Tiger Quant

环境准备

快速上手

1. 导出项目到本地

2. 编写策略

 public class BestLimitAlgo extends AlgoTemplate {
 public BestLimitAlgo() {
 }
 public BestLimitAlgo(Map<String, Object> settings) {
 super(settings);
 }
 @Override
 public void init() {
 this.direction = (String) settings.get("direction");
 this.volume = (Integer) settings.get("volume");
 this.symbol = (String) settings.get("symbol");
 }
 @Override
 public void onStart() {
 //
 barGenerator = new BarGenerator(bar -> onBar(bar));
 //订阅 AAPL 行情
 List<String> symbols = new ArrayList<>();
 symbols.add("AAPL");
 subscribe(symbol);
 }
 
 @Override
 public void onTick(Tick tick) {
 
 }
 
 private void buyBestLimit() {
 int orderVolume = volume - traded;
 orderPrice = lastTick.getBidPrice();
 if (orderPrice < 10) {
 buy(symbol, orderPrice, orderVolume, OrderType.LMT);
 }
 }
 
 private void sellBestLimit() {
 int orderVolume = volume - traded;
 orderPrice = lastTick.getAskPrice();
 if (orderPrice > 12) {
 sell(symbol, orderPrice, orderVolume, OrderType.LMT);
 }
 }
 
 @Override
 public void onOrder(Order order) {
 
 }
 
 @Override
 public void onTrade(Trade trade) {
 }
 
 @Override
 public void onBar(Bar bar) {
 log("onBar {}", bar);
 }
}

3. 完成对应的策略配置及券商接入配置

4. 编译运行策略

mvn -U clean install -Dmaven.test.skip=true

 java -jar tquant-bootstrap-1.0.0-jar-with-dependencies.jar -a /yourpath/algo_setting.json -g /yourpath/tiger_gateway_setting.json

5. 停止执行策略

 ps -ef|grep TigerQuantBootstrap
 kill {pid}

配置说明

策略配置

{
 "BestLimitAlgo": {
 "enable": false,
 "class":"com.tquant.algorithm.algos.BestLimitAlgo",
 "direction": "BUY",
 "volume": 100,
 "symbol": "00700"
 },
 "DmaAlgo": {
 "enable": false,
 "class":"com.tquant.algorithm.algos.DmaAlgo",
 "direction": "BUY",
 "price": 13.2,
 "volume": 100
 },
 "SpreadAlgo": {
 "enable": true,
 "class":"com.tquant.algorithm.algos.MacdAlgo",
 "symbol": "SPY",
 "bars": 100
 }
}

全局配置

接入券商配置

{
 "gateway": "TigerGateway",
 "apiLogEnable": true,
 "apiLogPath": "log/",
 "tigerId": "2015xxxx",
 "account": "20190419163707900",
 "privateKey": "MIICeQIBADANBgkqhkiG9w0BAQEFAASCAmMwggJfAgEAAoGBAL7..."
 }

