Royal Dutch Shell Evaluation of Oil Reserves — страница 15

factor of 2 in order to x E ) =( M 1( x ) − M 2( x )) ⋅ e 5 . 1 σ 2 (7) calculate the actual restatement in expected value of oil and gas reserves (the calculations are represented in Exhibit 3.2) 3.2 Estimation of Market Capitalization Discount for Parental Companies Another aspect one should cover in this event analysis is the discount is the market capitalization. As it was mentioned above, stocks of the parental companies reacted quite sharply on the announcements and therefore allowed to assess the investors’ reaction on the oil reserves restatement. In order to do so the event study methodology was used and the cumulative abnormal stock returns (CAR) were calculated. The returns of RD and Shell stocks were compared to the returns of S&P 500 stock

index. Although RD and Shell stocks are traded not only on NYSE but also on LSE (denominated in British Pounds) and in Amsterdam (denominated in Euros), the oil is priced in US$ and it makes sense to use a comprehensive index for the dollar denominated stocks in order to filter currency effect. It is important to remember that RDS’ assets are strongly dependent on the movement in the oil prices and the upswing in oil prices (or perhaps an expected upswing in oil prices) will most probably lead to stock appreciation. Usually this will not be the case for S&P 500 index, which includes large portion of stocks of oil consuming (industry) companies that are expected to fall on the oil price upturn. In order to control for this effect AMEX Oil & Gas was used. AMEX index

consists of major oil and gas producing companies stocks, and therefore would react on the moves in prices of oil it the same manner and to the same extend as the stocks of RD and Shell do. The problem with AMEX index of course, is that the weight of RD and Shell stocks in it is much larger than in S&P 500, in addition other stocks may have been traded down on restatement announcement, as the markets suspected that similar problems with oil reserves may well have existed in other oil companies. All this would result in smaller negative CAR of AMEX index, than of S&P 500. To improve the results of event studies the CAR on both indexes were calculated and then compared. The control window for both indexes was 240 days (from the beginning of 2003) and several event windows

were calculated. First is the event window of (-5; +20). The window is non-symmetric, since there was virtually no information about the restatement before the announcement and therefore adding more days before the announcement to the event has a very low added value. First event window ends on February 6, 2004 and covers only first announcement. In order to cover the second announcement another event window (-5; +53) was introduced. This Royal Dutch Shell: Evaluation of Oil Reserve window ends on March 25 and therefore covers first and second announcements and allows 5 additional days after the second announcement, which took place on March 18. This event window is rather long, so the results it would produce are less statistically reliable. To capture all three announcements

another event window of (-5; +70) was introduced. This window stretches up until April 19, 2004, but the results it produces are hardly reliable both from statistical and economical point of view. The calculations and results of event study summarized in Table 3.2 are shown in Exhibit 3.3 and 3.4 and the full CAR charts are shown in Exhibit 3.5 Table 3.2 Window Shell Royal Dutch S&P AMEX S&P AMEX (-5; +20) 16.4% 12.7% 14.8% 11.2% (-5; +53) 15.4% 13.4% 16.3% 14.3% (-5; +70)** 5.7% 8% 10.5% 12.8% ** - statistically insignificant at 10% level First thing that one would notice by looking at table 3.2 is that the abnormal returns of two parental companies are different quite significantly. In fact, as one calculates the correlation of adjusted dollar denominated returns of two

companies stocks the correlation will not be 100%, but rather about 0.92 (see Exhibit 3.6). Theoretically, the only property of both companies is the stake in RDS Group and as the returns are measured in the same currency, they should have the correlation of 1. The reason behind this difference can be a slightly different dividend policy. For example, in 2003 the dividends paid to Shell shareholders were about $2610 million and dividends to RD shareholders $4292 million (RDS Group: Financial and Operational Information, p 6sq, 2003). The ratio is different from 40:60 holding ratio of the companies. Another reason might be different holding structure of two companies and therefore different liquidity. Overall, the unification of Royal Dutch and Shell mentioned in Chapter 1 is