Crypto · Global · cryptos · Weekly · Global Free

Value Factor in Cryptocurrencies

Academic paper

'Know When to Hodl Em, Know When to Fodl Em': An Investigation of Factor Based Investing in the Cryptocurrency Space

AuthorsStefan Hubrich

Institute

T. Rowe Price Program for Charitable Giving
?T.Rowe Price

Paper linkhttps://ssrn.com/abstract=3055498

Strategy in a nutshell

Trades 11 cryptocurrencies using a standardized on-chain value factor (market value ÷ 7-day transaction average). Positions are equally weighted with 10% portfolio exposure, long-only, and rebalanced weekly. Optional risk-based weighting is also discussed.

Economic rationale

High volatility enhances returns with smaller capital allocations. The value factor adds modest alpha and can complement momentum strategies in crypto portfolios, showing meaningful risk-adjusted performance despite the short sample period.

Backtest performance

Annualised return8%

Volatility7.1%

Beta0.071

Sharpe ratio1.13

Sortino ratio1.093

Win rate68%

Full Python code

1from AlgorithmImports import *
2from typing import List, Dict
3class ValueFactorInCryptocurrencies(QCAlgorithm):
4def Initialize(self) -> None:
5 self.SetStartDate(2015, 1, 1)
6 self.SetCash(1_000_000)
7 
8 self.period: int = 7
9 self.percentage_traded: float = 0.1
10 
11 self.symbols: Dict[str, str] = {
12     'BTC' : 'BTCUSD',
13     'ETH' : 'ETHUSD', 
14     'LTC' : 'LTCUSD', 
15     'ETC' : 'ETCUSD',
16     'XMR' : 'XMRUSD',
17     'ZEC' : 'ZECUSD',
18 }
19  
20 self.data: Dict[str, SymbolData] = {}
21 
22 self.SetBrokerageModel(BrokerageName.Bitfinex)  
23 for crypto, ticker in self.symbols.items():
24     data: Securities = self.AddCrypto(ticker, Resolution.Daily, Market.Bitfinex)
25     
26     self.AddData(CryptoNetworkData, crypto, Resolution.Daily)
27     self.data[crypto] = SymbolData(self.period)
28 
29 self.Settings.MinimumOrderMarginPortfolioPercentage = 0.
30def OnData(self, data: Slice) -> None:
31 crypto_data_last_update_date: Dict[Symbol, datetime.date] = CryptoNetworkData.get_last_update_date()
32 # Store daily data.
33 for crypto, ticker in self.symbols.items():
34     if crypto in data and data[crypto]:
35         cap_mrkt_cur_usd: float = data[crypto].Capmrktcurusd
36         txtfr_val_adj_usd: float = data[crypto].Txtfrvaladjusd
37         self.data[crypto].update_attributes(cap_mrkt_cur_usd, txtfr_val_adj_usd)
38 if self.Time.date().weekday() != 0:
39     return
40 long: List[str] = []
41 short: List[str] = []
42 
43 for crypto, ticker in self.symbols.items():
44     if self.Securities[crypto].GetLastData() and self.Time.date() > crypto_data_last_update_date[crypto]:
45         continue
46     
47     if self.data[crypto].is_ready():
48         valuation_metric = self.data[crypto].valuation_metric()
49         
50         if valuation_metric > 0.:
51             long.append(self.Symbol(ticker))
52         else:
53             short.append(self.Symbol(ticker))
54 
55 # Trade Execution
56 invested: List[Symbol] = [x.Key for x in self.Portfolio if x.Value.Invested]
57 for symbol in invested:
58     if symbol not in long + short:
59         self.Liquidate(symbol)
60 for i, portfolio in enumerate([long, short]):
61     for symbol in portfolio:
62         if symbol in data and data[symbol]:
63             self.SetHoldings(symbol, round((((-1) ** i) / len(portfolio)) * self.percentage_traded, 2))
64 
65class SymbolData():
66def __init__(self, period: int) -> None:
67 self.transactions: RollingWindow = RollingWindow[float](period)
68 self.curr_market_cap: float = 0.
69 
70def update_attributes(self, current_market_value: float, num_of_transactions: int) -> None:
71 self.transactions.Add(num_of_transactions)
72 self.curr_market_cap = current_market_value
73 
74def valuation_metric(self) -> float:
75 trailing_data: np.ndarray = np.array([x for x in self.transactions])
76 return self.curr_market_cap / np.mean(trailing_data)
77 
78def is_ready(self) -> bool:
79 return self.transactions.IsReady
80 
81# Crypto network data.
82# NOTE: IMPORTANT: Data order must be ascending (datewise)
