Commodity · Global · CFDs, ETFs, forwards, futures, swaps · Quarterly · Global Free

Political Uncertainty and Commodity Prices

Academic paper

AuthorsKewei Hou; Ke Tang; Bohui Zhang

Institute

?Ohio State University (OSU) - Department of Finance
Tsinghua University
?Institute of Economics, School of Social Sciences, Tsinghua University
Chinese University of Hong Kong, Shenzhen
?The Chinese University of Hong Kong, Shenzhen

Paper linkhttps://ssrn.com/abstract=3064295

Strategy in a nutshell

Targets 78 commodities (via ETF, futures, or swaps). Short commodities on June 30 before a U.S. presidential election quarter and cover on September 30, exploiting election-driven market uncertainty.

Economic rationale

Commodity prices fall before U.S. elections due to reduced demand and policy uncertainty. The effect is strongest in globally integrated commodities, close elections, and recessions, reflecting political uncertainty’s impact on consumption and pricing.

Backtest performance

Annualised return6.4%

Beta-0.007

Sharpe ratio-0.47

Sortino ratio-0.174

Win rate38%

Full Python code

1from AlgorithmImports import *
2class PoliticalUncertainty(QCAlgorithm):
3def Initialize(self):
4 self.set_start_date(1996, 1, 1)
5 self.set_cash(100_000)
6 
7 self.symbol: Symbol = self.add_data(QuantpediaFutures, 'CME_GI1', Resolution.DAILY).symbol
8 self.securities[self.symbol].set_fee_model(CustomFeeModel())
9 self.securities[self.symbol].set_leverage(2)
10 
11 self.schedule.on(self.date_rules.month_end(self.symbol), self.time_rules.at(0, 0), self.rebalance)
12def rebalance(self) -> None:
13 if self.time.date() > QuantpediaFutures.get_last_update_date()[self.symbol.value]:
14     self.liquidate()
15     return
16 year = self.time.year
17 if year % 4 == 0:   # every 4th year
18     if self.time.month == 6:
19         self.set_holdings(self.symbol, -1)
20     elif self.time.month == 9:
21         if self.portfolio.invested:
22             self.liquidate()
23     
24# Quantpedia data
25# NOTE: IMPORTANT: Data order must be ascending (datewise)
26class QuantpediaFutures(PythonData):
27_last_update_date: Dict[str, datetime.date] = {}
28@staticmethod
29def get_last_update_date() -> Dict[str, datetime.date]:
30return QuantpediaFutures._last_update_date
31def GetSource(self, config:SubscriptionDataConfig, date:datetime, isLiveMode:bool) -> SubscriptionDataSource:
32 return SubscriptionDataSource("data.quantpedia.com/backtesting_data/futures/{0}.csv".format(config.Symbol.Value), SubscriptionTransportMedium.RemoteFile, FileFormat.Csv)
33def Reader(self, config:SubscriptionDataConfig, line:str, date:datetime, isLiveMode:bool) -> BaseData:
34 data = QuantpediaFutures()
35 data.Symbol = config.Symbol
36 
37 if not line[0].isdigit(): return None
38 split = line.split(';')
39 
40 data.Time = datetime.strptime(split[0], "%d.%m.%Y") + timedelta(days=1)
41 data['back_adjusted'] = float(split[1])
42 data['spliced'] = float(split[2])
43 data.Value = float(split[1])
44 # store last update date
45 if config.Symbol.Value not in QuantpediaFutures._last_update_date:
46     QuantpediaFutures._last_update_date[config.Symbol.Value] = datetime(1,1,1).date()
47 if data.Time.date() > QuantpediaFutures._last_update_date[config.Symbol.Value]:
48     QuantpediaFutures._last_update_date[config.Symbol.Value] = data.Time.date()
49 return data
50# Custom fee model
51class CustomFeeModel(FeeModel):
52def GetOrderFee(self, parameters):
53 fee = parameters.Security.Price * parameters.Order.AbsoluteQuantity * 0.00005
54 return OrderFee(CashAmount(fee, "USD"))