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]
136 Can the vendor give you a higher quality LNB if the one you first receive is not strong enough? Does the vendor have a fault-tolerant system? In what manner is it fault-tolerant? • With disks running in image mode? • With backup power generators in case of power outages or brownouts? • With a redundant (secondary) system working as a backup? Data Accumulation and Correction Some vendors let you download data on the entire days trades in a very short period of time after the trading session has ended. Typically, service becomes available 30 minutes after the market session closes and transmission is repeated (recycled) every few hours thereafter. This service lets you update your database in case your computer, for whatever reason, failed to collect and process some quotes in real time. Exchanges and vendors sometimes release bad data, and most will make corrections if they are detected. Some vendors will retransmit the corrected information after the days session and some software charting products use this information to automatically update bad ticks. Other vendors require you to access their database and download what you need, then manually paste the missing data into your database. The former method is the way of the future. For example, FutureSource (FS) automatically corrects all bad ticks, and if your computer goes down, FS automatically replaces the real-time data that you missed. The same is true with Reuters Trend DataLink As for detecting bad ticks, WINdoTRADEr has a sophisticated bad price filter that catches ticks in error by as little as four ticks. Detection is based on the kinds of errors typically made by the typist at each exchange pit. Questions to Ask Does the data vendor filter out bad quotes? Does the datafeed server software filter out bad quotes? Does the server software make it easy to detect and fix bad quotes as well as enter missing quotes? Is the update process manual or automatic? Data Storage Generic Issues How do you plan to store large amounts of historical data? Some services that provide historical data expect you to store it on your hard drive. Make sure it can be set up in a format your trading software can read. Popular formats are MetaStock and
ASCII Text. Other vendors supply data on a CD and expect your software to be able to access it off the CD. The CD disk is the upcoming trend for mass storage of financial data. Disks have large capacity and are cheap to manufacture. However, there are issues regarding accuracy and convenience of use. For example, would the software let you transfer data from the CD to your hard drive, so you can fix any bad ticks? If so, will the next CD upgrade search for your clean data and automatically delete it, thinking it is doing you a service by removing "redundant" data? The data format on the CD may be incompatible with data stored on your disk, creating a problem for "on-the-fly" data integration. This issue will become more important over time as we connect data services to other software such as spreadsheets or homebrew software. Converting between these formats is time-consuming and error-prone. Since storage is so inexpensive now, there is less incentive to compress data than there has been in the past. In addition, NT offers compression built into the file system, simplifying storage and access in standard ASCII text format with good performance. So, why bother compressing data into proprietary formats? Isnt open access across applications more important than proprietary gains? As large vendors come to realize the power of open access and implement it (and they do), smaller vendors will likely follow suit. Questions to Ask What data formats (text, MetaStock, etc.) does the server support? What data format will the historical data be in? Can the software read data from a CD? If so, will it let you transfer data from the CD to your hard drive, so you can fix any bad ticks? If so, will the next CD upgrade search for your cleaned data on disk and automatically delete it, thinking it is doing you a service by removing redundant data? If so, can you disable that feature? Can the server roll together contracts on the fly? If so, can the user specify that particular contracts in the default directory are to be replaced with "corrected" ones somewhere on the users disk? Does the server support "Open Access," whereby other applications can receive data from and send data to the server by means of DDE, OLE, or some other common I/O link? Data analysis Generic Issues One consideration when choosing your charting and analysis software is cost. For example, owning a copy of charting and trading software can cost up to $2,000 or
more, a significant amount to a novice EOD trader trying to keep costs low and buying data at, say, $50 per month. However, the same $2,000 may seem insignificant when compared with the fees of real-time data services, some exceeding $1,500 per month. Another software consideration is whether you can live with the limited features of the product. Which capabilities and features do you want (see "Questions to Ask" for this section)? Trading the night session has unique requirements, besides a pot of coffee. Because you may find large spaces between bars on a time-base chart, you are almost forced to use tick-based charting software. You may experience difficulty trying to include the symbol of a new instrument into a commercial analysis software application. Sometimes the vendors are almost never accessible to solve these problems promptly, or blame the exchanges, or give their subscribers a runaround. Questions to Ask Is a payment plan available? Which software features do you want? Here are a few to get you started: Auto bad data detection Charting Auto alerts Auto stock picking Auto trade placement Auto bad data detection Internet user forum General programming language Graphics General programming language DLL, DDE, OLE connectivity Database management Fast and reputable tech support Online tech support User-definable trading systems Simulation trading Real-time analysis Portfolio management System optimization User-definable indicators User-definable trading systems Homebrew Software If you want to route the incoming data stream into your own software application, several approaches are available. For more information on the software mentioned here and their manufacturers, see Appendix for a list of Web sites. Plan A Use a ready-made server designed to be open for easy communication to your own applications: • The Universal Market Data Server (UMDS), by MarketStream Inc. is a realtime market data collection program for desktop computers running Microsoft Windows NT and Microsoft Windows 95. The server collects real-time data for all stocks, indices, futures, bonds, mutual funds, options, economic indicators, news, text services, fundamental data and other data provided by a real-time
[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]
|