Impact of Overnight Returns and Daytime Reversals to Future Stock Returns

1import numpy as np
2from AlgorithmImports import *
3import pandas as pd
4from pandas.core.frame import DataFrame
5class ImpactOfOvernightReturnsDaytimeReversals(QCAlgorithm):
6def Initialize(self):
7self.SetStartDate(2010, 1, 1)
8self.SetCash(100000)
9
10market:Symbol = self.AddEquity("SPY", Resolution.Daily).Symbol
11self.fundamental_count:int = 1000
12self.fundamental_sorting_key = lambda x: x.DollarVolume
13self.period:int = 13 * 21
14self.quantile:int = 10
15self.leverage:int = 5
16self.min_share_price:float = 5.
17self.exchange_codes:List[str] = ['NYS', 'NAS', 'ASE']
18self.long:List[Symbol] = []
19self.short:List[Symbol] = []
20
21self.selection_flag = False
22self.UniverseSettings.Resolution = Resolution.Daily
23self.AddUniverse(self.FundamentalSelectionFunction)
24self.Settings.MinimumOrderMarginPortfolioPercentage = 0.
25self.Schedule.On(self.DateRules.MonthEnd(market), self.TimeRules.AfterMarketOpen(market), self.Selection)
26self.settings.daily_precise_end_time = False
27def OnSecuritiesChanged(self, changes: SecurityChanges) -> None:
28for security in changes.AddedSecurities:
29    security.SetFeeModel(CustomFeeModel())
30    security.SetLeverage(self.leverage)
31        
32def FundamentalSelectionFunction(self, fundamental: List[Fundamental]) -> List[Symbol]:
33if not self.selection_flag:
34    return Universe.Unchanged
35selected:List[Fundamental] = [
36    x for x in fundamental if x.HasFundamentalData and x.Price >= self.min_share_price and \
37    x.Market == 'usa' and x.MarketCap != 0
38]
39if len(selected) > self.fundamental_count:
40    selected = [x for x in sorted(selected, key=self.fundamental_sorting_key, reverse=True)[:self.fundamental_count]]
41
42AB_NR:Dict[Fundamental, float] = {}
43
44for stock in selected:
45    symbol:Symbol = stock.Symbol
46    hist:DataFrame = self.History([symbol], self.period, Resolution.Daily)
47    if 'close' in hist.columns and 'open' in hist.columns:
48        closes:pd.Series = hist['close']
49        opens:pd.Series = hist['open']
50        if len(closes) == self.period and len(opens) == self.period:
51            # Calculate overnight and daily returns                    
52            RET_OC:pd.Series = pd.Series(closes / opens - 1)         # Open to close return
53            RET:pd.Series = pd.Series(closes).pct_change()        # Close to close return
54            RET_CO:pd.Series = ((1 + RET) / (1 + RET_OC)) - 1
55            
56            # Negative daytime reversal signal for last year                    
57            reversal_vector:List = [1 if co > 0 and oc < 0 else 0 for co, oc in zip(RET_CO, RET_OC)]
58            
59            # Slice it for every month
60            reversal_separate_months:List = [reversal_vector[x:x+21] for x in range(0, len(reversal_vector),21)]
61            NRIT:List = [month.count(1) / len(month) for month in reversal_separate_months]
62            NRIT_current_month:float = NRIT[-1]
63            NRTI_avg:float = np.average(NRIT[:-2])
64            
65            # AB_NR calc
66            AB_NR[stock] = NRIT_current_month / NRTI_avg
67
68if len(AB_NR) != 0:
69    # Sort by market cap and AB_NR
70    market_cap_values:List[float] = [x.MarketCap for x in AB_NR.keys()]
71    high_by_market_cap:List[Fundamental] = [x[0] for x in AB_NR.items() if x[0].MarketCap >= np.percentile(market_cap_values, 66)]
72
73    abnr_values:List[float] = list(AB_NR.values())
74    high_by_abnr:List[Fundamental] = [x[0] for x in AB_NR.items() if x[1] >= np.percentile(abnr_values, 80)]
75    low_by_abnr:List[Fundamental] = [x[0] for x in AB_NR.items() if x[1] <= np.percentile(abnr_values, 20)]
76
77    self.long = [x.Symbol for x in high_by_market_cap if x in high_by_abnr]
78    self.short = [x.Symbol for x in high_by_market_cap if x in low_by_abnr]
79
80return self.long + self.short
81
82def OnData(self, data: Slice) -> None:
83if not self.selection_flag:
84    return
85self.selection_flag = False
86# order execution
87targets:List[PortfolioTarget] = []
88for i, portfolio in enumerate([self.long, self.short]):
89    for symbol in portfolio:
90        if symbol in data and data[symbol]:
91            targets.append(PortfolioTarget(symbol, ((-1) ** i) / len(portfolio)))
92		
93self.SetHoldings(targets, True)
94self.long.clear()
95self.short.clear()
96    
97def Selection(self) -> None:
98self.selection_flag = True
99# Custom fee model.
100class CustomFeeModel(FeeModel):
101def GetOrderFee(self, parameters):
102fee = parameters.Security.Price * parameters.Order.AbsoluteQuantity * 0.00005
103return OrderFee(CashAmount(fee, "USD"))