back start next


[start] [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99] [100] [101] [102] [103] [104] [105] [106] [107] [108] [109] [110] [111] [112] [113] [114] [ 115 ] [116] [117] [118] [119] [120] [121] [122] [123] [124] [125] [126] [127] [128] [129] [130] [131] [132] [133] [134] [135] [136] [137] [138] [139] [140] [141] [142] [143] [144] [145] [146] [147] [148] [149] [150] [151] [152] [153] [154] [155] [156] [157] [158] [159] [160] [161] [162] [163] [164] [165] [166] [167] [168] [169] [170] [171] [172] [173] [174] [175] [176] [177] [178] [179] [180] [181] [182] [183] [184] [185] [186] [187] [188] [189] [190] [191] [192] [193] [194] [195] [196] [197] [198] [199] [200] [201] [202] [203] [204] [205]


115

It might be that there is a distiQction between the weeks that begin with an upward move on Monday rather than a lower price, but both cases were combined. This assumes that the pattem, rather than the direction, is most important, it would also be reasonable to have assumed that the upward bias of the stock and interest rate markets during the past 10 years would justify a separation of pattems according to their initial direction.

Daily Sequences

Figure 15-6 shows the aggrate results of the weekday pattems for a sampling of 23 futures markets tested during the period 1976 through 1985, appearing in the previous edition of this book. The ramibers next to the corresponding pattems indicate the percentage of occurrences of each new days pattem. For example, if the Monday-Tuesday pattem was XX, indicating that prices moved the same direction, then there was a 58**o chance that an 0 followed, indicating areversal.

Returning to the original premise for studying weekday pattems, we can use the broad set of maricel to see if any dominant behavior exists. We see that there was a 58**o chance of a reversal on Tuesday (XO) anda53**o chance of that reversal continuing into Wednesday (XOO). If areversal did not occur on Tuesday, there was a 58**o chance that it would occur on Wednesday (XOO). For a large ramiber of tests, 58**o shows a definite pattem that confirms the original view. There are very few significant \alues that appear on Thursday, but three pattems of interest appear on Friday-XOXXX, XOOXX, and XXOOX-showing 57<»o, 64** o, and 60** o, respectively. These three pattems are important because they all show midweek reversal and a continuation of the trend by the end of the week.

The weekday pattems for the same six maricets studied in the previous section ("Opening Gaps") were also tested for the period Januarj 1 * 86 through February 19*36 and appear

FIGURE 15-6 23 fiitures madiets, 1976-1985.

in Figure 15-7a-f To imderstand the accuracy of these tests, each one includes approximately 480 complete weeks of information used to compile the pattems; therefore, each of the two combinations appearing on Tuesday have 240 entries, each pattem on Wednesday represents 120 weeks. Thursday 60. and Friday 30. assuming an equal distribution of pattems. All of these maikets are actively traded.

Looking first at the Deutschemarks (Figure 15-7a), we see the same pattem as in the 10 years prior There is a 55**o chance of a Tuesday reversal (XO) and a 58**o chance of a Wednesday reversal (XXO). If a reversal does not occur by Wednesday, there is a 57**o chance of a reversal on Thursday (XXXO), but only a



28** chance that a reversal will occur on Friday (XXXXO), if it hasnt yet happened. The reverse situation shows that, if prices moved in the same direction Monday through Thursday, then there was a 72**o chance they would move in that direction on Friday.

The same pattem seen in Deutschemarks is also apparent in the S&P (Figure 15-7b), where the chance of a reversal is 56**o, 54**o, 62**o, and 38**o. The most dominant Friday pattem is a trend continuation following a similar d on Thursday (XXXXX and XXOXX). Treasury bonds (Figure 15-7c) show a slightly greater consistencj for reversal-57**o, 59**o, 55**o, and 53**o. For bonds, the chance of continuing in the same direction for the entire week is small. Gold (Figure 15-7e) also reflects similar reversal pattems-59**o, 55** o, 58** o, and38**o. In view of the steady downtrend in gold prices during this 10-year period, this pattern appears to show a consistent method in the behavior of the traders. Soybeans (Fig

FIGUREl 5-7 Weekd study results, (a) Deutschemarb. (b) S&P 500. (c) Treasurj bonds, (d) Heating oil. (e) Gold, (f) Soybeans.

Monday Tuesday Wednesday TTursday Fnday

L>ooo «3-Q

XXXX 43 I-

[3-

[ -

FIGURE 15-7 (Continued)



-1»°° "

□ -

-* ]-1

tXXO 52 I-

15-7, maiket clearly out of the financial circle, show their reversal tendency as 56** o, 58** o, 59** o, and 58**o. This variation on Friday does not change the uniform maiket reversal pattern seen early in the week.

Heating oil pattems (Figure 15-7d) are noticeably different fiom the other five markets. There seems to be no dominant pattem on Tuesdays and Wednesdays and only minor separation on Thursday. The three best pattems on Thursday, XOXO, MO, and XXOO, all show that Thursday moves in the direction opposite to Monday. The three best pattems on Friday, XOOXO, MOO, and XXOXO also show that Fridays reverse fiom the opening Monday direction. After seeing the consistency of the other markets, these pattems indicate a very different participation We might speculate that midweek release of oil stocks by the API, combined with weekend news about OPEC, have a tendency to create different reactions fiom traders.

Trading Weekday Pattems

The persistence of the weekday reversal over the past 20 years appears to be based on the nature of maiket participation. The pattems span nearly all markets with very similar results. To trade these weekday pattems, positions must be entered on the close of the prior day. Therefore, if soybeans move higher on Monday and Tuesday, we set a short on Tuesdays close with a 58*o chance of a profit on Wednesday. If thai does not occur, we can hold that short with a 5* ** chance Of recovering at least some of that loss on Thursday, if Wednesday proved profitable, neither Thursday nor Friday have interesting pattems and we would close out the trade.

FIGURE 15-7 (Continued)



[start] [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99] [100] [101] [102] [103] [104] [105] [106] [107] [108] [109] [110] [111] [112] [113] [114] [ 115 ] [116] [117] [118] [119] [120] [121] [122] [123] [124] [125] [126] [127] [128] [129] [130] [131] [132] [133] [134] [135] [136] [137] [138] [139] [140] [141] [142] [143] [144] [145] [146] [147] [148] [149] [150] [151] [152] [153] [154] [155] [156] [157] [158] [159] [160] [161] [162] [163] [164] [165] [166] [167] [168] [169] [170] [171] [172] [173] [174] [175] [176] [177] [178] [179] [180] [181] [182] [183] [184] [185] [186] [187] [188] [189] [190] [191] [192] [193] [194] [195] [196] [197] [198] [199] [200] [201] [202] [203] [204] [205]