Effects of opportunistic behaviors on security markets: an experimental approach to insider trading and earnings management

Abstract

The current paper produces the very first experiment on the relationship between insider trading and earnings management in an unregulated scenario, where the decision of misreporting only depends on ethical considerations. Using a price formation model based upon the excess of supply/demand, but also on the difference between the reported and real earnings of the firm, our results indicate that (1) insiders systematically misreport earnings, (2) earnings management is perceived as a self-dealing strategy, (3) insiders time opportunistically their transactions, thus, achieving higher returns than outsiders, and (4) insider information is not properly conveyed in prices, what erodes market efficiency. Therefore, earnings management and insider trading induce inefficiencies on stock prices and affect the reliability and confidence on financial markets. We thus advocate for the enforcement of financial market regulation since the ethical reasoning of investors is not enough to prevent opportunistic practices.

Appendix: Instructions

This experiment simulates a market, in which stocks of a firm named X are traded. Participants are randomly matched in groups of four that remain unchanged for the entire session. For each group, one participant is labeled as “Type 1” and the other three as “Type 2”. Every participant will be given an initial endowment of 100 ECUs (Experimental Currency Units) and 5 stocks of the firm X. The initial stock price is 20 ECUS, and the initial wealth of 200 ECUs per agent. All this information is displayed in the laboratory screen.

$${\text{Initial wealth }} = { 1}00 \, + \, \left( { 5 \times 20} \right) \, = { 2}00{\text{ ECUs}}$$

The experiment consists of 40 rounds and every round has two steps. In the first step, Type 1 privately receives information about the variation rate of firm X earnings in the current round t, [variable referred to as B_t]. This information will be randomly drawn and may be one of the following five values:

Bt	Firm’s X earnings rate of change
−20	20 % decrease
−10	10 % decrease
0	Unchanged
+10	10 % increase
+20	20 % increase

With this information, Type 1 will send a message to the other three participants reporting a firm earnings figure ($({\text{B}}_{\text{t}}^{*})$) from the table above that do not necessarily has to be the true earnings figure (B_t).

In the second step the four participants of every group will make one of the three possible decisions:

• Buy ONE stock of the firm X

• Sell ONE stock of the firm X

• Neither buy nor sell stocks of the firm X.

You will be also asked about your belief about the movement of the price (whether it is going to increase or decrease).

Once your decisions are made the price of the stock (P_t) will be computed and displayed in the computer interface, and the stocks, income and wealth in your portfolio updated according to the placed orders. In particular, P_t will depend on the price of the previous round (P_t−1), the difference between the number of stocks that participants within the same group are willing to buy (D_t) and sell (S_t) in the current round and the difference between reported ($({\text{B}}_{\text{t}}^{*})$) and real (B_t) earnings, according to the following equation:

$${\text{P}}_{\text{t}} = {\text{ P}}_{{{\text{t}} - 1}} + { 2}\left( {{\text{D}}_{\text{t}} {-}{\text{ S}}_{\text{t}} } \right) \, + \, \left( {{\text{B}}_{\text{t}}^{*} {-}{\text{ B}}_{\text{t}} } \right)/ 10$$

For example, let us assume that earnings experienced a 10 % increase in the first round (B_t = 10) but the Type 1 participant reported a 20 % increase ($({\text{B}}_{\text{t}}^{*})$ = 20). Given this information (note that B_t is only known by the Type 1), if two participants are willing to buy ONE stock each (D_t = 2), one participant is willing to sell ONE stock (S_t = 1) and one participant does not want to either buy or sell, the price at which all the orders in the first round will be 23 ECUs:

$${\text{P}}_{\text{t}} = { 2}0_{{}} + { 2}\left( { 2_{{}} {-}{ 1}} \right) \, + \, \left( { 20 \, {-}{ 1}0} \right)/ 10 = 2 3$$

Note that in every round the stock price may rise a maximum of 12 ECUs with respect to the price in the previous round. This will happen when

• All participants are willing to buy D_t = 4

• Earnings experience a 20 % decrease B_t = −20

• Participant Type 1 reports a 20 % earnings increase $({\text{B}}_{\text{t}}^{*})$ = 20

On the other hand the stock price may fall a maximum of de 12 ECUs with respect to the price in the previous round. This will happen when

• All participants are willing to sell S_t = 4

• Earnings experience a 20 % increase B_t = 20

• Participant Type 1 reports a 20 % earnings decrease $({\text{B}}_{\text{t}}^{*})$ = −20

There are no short sales. The stock price cannot go below 1 ECU and in case this happens all the orders will be closed at this price.

Every 5 rounds you will win an extra dividend of 2 ECUs per stock that you possess once the orders placed on the dividend payout round are closed (the dividend will increase your wealth in the payout period).

1.1 Payoffs

At the end of the experiment participants will receive the quantity of ECUs (wealth) that they had accumulated throughout the experiment. This wealth will be the sum of the available ECUs on your account plus the value in ECUs of the stocks on your portfolio at the end of the experiment. These stocks will be valued at the stock price at the end of the experiment. This wealth will be converted in Euros at the exchange rate 50 ECUs = €1.

1.2 Test rounds

Next, in order to be familiar with the experiment you will be able to practice in your computer for 5 rounds. At the end of every round you will be asked some control questions to ensure that you have understood the instructions. Once you finish the test rounds and after successfully answering the control questions the experiment will start. The results of the 5-round test will not be considered for computing the final payoff of the experiment.