Thetagang策略说明

 "ThetaGangAlgo": {
 "enable": true, //是否启用策略
 "class":"com.tquant.algorithm.algos.ThetaGangAlgo", //对应策略实现的代码路径
 "account": {
 "account_id": "20190419163107900", //使用的账号信息,可以配置为模拟账号或综合账号
 "cancel_orders": true, //策略执行前,是否要取消已经挂出去的订单
 "margin_usage": 0.5, //该策略要使用的资金占总资产的比例,如 0.5 表示为 50%
 },
 //期权链是延迟加载的,在你确定期权希腊值(delta)或期权价格之前,你需要先扫描期权链。
 //这里的设置是告诉thetagang策略要加载多少个合约。不要让这些值太大,因为它们会导致扫描期权链过多,可能会失败。
 //如果你遇到thetagang找不到合适的contract的问题,可以试试略微增加这些值。
 "option_chains": {
 "expirations": 4, //从期权链上加载的到期日数量
 "strikes": 15 //从期权链上加载的行权价数量
 },
 "roll_when": {
 "pnl": 0.9, //盈亏(pnl)到达 90% 时,需要滚动持仓
 //或者,当离到期日<=15天,并且盈亏至少到min_pnl (min_pnl默认为0)时。
 //注意:对于期权最终是深度 ITM 的情况,特别是在卖出套期保值(covered call)的时候,盈亏有可能是负数、
 //示例:如果你想在这种情况下进行滚动,请将min_pnl设置为一个负值,如-1(代表-100%)。
 "dte": 15, 
 "min_pnl": 0.2,
 //可选的: 当盈亏达到这个阈值时,创建一个平仓单。
 //这会覆盖其他参数,也就是说,它忽略了dte和其他参数。如果不指定,它没有任何作用。
 //这可以处理这样的情况,即你有长期的期权,已经慢慢变得毫无价值,你只是想把它们从你的投资组合中删除。
 "close_at_pnl": 0.99,
 "calls": {
 "itm": true,
 "credit_only": false //只有在有合适的contract可用时,才会进行滚动,从而获得一个 credit。
 },
 "puts": {
 "itm": false,
 "credit_only": false //只有在有合适的contract可用时,才会进行滚动,从而获得一个 credit。
 }
 },
 "write_when": {
 "calls": {
 "green": true, //可选的,只有在对应标的上涨时才会write。
 //有了套期保值(covered call),我们就可以通过这个因素来限定写的套期保值的数量。
 //在1.0的时候,我们对100%的头寸写覆盖性看涨。
 //在0.5时,我们只写 50%的头寸。这个值必须在1和0之间(含)。
 "cap_factor": 1
 },
 "puts": {
 "red": true //可选的,只有在对应标的下跌时才会write。
 }
 },
 "target": {
 "dte": 45, // Target 45 or more days to expiry
 "delta": 0.3, //Target delta of 0.3 or less. Defaults to 0.3 if not specified.
 // When writing new contracts (either covered calls or naked puts), or rolling
 // before `roll_when.dte` is reached, never write more than this amount of
 // contracts at once. This can be useful to avoid bunching by spreading contract
 // placement out over time (and possibly expirations) in order to protect
 // yourself from large swings. This value does not affect rolling existing
 // contracts to the next expiration. This value is expressed as a percentage of
 // buying power based on the market price of the underlying ticker, as a range
 // from [0.0-1.0].
 //
 // Once the `roll_when.dte` date is reached, all the remaining positions are
 // rolled regardless of the current position quantity.
 //
 // Defaults to 5% of buying power. Set this to 1.0 to effectively disable the
 // limit.
 "maximum_new_contracts_percent": 0.05, 
 // Minimum amount of open interest for a contract to qualify
 "minimum_open_interest": 10
 },
 // Optional: specify delta separately for puts/calls. Takes precedent over
 // target.delta.
 //
 // [target.puts]
 // delta = 0.5
 // [target.calls]
 // delta = 0.3
 "symbols": {
 # NOTE: Please change these symbols and weights according to your preferences.
 # These are provided only as an example for the purpose of configuration. These
 # values were chosen as sane values should someone decide to run this code
 # without changes, however it is in no way a recommendation or endorsement.
 #
 # You can specify the weight either as a percentage of your buying power (which
 # is calculated as your NLV * account.margin_usage), or in terms of parts. Parts
 # are summed from all symbols, then the weight is calculated by dividing the
 # parts by the total parts.
 #
 # You should try to choose ETFs or stocks that:
 #
 # 1) Have sufficient trading volume for the underlying
 # 2) Have standard options contracts (100 shares per contract)
 # 3) Have options with sufficient open interest and trading volume
 #
 # The target delta may also be specified per-symbol, and takes precedence over
 # `target.delta` or `target.puts/calls.delta`. You can specify a value for the
 # symbol, or override individually for puts/calls.
 "SPY": {
 "weight": 0.4
 },
 "AAPL": {
 "weight": 0.3
 },
 "MSFT": {
 "weight": 0.3
 }
 }
 }

外部依赖

ta4j

问题反馈