ARIMA_with_eval.py

# import necessary Packages below:
import numpy as np
from pmdarima import auto_arima

def eval_model(predicted, observed):
    """
    evaluation function to assess the sign accuracy for
    validation period
    """
    pred = [1 if value < 0 else 0 for value in predicted]
    obs = (observed < 0).astype(int)
    results = (pred == obs).astype(int)
    accuracy = np.mean(results)
    return accuracy
    
def myTradingSystem(DATE, OPEN, HIGH, LOW, CLOSE, VOL, exposure, equity, settings):
    nMarkets=CLOSE.shape[1]
    log_diff = np.diff(np.log(CLOSE),axis=0)
    weights = np.zeros(nMarkets)
    # set the period for train and evaluate the built model
    periodLonger=200
    print('{} {}'.format(DATE[0],DATE[-1]))
    # set the No. of days until a new model is built (ARIMA/SARIMA order to be re-identified)
    build = settings['counter']%28==0
    
    for market in range(nMarkets):
        curr_market = log_diff[-periodLonger:,market]
        train = curr_market[:190]
        # left the last 10 days to be the validation period to evaluate the model
        test = curr_market[190:]
                 
        try:
            if build:
                model = auto_arima(train, error_action='ignore', suppress_warnings=True, seasonal=True, m=4,stepwise=True)#sarima
                #model = auto_arima(curr_market, error_action='ignore', suppress_warnings=True)#arima
                predicted = []
                
                # evluate the sign prediction for the model
                for day in range(10):
                    new_train = curr_market[:190+day]
                    model.fit(new_train)
                    prediction = model.predict(n_periods=1)[0]
                    predicted.append(prediction)
                print('==={}:{}'.format(settings['markets'][market], model.params()))
                accuracy = eval_model(predicted, test)
                print('accuracy:',accuracy)
                
                # only when model is evaluated to be accurate, it will be chosen to make an investment
                if accuracy >= 0.6:    
                    settings['models'][market] = model
                    #print('==={}:{}'.format(settings['markets'][i], model.params()))
                else:
                    model = None
            
            # model will not be re-built if it doesn't reach 28 days' period
            else:
                model = settings['models'][market]
                print('==={}'.format(settings['markets'][market]))
                pred = 0
            
            # if there's a model, make a prediction
            if model:
                #print('model')
                model.fit(curr_market)
                pred = model.predict(n_periods=1)[0]
                print('predicted: ',pred)
            
            # use the preidction to decide portfolio allocation
            if pred:
                weights[market] = pred
                
        #except:
            #print('cash')
        except Exception as e: 
            #print(e) 
            pass
            
    settings['counter']+=1
    return weights, settings


def mySettings():
    ''' Define your trading system settings here '''

    settings= {}

    # Futures Contracts
    settings['markets']  = ['CASH','F_AD','F_BO','F_BP','F_C','F_CC','F_CD','F_CL',
                            'F_CT','F_DX','F_EC','F_ES','F_FC','F_FV','F_GC','F_F',
                            'F_HG','F_HO','F_JY','F_KC','F_LB','F_LC','F_LN','F_MD',
                            'F_MP','F_NG','F_NQ','F_NR','F_O','F_OJ','F_PA','F_PL',
                            'F_RB','F_RU','F_S','F_SB','F_SF','F_SI','F_SM','F_TU',
                            'F_TY','F_US','F_W','F_XX','F_YM','F_AX','F_CA','F_DT',
                            'F_UB','F_UZ','F_GS','F_LX','F_DL','F_ZQ','F_VX','F_SS',
                            'F_AE','F_BG','F_BC','F_LU','F_DM','F_AH','F_CF','F_DZ',
                            'F_FB','F_FL','F_FM','F_FP','F_FY','F_GX','F_HP','F_LR',
                            'F_LQ','F_ND','F_NY','F_PQ','F_RR','F_RF','F_RP','F_RY',
                            'F_SH','F_SX','F_TR','F_EB','F_VF','F_VT','F_VW','F_GD', 'F_F']
    # Back testing
    settings['beginInSample'] = '20180119'
    settings['endInSample'] = '20200331'
    # # Validation
    # settings['beginInSample'] = '20171030'
    # settings['endInSample'] = '20191231'
    settings['lookback']= 504
    settings['budget']= 10**6
    settings['slippage']= 0.05
    settings['models'] = [None]*len(settings['markets'])
    settings['counter']=0

    return settings

# Evaluate trading system defined in current file.
if __name__ == '__main__':
    import quantiacsToolbox
    results = quantiacsToolbox.runts(__file__)