相应的测试报告 系统交易测试报告 测试证券数: 888 净利润: 10507212.00 净利润率: 12.58% 总盈利: 44225104.00 总亏损: -33717892.00 交易次数: 2342 胜率: 36.17% 年均交易次数: 259.98 盈利/亏损交易次数: 847/1495 多头交易次数: 2342 空头交易次数: 0 多头盈利次数: 847 空头盈利次数: 0 最大单次盈利: 654664.50 最大单次亏损: -342463.72 平均盈利: 52213.82 平均亏损: 29582.01 平均利润: 18883.48 平均盈利/平均亏损: 1.77 最大连续盈利次数: 8 最大连续亏损次数: 24 交易平均周期数: -298.83 盈利交易平均周期: -826.28 亏损交易平均周期: -1861.39 盈利系数: 0.13 最大浮动盈利: 1405026.75 最大浮动亏损: -343938.22 总投入: 83500000.00 有效投入: 70730168.00 年回报率: 1.32% 年有效回报率: 1.55% 简单买入持有回报: 0.00% 买入持有年回报率: 0.00% 总交易额: 563107072.00 交易费: 2815533.00 交易费/利润: 0.27 测试时间: -- 共 3288 天 测试周期数: 0 平均仓位: 0.00% 最大仓位: 100.00% 平均持仓量: 0.00 最大持仓量: 159500.00 空仓周期数: 0 空仓周期比例: 0.00% 最长空仓周期: 1427 平均空仓周期: 0.00 ------------------------------买入信号统计--------------------------------- (统计所有买入信号点情况，不考虑交易测试中资金及策略造成的信号删除问题) 成功率: 57.44% 信号数量: 5319 年均信号数量: 590.46 五日获利概率: 53.77% 十日获利概率: 51.63% 二十日获利概率: 50.57% 目标周期获利概率: 50.57%
更改后的测试报告 系统交易测试报告 测试证券数: 888 净利润: 17623768.00 净利润率: 21.11% 总盈利: 83963000.00 总亏损: -66339232.00 交易次数: 6830 胜率: 34.28% 年均交易次数: 758.20 盈利/亏损交易次数: 2341/4489 多头交易次数: 6830 空头交易次数: 0 多头盈利次数: 2341 空头盈利次数: 0 最大单次盈利: 971625.19 最大单次亏损: -215982.39 平均盈利: 35866.30 平均亏损: 18704.17 平均利润: 12293.26 平均盈利/平均亏损: 1.92 最大连续盈利次数: 6 最大连续亏损次数: 19 交易平均周期数: -310.61 盈利交易平均周期: -906.23 亏损交易平均周期: -2046.93 盈利系数: 0.12 最大浮动盈利: 1297123.13 最大浮动亏损: -236556.48 总投入: 83500000.00 有效投入: 81486432.00 年回报率: 2.15% 年有效回报率: 2.20% 简单买入持有回报: 0.00% 买入持有年回报率: 0.00% 总交易额: 1889339904.00 交易费: 9446692.00 交易费/利润: 0.54 测试时间: 1996/01/01 -- 2005/01/01 共 3288 天 测试周期数: 0 平均仓位: 0.00% 最大仓位: 100.00% 平均持仓量: 0.00 最大持仓量: 211200.00 空仓周期数: 0 空仓周期比例: 0.00% 最长空仓周期: 648 平均空仓周期: 0.00 ------------------------------买入信号统计--------------------------------- (统计所有买入信号点情况，不考虑交易测试中资金及策略造成的信号删除问题) 成功率: 41.36% 信号数量: 24851 年均信号数量: 2758.70 五日获利概率: 44.37% 十日获利概率: 45.06% 二十日获利概率: 45.02% 目标周期获利概率: 45.02%
到昨天为止，我终于初步译完了第6章。 接下来整理附图和附表。 基本上是按计划完成的。 等全部修改编排好，就发给合作的朋友。 请各位通告一下进度，看看坚持下来的有几位。 另－－林先生的部分译稿已收到，还未拜读，读后再交流。
LP77兄我阐白，我已停止了直接翻译，想直接进入：用飞狐或分析家创建盈利交易系统。但看了LP77兄的进度报告，多少觉得汗颜，真的很汗颜啊。 通过此事我再一次的觉得：一个人的成功与其对一件事的把持度成正比。选定，认定目标，不受诱惑，坚持到底。就会出成绩。 从明天起坚决地回到一天一页的轨道上。
有手录的就有翻译的对应 Before you can design a good trading strategy, you must first know what one is. The validity of a strategy is defined by its trading performance. In this chapter, we cover the key performance statistics that separate the good strategies from the bad. Today, with the use of TradeStation, EasyLanguage, and backadjusted data, we can test an infinite number of different strategies on thousands of different tradable instruments. Twenty years ago only a handful of people throughout the world had this power. Power can be used or abused. In addition to performance statistics, the concepts of parameter orbustness and overoptimization are discussed. Omega Research’s System Writer was one of the first software packages to provide back testing capabilities and trading performance analysis. Through the years and a couple of name changes, this software package has evolved. In its latest incarnation, back testing has became much easier to facilitate, and the number of trading performance statistics has increased ten-fold. Unfortunately, throughout the years, one key analysis tool has not found its way into TradeStation: portfolio analysis. We have the ability to microscopically analyze a trading strategy on a particular market, but we can’t analyze the interaction of a strategy on a portfolio of different markets. Sure we can have multiple tests of a strategy on