Can you post a few of those .SPX trades? Thnx. Also, I don't use Fidel. Here is some Esig data. Are we using the same data(date, time, open, high, low, close, vol)? Code: 9/18/2008 6:30 1157.08 1179.06 1157.08 1173.16 59 9/18/2008 6:45 1172.47 1180.82 1168.21 1179.67 60 9/18/2008 7:00 1179.59 1184.22 1176.42 1176.59 59 9/18/2008 7:15 1176.44 1178.88 1173.08 1177.91 59 9/18/2008 7:30 1177.68 1178.08 1172.93 1172.93 60 9/18/2008 7:45 1172.66 1172.66 1164.39 1164.42 60 9/18/2008 8:00 1164.66 1166.61 1161.15 1161.23 60 9/18/2008 8:15 1160.71 1160.82 1155.33 1158.84 59 9/18/2008 8:30 1159.17 1164.48 1156.65 1158.7 59 9/18/2008 8:45 1158.07 1158.61 1149.68 1158.14 60 9/18/2008 9:00 1157.89 1157.89 1149.96 1154.66 60 9/18/2008 9:15 1154.7 1158.45 1150.97 1151.87 60 9/18/2008 9:30 1153.06 1153.06 1143.57 1145.3 60 9/18/2008 9:45 1144.37 1144.37 1134.26 1134.26 60 9/18/2008 10:00 1133.59 1152.77 1133.5 1149.65 60 9/18/2008 10:15 1150.15 1157.67 1150.15 1154.88 59 9/18/2008 10:30 1155.65 1173.78 1155.65 1164.86 60 9/18/2008 10:45 1165.08 1171.37 1164 1167.64 60 9/18/2008 11:00 1167.82 1169.11 1155.94 1161.45 59 9/18/2008 11:15 1161.37 1163.43 1157.89 1160.09 60 9/18/2008 11:30 1159.16 1164.65 1154.67 1157.36 58 9/18/2008 11:45 1158.16 1169.67 1157.86 1167.31 60 9/18/2008 12:00 1167.91 1199.12 1166.82 1195.19 59 9/18/2008 12:15 1195.02 1204.01 1190.12 1197.46 59 9/18/2008 12:30 1198.3 1208.51 1197.66 1206.66 60 9/18/2008 12:45 1207.67 1211.14 1196.57 1200.5 60 9/19/2008 6:30 1213.11 1259.82 1213.11 1256.55 60 9/19/2008 6:45 1255.88 1260.89 1250.25 1250.25 60 9/19/2008 7:00 1249.86 1250.38 1238.58 1240.93 60 9/19/2008 7:15 1240.85 1247.67 1240.8 1242.26 58 9/19/2008 7:30 1242.17 1244.72 1238.72 1240.46 60 9/19/2008 7:45 1240.07 1240.5 1236.09 1238.49 60 9/19/2008 8:00 1238.55 1250.81 1238.55 1247.27 60 9/19/2008 8:15 1247.59 1250.47 1246.31 1246.48 59 9/19/2008 8:30 1246.64 1249.02 1244.16 1248.31 59 9/19/2008 8:45 1248.48 1255.17 1247.69 1255.06 59 9/19/2008 9:00 1255.92 1257.61 1254.42 1255.85 60 9/19/2008 9:15 1256.54 1262.02 1255.94 1262.02 59 9/19/2008 9:30 1262.49 1265.12 1260.5 1260.86 59 9/19/2008 9:45 1260.59 1260.7 1256.81 1258.37 60 9/19/2008 10:00 1258.29 1258.29 1253.41 1255.88 59 9/19/2008 10:15 1255.42 1258.66 1255.25 1257.8 58 9/19/2008 10:30 1257.37 1257.37 1250.69 1251.29 60 9/19/2008 10:45 1251.16 1252.29 1249.84 1250.04 57 9/19/2008 11:00 1249.97 1250.39 1246.6 1249.6 60 9/19/2008 11:15 1249.28 1249.96 1246.41 1249.96 58 9/19/2008 11:30 1250.02 1252.08 1247.14 1250.55 58 9/19/2008 11:45 1250.34 1253.12 1248.39 1249.34 59 9/19/2008 12:00 1249.47 1251.82 1239.52 1239.52 59 9/19/2008 12:15 1239.42 1246.17 1237.17 1243.78 60 9/19/2008 12:30 1245.24 1250.74 1240.46 1249.38 60 9/19/2008 12:45 1249.51 1253.19 1248.06 1251.72 60
Guys this isn't rocket science. If the system's worth a damn it should also work on daily bars. Why can't we isolate long entries for example and see how they perform on a few thousand stocks over a multi year period? That should give us plenty of samples. With few exceptions, a system that only works on one tradable or time frame is curve fitted mush. This cow pie uses MACD and stochastics... it shouldn't be too difficult to cut through the BS.
I can't really help you, I only have fidelity data. And I also found if you filter data outside of market hours it won't work. So you need that data, and I hope that helps., Yeah the problem is it works basically on 8 am to 5 pm EST data.
If you can't duplicate the results, either bwolinsky made a mistake someplace or this system is super curve fit (my opinion) and performs well while looking through a small window of data. There might be something here, but I'm not sure this effort is showing what it is.
That data is RTH. Should be exactly the same as Fido RTH, just shifted for West Coast. If you are telling me I need to have AH data in there too, then that's okay. But I don't see that requirment in the script. But the fact remains, that you are getting very unrealistic results. You have something wrong. If the Jackers had anything remotely that good, they would have shut up years ago and just traded their way to billions. Good luck.
I'm just as astonished, but I can say the code is exactly the same, save for mine that doesn't have an EOD close, or at least one that works.
If you're talking about the volume condition: Volume[bar-1]>20000, there's nothing magic about a volume of 20000. This only shows the "system" has been optimized to a specific market at a specific time. Even in the same market, that number won't always be "right" as participation and conditions change. If it can't be made adaptable (something like x% of a y period moving average of V) then it has no value.