83# Data source: https://coinmetrics.io/community-network-data/
84class CryptoNetworkData(PythonData):
85_last_update_date: Dict[Symbol, datetime.date] = {}
86@staticmethod
87def get_last_update_date() -> Dict[Symbol, datetime.date]:
88return CryptoNetworkData._last_update_date
89def GetSource(self, config: SubscriptionDataConfig, date: datetime, isLiveMode: bool) -> SubscriptionDataSource:
90 return SubscriptionDataSource(f"data.quantpedia.com/backtesting_data/crypto/{config.Symbol.Value}_network_data.csv", SubscriptionTransportMedium.RemoteFile, FileFormat.Csv)
91# File exmaple:
92# date,AdrActCnt,AdrBal1in100KCnt,AdrBal1in100MCnt,AdrBal1in10BCnt,AdrBal1in10KCnt,AdrBal1in10MCnt,AdrBal1in1BCnt,AdrBal1in1KCnt,AdrBal1in1MCnt,AdrBalCnt,AdrBalNtv0.001Cnt,AdrBalNtv0.01Cnt,AdrBalNtv0.1Cnt,AdrBalNtv100Cnt,AdrBalNtv100KCnt,AdrBalNtv10Cnt,AdrBalNtv10KCnt,AdrBalNtv1Cnt,AdrBalNtv1KCnt,AdrBalNtv1MCnt,AdrBalUSD100Cnt,AdrBalUSD100KCnt,AdrBalUSD10Cnt,AdrBalUSD10KCnt,AdrBalUSD10MCnt,AdrBalUSD1Cnt,AdrBalUSD1KCnt,AdrBalUSD1MCnt,AssetEODCompletionTime,BlkCnt,BlkSizeMeanByte,BlkWghtMean,BlkWghtTot,CapAct1yrUSD,CapMVRVCur,CapMVRVFF,CapMrktCurUSD,CapMrktFFUSD,CapRealUSD,DiffLast,DiffMean,FeeByteMeanNtv,FeeMeanNtv,FeeMeanUSD,FeeMedNtv,FeeMedUSD,FeeTotNtv,FeeTotUSD,FlowInExNtv,FlowInExUSD,FlowOutExNtv,FlowOutExUSD,FlowTfrFromExCnt,HashRate,HashRate30d,IssContNtv,IssContPctAnn,IssContPctDay,IssContUSD,IssTotNtv,IssTotUSD,NDF,NVTAdj,NVTAdj90,NVTAdjFF,NVTAdjFF90,PriceBTC,PriceUSD,ROI1yr,ROI30d,RevAllTimeUSD,RevHashNtv,RevHashRateNtv,RevHashRateUSD,RevHashUSD,RevNtv,RevUSD,SER,SplyAct10yr,SplyAct180d,SplyAct1d,SplyAct1yr,SplyAct2yr,SplyAct30d,SplyAct3yr,SplyAct4yr,SplyAct5yr,SplyAct7d,SplyAct90d,SplyActEver,SplyActPct1yr,SplyAdrBal1in100K,SplyAdrBal1in100M,SplyAdrBal1in10B,SplyAdrBal1in10K,SplyAdrBal1in10M,SplyAdrBal1in1B,SplyAdrBal1in1K,SplyAdrBal1in1M,SplyAdrBalNtv0.001,SplyAdrBalNtv0.01,SplyAdrBalNtv0.1,SplyAdrBalNtv1,SplyAdrBalNtv10,SplyAdrBalNtv100,SplyAdrBalNtv100K,SplyAdrBalNtv10K,SplyAdrBalNtv1K,SplyAdrBalNtv1M,SplyAdrBalUSD1,SplyAdrBalUSD10,SplyAdrBalUSD100,SplyAdrBalUSD100K,SplyAdrBalUSD10K,SplyAdrBalUSD10M,SplyAdrBalUSD1K,SplyAdrBalUSD1M,SplyAdrTop100,SplyAdrTop10Pct,SplyAdrTop1Pct,SplyCur,SplyExpFut10yr,SplyFF,SplyMiner0HopAllNtv,SplyMiner0HopAllUSD,SplyMiner1HopAllNtv,SplyMiner1HopAllUSD,TxCnt,TxCntSec,TxTfrCnt,TxTfrValAdjNtv,TxTfrValAdjUSD,TxTfrValMeanNtv,TxTfrValMeanUSD,TxTfrValMedNtv,TxTfrValMedUSD,VelCur1yr,VtyDayRet180d,VtyDayRet30d
93# 2009-01-09,19,19,19,19,19,19,19,19,19,19,19,19,19,0,0,19,0,19,0,0,0,0,0,0,0,0,0,0,1614334886,19,215,860,16340,0,0,0,0,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,9.44495122962963E-7,0,950,36500,100,0,950,0,1,0,0,0,0,1,0,0,0,0,11641.53218269,1005828380.584716757433,0,0,950,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,950,950,950,950,950,950,950,950,950,950,950,950,950,0,0,0,0,0,0,0,0,0,0,0,0,0,950,50,50,950,17070250,950,1000,0,1000,0,0,0,0,0,0,0,0,0,0,0,0,0
94def Reader(self, config: SubscriptionDataConfig, line: str, date: datetime, isLiveMode: bool) -> BaseData:
95 data: CryptoNetworkData = CryptoNetworkData()
96 data.Symbol = config.Symbol
97 try:
98     cols:str = ['CapMrktCurUSD', 'TxTfrValAdjUSD']
99     if not line[0].isdigit():
100         header_split = line.split(',')
101         self.col_index = [header_split.index(x) for x in cols]
102         return None
103     split = line.split(',')
104     
105     data.Time = datetime.strptime(split[0], "%Y-%m-%d") + timedelta(days=1)
106     
107     for i, col in enumerate(cols):
108         data[col] = float(split[self.col_index[i]])
109     data.Value = float(split[self.col_index[0]])
110         
111     if config.Symbol.Value not in CryptoNetworkData._last_update_date:
112         CryptoNetworkData._last_update_date[config.Symbol.Value] = datetime(1,1,1).date()
113     if data.Time.date() > CryptoNetworkData._last_update_date[config.Symbol.Value]:
114         CryptoNetworkData._last_update_date[config.Symbol.Value] = data.Time.date()
115 except:
116     return None
117 return data