Equity · Emerging Markets · ETFs, funds · Monthly · Emerging Markets Free

Volatility-Weighted Short-Term Reversal Strategy in Emerging Market Equities

Academic paper

Maximize Market Timing Returns: Implementing Volatility-Weighted Bets

AuthorsPlamen Patev; Kaloyan Petkov

Institute

TWI-Shou University
?ABIR Analytics
?I Shou University, Department of International Finance
BGD. A. Tsenov Academy of Economics

Paper linkhttps://papers.ssrn.com/sol3/papers.cfm?abstract_id=3339268

Strategy in a nutshell

Invests in 15 emerging market ETFs, going long when the current month’s return is below the 3-month average and shifting to cash otherwise. Country allocations are volatility-weighted and rebalanced monthly.

Economic rationale

Volatility-weighted market timing outperforms traditional equally-weighted approaches by focusing on high-volatility markets and periods with significant expected movements, enhancing returns while managing risk.

Backtest performance

Annualised return6.14%

Volatility9.1%

Beta0.736

Sharpe ratio0.67

Sortino ratio-0.135

Win rate54%

Full Python code

1from AlgorithmImports import *
2import numpy as np
3from collections import deque
4class VolatilityWeightedShortTermReversal(QCAlgorithm):
5def Initialize(self):
6self.SetStartDate(2012, 1, 1)
7self.SetCash(100000)
8
9self.tickers:List[str] = ['FXI', 'EWH', 'EWT', 'EIDO', 'EPHE', 'EWM', 'THD', 'EWS', 'TUR', 'EWZ', 'ARGT', 'ECH', 'EPOL']
10                
11self.daily_period:int = 21
12self.monthly_period:int = 3
13self.data:Dict[Symbol, SymbolData] = { self.AddEquity(ticker, Resolution.Daily).Symbol : SymbolData(self.daily_period, self.monthly_period) for ticker in self.tickers }
14self.recent_month:int = -1
15
16def OnData(self, data: Slice) -> None:
17rebalance_flag:bool = False
18volatility:Dict[Symbol, float] = {}
19# update daily price
20for symbol, symbol_data in self.data.items():
21    if symbol in data and data[symbol]:
22        # set rebalance flag once a month
23        if self.recent_month != self.Time.month:
24            rebalance_flag = True
25            self.recent_month = self.Time.month
26        price:float = data[symbol].Value
27        symbol_data.update_price(price)
28        if rebalance_flag:
29            if symbol_data.prices_are_ready():
30                performance:float = symbol_data.performance()
31                symbol_data.update_perf(performance)
32
33                if symbol_data.returns_are_ready():
34                    # current month realized return is below the 3-month average
35                    if performance < symbol_data.monthly_avg():
36                        volatility[symbol] = symbol_data.volatility()
37
38if rebalance_flag:
39    # Trade execution
40    self.Liquidate()
41    if len(volatility) != 0:
42        total_vol:float = sum(1. / x[1] for x in volatility.items() if x[0] in volatility)
43    
44        for symbol in volatility:
45            # Volatility weighting
46            weight:float = (1. / volatility[symbol]) / total_vol
47            self.SetHoldings(symbol, weight)
48        
49class SymbolData():
50def __init__(self, period:int, month_period:int) -> None:
51self._prices:RollingWindow = RollingWindow[float](period)
52self._returns:RollingWindow = RollingWindow[float](month_period)
53
54def prices_are_ready(self) -> bool:
55return self._prices.IsReady
56
57def returns_are_ready(self) -> bool:
58return self._returns.IsReady
59
60def performance(self) -> float:
61return self._prices[0] / self._prices[self._prices.Count - 1] - 1
62
63def update_price(self, price:float) -> None:
64self._prices.Add(price)
65def update_perf(self, perf:float) -> None:
66self._returns.Add(perf)
67
68def volatility(self) -> float:
69prices:np.ndarray = np.array([x for x in self._prices])
70returns:np.ndarray = prices[:-1] / prices[1:] - 1
71return np.std(returns)
72
73def monthly_avg(self) -> float:
74return np.mean(list(self._returns))