# 俠之大者 一起賺錢 > 來源：https://uqer.io/community/share/554048dff9f06c1c3d687fa5 在闊別七年的又一輪牛市里,炒股已經成為人們每天討論的話題. 小老弟一直以為:"俠之大者,一起賺錢,一起嗨". 故借寶地獻出珍藏多年的交易秘籍. 首先講述一下策略思路: + 標的: 流通性較好,深受大媽喜愛的滬深300成分股, 乃策略標的最佳選擇. + 買賣點: 追漲殺跌是本策略的核心思路. 在股價,成交量向上突破最近20日最高價格(量)時買入. 在股價向下突破最近10日最低價格賣出. + 頭寸規模：每只股票最多占1/10倉位. 話不多說, 上代碼: ```py start = datetime(2013, 1, 1) end = datetime(2015, 5, 25) benchmark = 'HS300' universe = set_universe('HS300') capital_base = 100000 pos_pieces = 10 enter_window = 20 exit_window = 10 def initialize(account): pass def handle_data(account): highest_price = account.get_attribute_history('highPrice', enter_window) lowest_price = account.get_attribute_history('lowPrice', exit_window) close_price = account.get_attribute_history('closePrice', exit_window) turnover_vol = account.get_attribute_history('turnoverVol', enter_window) for stock in account.universe: cnt_price = close_price[stock][-1] #account.referencePrice[stock] cnt_turnover = turnover_vol[stock][-1] if cnt_price > highest_price[stock][:-1].max() and cnt_turnover > turnover_vol[stock][:-1].max() and account.position.secpos.get(stock, 0)==0: order(stock, capital_base/pos_pieces/cnt_price) elif cnt_price < lowest_price[stock][:-1].min(): order_to(stock, 0) ```  也許已經有人發現, 其實這就是海龜交易系統. 海龜交易系統是一個完整的交易系統,它有一個完整的交易系統所應該有的所有成分，涵蓋了成功交易中的每一個必要決策： + 市場：買賣什么？ + 頭寸規模：買賣多少？ + 入市：什么時候買賣？ + 止損：什么時候放棄一個虧損的頭寸？ + 退出：什么時候退出一個盈利的頭寸？ + 戰術：怎么買賣？ 在上面的策略中, 每只股票的頭寸規模為1/10的初始資金. 《海龜交易法則》介紹了一種頭寸規模控制方法, 將頭寸分為一個個單位, 下面的策略將展示將頭寸分為N個單位, 每次產生買入信號時, 僅買入一個單位. ```py start = datetime(2013, 1, 1) end = datetime(2015, 5, 25) benchmark = 'HS300' universe = set_universe('HS300') capital_base = 100000 pos_pieces = 10 enter_window = 20 exit_window = 10 N = 4 def initialize(account): account.postion_size_hold = {} for stk in universe: account.postion_size_hold[stk] = 0 def handle_data(account): highest_price = account.get_attribute_history('highPrice', enter_window) lowest_price = account.get_attribute_history('lowPrice', exit_window) close_price = account.get_attribute_history('closePrice', exit_window) turnover_vol = account.get_attribute_history('turnoverVol', enter_window) for stock in account.universe: cnt_price = close_price[stock][-1] #account.referencePrice[stock] cnt_turnover = turnover_vol[stock][-1] if cnt_price > highest_price[stock][:-1].max() and cnt_turnover > turnover_vol[stock][:-1].max() and account.postion_size_hold[stock]<N: order(stock, capital_base/pos_pieces/cnt_price/N) account.postion_size_hold[stock] += 1 elif cnt_price < lowest_price[stock][:-1].min(): order_to(stock, 0) account.postion_size_hold[stock] = 0 ```  我們發現回撤和波動率有所下降,而收益率竟然上升了. 其實原因很簡單, 分N次買入時, 如果信號正確, 可能會提高一定的持倉成本,降低收益率; 反之如果信號有誤, 也能夠快速止損, 減少回撤. 也就是說, 頭寸規模有效的控制了風險. 以上兩個策略屬于唐安奇趨勢系統, 結束之前, 再介紹一下布林格突破系統. 布林線定義： 布林線是通過350日平均收盤加減2.5倍標準差得到的。 布林線方法： + 如果前一日的收盤價穿越了通道的頂部，則開盤做多 + 如果前一日的收盤價跌破了通道的底部，則開盤做空 在我們的這個股票策略里,我們以60日平均收盤加減2.5倍標準差作為波幅通道. ```py import numpy as np start = datetime(2013, 1, 1) end = datetime(2015, 5, 25) benchmark = 'HS300' universe = set_universe('HS300') capital_base = 100000 longest_history = 60 pos_pieces = 10 enter_window = 20 exit_window = 10 N = 4 def initialize(account): account.postion_size_hold = {} for stk in universe: account.postion_size_hold[stk] = 0 def handle_data(account): close_prices = account.get_attribute_history('closePrice', longest_history) for stock in account.universe: cnt_price = close_prices[stock][-1] #account.referencePrice[stock] mean_cp = close_prices[stock].mean() bias = 2.5*np.std(close_prices[stock]) high_channel = mean_cp + bias low_channel = mean_cp - bias if cnt_price >= high_channel and account.postion_size_hold[stock]<N: order(stock, capital_base/pos_pieces/cnt_price/N) account.postion_size_hold[stock] += 1 elif cnt_price <= low_channel: order_to(stock, 0) account.postion_size_hold[stock] = 0 ```  參考自:《海龜交易法則》 作者: 柯蒂斯·費思