different markets, but we can’t see how trading several markets simultaneously will affect the bottom line. To illustrate this point, let’s assume that we have developed a trading strategy and we tested it on the U.S. Bond and soybean markets. After applying the strategy to the two charts, we end up with two performance reports. From these reports, we extract the profit or loss and maximum draw-down figures ( maximum draw down is the biggest $ drop in account value—actual or paper loss). To get the total profit or loss of the two markets, we simply add these two results together.
到目前为止，本项目参加人员共四人： amkr1015，lq77, 思迷思，林玲玲 目前进度： 思迷思，第一章，(已完成)，第五章，（进度未知） 林玲玲，第三章，（已完成） lq77, 第六章，（已完成） amkr1015，第十章，（进度未知） 望各位老大通报当前进度和下一步的计划。 另欢迎对系统交易有兴趣有翻译能力的朋友加入合作小组！ 这是海洋系统交易者的合作计划NO.1 第一炮能不能打响就看大家的努力和支持了！
长征第二步 To get the maximum draw down of the two markets, you can’t simply add the two values together. Draw down for anticorrelated markets almost always occur at different times. Therefore, the overall maximum draw down of the two markets will always be equal to or less than the combined draw downs. This is the magic of diversification. When you are looking at a single market analysis, you can’t see the interaction of the other markets in your portfolio. If you are actually trading the U.S. Bonds and soybeans, you know that these two markets move independently of each other. If you are losing in a bond position, you may be wining in a soybean position. In effect, offsetting a loser with a winner.. Table 5.1 illustrates the magic of diversification. We feel this is a huge oversight. Fortunately, RIAN Systems, a thirdparty software developer, has created a solution for this problem. With their software and TradeStation, you can test a basket of different commodities or stocks and evaluate how a particular strategy performs at the portfolio level. Since we have discussed portfolio analysis with such gusto, you may be asking, “Is testing a portfolio of markets simultaneously necessary in the determination of a good trading strategy?” The answer would be no. We feel that even with this limitation, you can still develop a good strategy by testing over a well-diversified portfolio. If a strategy tests well on most of the markets in a diversified portfolio, then you may have a good strategy. If a strategy only does well in a small sector of markets that are highly correlated (markets that move somewhat in unison), then you may have an okay system that will probably generate higher draw downs. Testing markets on a simultaneous basis can give a more accurate initial capital allocation reading and open the door for in-depth money management research Even through we can’t test at the portfolio level, let’s forge ahead and utilize the tools that are available TRADESTATION’S SUMMARY REPORT. Let’s take a look at the performance statistics generated by TradeStation by plotting a chart of 2000 daily bars of the Japanese Yen and applying MyStrategy-1 ( the first strategy we developed in Chapter 1). After you have applied the strategy, go under the View menu and select view strategy performance. Click on the Summary tab. A report similar to the one in Table 5.2 should show up on your screen. This summary report has most of the statistics you need to help in your search for a good trading strategy. There is a lot of information here, but some of the statistics, in our opinion, are superfluous. We will only touch upon the most influential ones.