally not be the "correct" one, but only some approximation of it.

The problem with both these propositions (and here I certainly do not want to fault you for adopting what seems to be the conventional view in the economics literature and for not noticing how inappropriate it is for present purposes) is that it all assumes a mechanism for price formation that simply does not work most of the time. Here I have reference (and this is confirmed by your misleading reference to simplified traditional supply and demand analysis) to the implicit idea that a stock's price is formed by the marginal trader with information who arbitrages the market until the price is right.

That is standard price theory, and it even holds true in the case of a purchase of substantially all the stock of a given company (say in a takeover). But it is a far cry from reality, as I think some of the better behavioral economics now teaches us. There is clearly a different mechanism at work here, something that borrows on "the wisdom of crowds" notion of James Suroweicki, a kind of weighted averaging (nothing like the "consensus" you referred to) that results in a (miraculously?) efficient pricing scheme. Clearly there are these two mechanisms for formation of prices either one of which may doninate at any given moment. However, the latter is more likely to be operating in run-of-the-mill stock trading. Note further that the second has little or nothing to do with suppply and demand concepts and yet there is nothing in it that contradicts the findings of efficient market researchers. Insider trading long did and undoubtedly still does play a large role in keeping the stock market as efficient as it is and useful in all the ways that Don Boudreaux described.

[MDF]

I recognize that data on insider trading in the US is problematic, for the reasons you note above. However, couldn't empirical studies on this topic draw from data outside the US? In the UK, for example, the FSA has noted that fully a quarter of all mergers involve significant insider trading (notwithstanding laws against the practice). For that matter, even in the US, there is frequently substantial market movement leading up to the announcement of previously "material nonpublic information" -- certainly some of this data could be used, properly structured to take into account the activity's necessarily surreptitious nature. Or just an analysis of the German market in the early 1980s or the Taiwan or Shanghai markets today (where insider trading convictions are so infrequent as to be effectively nonexistent.)

[Jason Reaves]

I can see how the legalization of insider trading might (or might not) increase market efficiency, but the agency costs would be astronomical. Legalization would further erode the tenuous connection between executive pay and company performance by allowing feckless corporate captains to abandon ship before the passengers have even had the chance to line up for life vests. Some degree of insider trading is already commonplace, but the practice should remain illegal as the public needs the occasional perp walk (as with Rajaratnam) to maintain the pretense basic fairness.

[Henry G. Manne]

Jason Reaves, where have you been? Prior to the mid-1970s, there was no significant enforcement of insider trading laws anywhere in the world. Were agency costs astronomical then? I think not. Was there more of a disconnect between compensation and performance then? I think not. Did the public need pep talks to encourage false confidence, when everyone already knew that insider trading was a fact of financial life? No. Were markets less efficient before this silly non-enforceable law was pushed? I seriously doubt it. Did capital markets fail from investors' lack of confidence in the integrity of the market, as SEC blather would imply? Not that I know of. Just a little bit of historical knowledge can go a long way to clearing up the fog surrounding the insider trading debate.