Anomalous Price Behavior Around Repurchase Tender Offers days before the repurchase announcement. On average, 16.41 percent of the shares outstanding were purchased, but a much larger fraction(87 percent)of the shares tendered was purchased by the company. apparently, many investors decided not to tender their shares. By tendering shares, investors realize capital gains which are substantial if the base price is low relative to the tender offer price. Therefore tax considerations might explain why many investors do not tender. However, tax considerations are unlikely to provide a full explanation given the widespread ownership of shares by tax-exempt institutions. We also computed the cumulative average abnormal return to the nontendering shareholders from five days before the announcement until ten days after the expiration date(CAR)and the weighted average abnormal return to the tendering and nontendering shareholders(TOTALR). On average, nontendering share- holders earn an abnormal return of 12. 54 percent, which is significantly smaller than the 21. 79 percent premium that the tendering shareholders receive. Because on average, 16.41 percent of the outstanding shares are repurchased, the repur chase announcement increases the total value of the firm by 14.29 percent (TOTALR) To test for the stability of the results, we divided the sample into two subperiods: 1)1962-1979, which approximately corresponds to the period exam ined in past research, and 2) the more recent period, 1980-1986. Table I shows that, in the second subperiod, returns to both the tendering and nontendering shareholders became smaller. We elaborate on these results later Il Trading Rules Around the Expiration Date A Methodology In a repurchase tender offer, firms offer to buy back a fraction of their shares at a tender price, Pt, before a specific date(the expiration date). The first trading rule consists of buying shares before the original expiration date and tendering to the firm. In order to compute the gains from such a strategy, it is important to keep in mind the rules governing the repurchase tender mechanism If the offer is undersubscribed (i. e, the fraction of shares tendered, FT, is less than the fraction of shares sought), the firm will repurchase all shares tendered. This is true even when the offer is extended and becomes oversubscribed later on. If the offer is oversubscribed, the firm will either buy back all the shares endered or allocate pro-rata so that each shareholder sells the fraction Fn/Frof his or her shares to the company(where Fp is the fraction purchased Thus, the profit from buying a stock at a price PB during the tender period and tendering it to the firm is equal to Pr- Pb if the offer is undersubscribed at The benchmark return used was based on the CrsP equally weighted market index. Because of the short period, the results are not sensitive to the various methods of computing abnormal returns An offer may be extended. In our sample, 99 of the 258 offers were undersubscribed at the initial expiration day: 43 were extended and 1l of these became subsequently oversubscribe g To reduce shareholder servicing costs, companies in many cases buy all the shares from the hareholders who tendered a small number of shares, typically 100 or less. However, this has a very small impact on the F/Fr ratio
The Journal of finance the first expiration date. For oversubscribed offers(at the first expiration date) the profit from the trading strategy is Py(F/FT)+ PE(1-Fp/FT)-PB where Pe is the price after the expiration day In order to allow for transaction costs, no trading was undertaken in cases where Pg>0.97 Pr. An investor who buys shares and tenders does not incur transaction costs on the repurchase shares. Therefore, the transaction costs of this strategy are less than the costs of a round-trip transaction To test the sensitivity of the results to the timing of purchases and sales different trading rules were examined. The first trading rule assumes buying and ay and, if the offer is ascribed, selling the nonrepurchased shares two days after the expiration day. In the second trading rule, the purchase is made on the day before the expiration and the nonrepur chased shares are sold two days after the expiration. Finally, the third trading rule assumes that we buy on the day before the expiration day and sell four days after the expiration. In each case, results are provided for the total period and the two subperiods Note that the trading rule assumes that we can actually sell the shares two days after the expiration date. In general, a day after the offer expires, the company announces the outcome of the offer. If the offer is oversubscribed and the company decides not to buy all the shares, a preliminary pro-rata decision is announced. The preliminary pro-rata number is always very close to the final pro-rata allocation, which is announced within ten business days. However, an investor does not have to wait until the final pro-rata decision is made before lling the remaining shares Under the usual settlement procedure, an investor who buys shares will officially become the owner of the stock after five business days. Hence, buyin and tendering one day before the expiration day is not feasible. However, by following a cash-settlement procedure, it is possible to obtain title to the shares on the same day. Discussions with brokers revealed that cash settlement implies larger transaction costs for relatively small trades (i.e, a small investor cannot use discount brokers). Thus, we also tested the trading strategy by assuming that the stock was purchased six business days before the expiration day and the nonrepurchased shares are sold twelve business days after the expiration day The long post-expiration period reflects the fact that the exact pro-rata decision is known within twelve business days. B. Results The results of the trading strategy are shown in Table Il. The abnormal returns presented are adjusted for the market movement based on the CRSP equally 10 Shares are, in general, kept in the depository under the age name and transferred to the the receipt of payment for the acquin tendered, the brokerage house makes a commitment to ha ndered shares available, Because the preliminary pro-rata decision is very close to the final allocation, the nonrepurchased shares are released by the brokerage house after the preliminary announcement
Anomalous Price Behavior Around Repurchase Tender offers 461 Table ll Abnormal returns from Buying and Tendering Abnormal returns(in percent)from trading rule around expiration. according to the ading rule, shares are bought T days prior to the expiration date wl market price is at least three percent below the tender offer price. In oversubscribed offers, shares not repurchased by the company are sold T2 days after the expiration Mean t-value Positive Panel A: Total sample(1962-1986)(109 observations) Trading rule #1 5.88 94.5 0,T2=2) Trading rule #2 9.46 7.76 552 972 T2=2) Trading rule #3 9.26 758 5.59 27 1,T2=4) Panel B: First subperiod (1962-1979)(63 observations) Trading rule #1 Trading rule 8.01 758 (T1=-1,T2=2) Trading rule #3 726 5.59 (T1=-1,T2=4) Panel C: Second subperiod(1980-1986)(46 observations Trading rule 100.0 Trading rule #2 100.0 Trading rule #3 11.39 4.44 95.5 (T1=-1,T2=4) weighted index. Because the holding period is short, the results are not sensitive to the assumption made about the process that generates returns.( See Brown and Warner(1985). )However, only expected nonrepurchased shares are affected by the market risk. As shown in Table I, the fraction of shares purchased relative to the shares tendered exceeds 80 percent, and therefore this strategy is less risky than the benchmark portfolio Panel A of the table (which covers the total sample period)shows that, in 109 of the 258 cases, the market price is at least three percent below the tender offer price before the expiration date. The trading rules generate a rather impressive abnormal return of more than nine percent on average(not annualized), with t statistics larger than 7.5. The median value is smaller(5.5 percent), and at least 93 percent of the abnormal returns are positive. There seems to be no substantial difference among the various trading rules. Panels B and C show similar results for the two subperiods. The difference in average abnormal returns over the two subperiods is not statistically significant, although the abnormal returns are higher in the second subperiod. The median abnormal returns are essentially the same across the two subperiods, around 5.5 percent. The persistence of abnormal returns is striking: for example, in the