Tbe Revtew of Financial Studies /v 6n 4 1993 Table 1 Returns from four-month event window for both trader-quote data and s&P bond guide data Trader quote S&P bond guide DTRET DTRETR DTRET DTRETR A:With RJR Nabisco bonds One bond/firm,N=13 Mean -2.75% -6.55% 2.03% -1.70% (-1.06) (-2.78) (1,74) (-1.17) All Bonds,N=36 Mean -4.60% -7.33% -1.10% -3.83% (-3.65) (-5.62) (-1.14) (-3.25) B:Without RJR Nabisco bonds One bond/firm,N=12 Mean -2.32% -6.18% 2.63% -1.14% (-0.84) (-2.44) (2.45) (-0.78) All bonds,N=22 Mean -2.50% -5.00% 1.55% -0.93% (-1.41) (-2.80) (1.38) (-0.62) f-statistics are in parentheses below the coefficients.The 36 bonds from 13 companies examined here represent all bonds traded at Lehman Brothers over the period January 1985 through April 1989 that have both a consecutive time series of dealer quotes available and prices reported in the S&P Bond Guide around the month of the buyout announcement.Risk-adjusted bond returns (DTRETR)are calculated by subtracting from the raw bond return(DTRET)the return of an index with rating and maturity characteristics similar to the bond of interest.The "adjustment"index was constructed from eight Lehman Brothers Corporate Bond indexes by linear interpolation of the two closest indexes in the dimensions of rating and maturity (again,bonds outside the available range of characteristics are benchmarked against the closest index).Under the null hypothesis that LBOs had no effect,the average risk-adjusted debt return should be zero.The event-window returns are calculated from the month-end two months preceding an LBO announcement to the end of the second month following it.For example,if the LBO announcement is March 20,the returns are calculated over February,March,April,and May.One bond/firm means that values are averaged first within firm and then across firms.Trader-quote data are from Lehman Brothers,and exchange-based data are from the S&P Bond Guide. are at best statistically insignificantly negative,at worst statistically significantly positive.In contrast,in this article the risk-adjusted trader- quoted return drops are considerably larger,ranging from 5.00 to 7.30 percent,and always statistically significant.Even unadjusted returns are always negative,ranging from -2.32 to -4.60 percent,although the unadjusted drops are only statistically significant when we include all 36 bonds as independent observations. Panel B of Table 1 presents the same information as panel A but with all 14 RJR bonds excluded from the analysis.Risk-adjusted returns from trader-quote data again reveal significant losses,but none are provided by the S&P data.It is noteworthy that the only significant figure in the s&P data in panel B is a positive unadjusted debt return on a firm-by-firm basis. Note that not all exchange data respond slowly.We can illustrate this with two Federated Department Store bonds that are exchange 966