The strategy trades crypto-assets, longing the smallest market-cap quintile and shorting the largest, using value-weighted portfolios rebalanced weekly, excluding stablecoins and select assets with limited fundamental data.
ASSET CLASS: cryptos | REGION: Global | FREQUENCY:
Weekly | MARKET: cryptos | KEYWORD: Size, Cryptocurrencies
I. STRATEGY IN A NUTSHELL
Rank crypto-assets by market capitalization weekly. Go long on the smallest-cap quintile and short the largest-cap quintile, using value-weighted portfolios rebalanced every Tuesday.
II. ECONOMIC RATIONALE
Smaller-cap cryptos earn higher returns due to lower liquidity and higher risk. The size factor, proven in equities, delivers a persistent premium in crypto markets, independent of other risk exposures.
III. SOURCE PAPER
Is There a Value Premium in Cryptoasset Markets? [Click to Open PDF]
Luca Liebi, University of St. Gallen – Swiss Institute of Banking and Finance
<Abstract>
This study examines if blockchain fundamentals determine cryptoasset prices. Previous research has shown that non-fundamental factors affect cryptoasset prices however, whether blockchain fundamentals affect cryptoasset prices, lacks empirical research. Using data for 652 cryptoassets, I specify active addresses-to-network value as a valuation metric that captures transaction benefits. I find anomalous returns that increase with active addressesto-network value ratio, a proxy for the value anomaly. Cryptoassets with a high active addresses-to-network value ratio yield on average 3.7 percentage points higher weekly returns than cryptoassets with low active addresses-to-network value ratio, and comparable size. A four-factor model directed at capturing the value pattern in average returns performs better than the three-factor model. Importantly, my results suggest that cryptoasset prices are related to their blockchain fundamentals.
IV. BACKTEST PERFORMANCE
| Annualised Return | 134.2% |
| Volatility | 70.31% |
| Beta | -0.003 |
| Sharpe Ratio | 1.91 |
| Sortino Ratio | -1.031 |
| Maximum Drawdown | N/A |
| Win Rate | 45% |
V. FULL PYTHON CODE
from AlgorithmImports import *
from typing import List, Dict
#endregion
class SizeInCryptocurrencies(QCAlgorithm):
def Initialize(self) -> None:
self.SetStartDate(2015, 1, 1)
self.SetCash(1_000_000)
self.period: int = 30
self.quantile: int = 5
self.percentage_traded: float = 0.1
self.leverage: int = 5
self.crypto_symbols: Dict[str, str] = {
'BTC' : 'BTCUSD',
'ETH' : 'ETHUSD',
'LTC' : 'LTCUSD',
'ETC' : 'ETCUSD',
'ZEC' : 'ZECUSD',
'EOS' : 'EOSUSD',
'OMG' : 'OMGUSD',
'NEO' : 'NEOUSD',
'BAT' : 'BATUSD',
'ZRX' : 'ZRXUSD',
'TRX' : 'TRXUSD',
}
self.weight: Dict[str, float] = {}
self.SetBrokerageModel(BrokerageName.Bitfinex)
self.cap_mrkt_cur_usd: Dict[str, float] = {}
for crypto, ticker in self.crypto_symbols.items():
data: Securities = self.AddCrypto(ticker, Resolution.Daily, Market.Bitfinex)
data.SetLeverage(self.leverage)
self.AddData(CryptoNetworkData, crypto, Resolution.Daily).Symbol
self.Settings.MinimumOrderMarginPortfolioPercentage = 0.
def OnData(self, data: Slice) -> None:
crypto_data_last_update_date: Dict[Symbol, datetime.date] = CryptoNetworkData._last_update_date
cap_mrkt_cur_usd: Dict[str, float] = {}
weight: Dict[str, float] = {}
# Store daily data.
for crypto, ticker in self.crypto_symbols.items():
if ticker in data and data[ticker]:
if crypto in crypto_data_last_update_date:
if self.Securities[crypto].GetLastData() and self.Time.date() <= crypto_data_last_update_date[crypto]:
cap_mrkt_cur: float = self.Securities[crypto].GetLastData().Capmrktcurusd
if cap_mrkt_cur != 0:
cap_mrkt_cur_usd[ticker] = cap_mrkt_cur
# tuesday
if self.Time.date().weekday() == 1:
# Number in quantile variable shouldn't be 0.
if len(cap_mrkt_cur_usd) < self.quantile:
self.Liquidate()
return
sorted_by_cap: List[str] = [x[0] for x in sorted(cap_mrkt_cur_usd.items(), key = lambda item: item[1], reverse=True)]
quantile: int = int(len(sorted_by_cap) / self.quantile)
# Long (short) the quintile of crypto-assets with the lowest (highest) size measure.
long: List[str] = sorted_by_cap[-quantile:]
short: List[str] = sorted_by_cap[:quantile]
for i, portfolio in enumerate([long, short]):
mc_sum: float = sum(list(map(lambda x: cap_mrkt_cur_usd[x], portfolio)))
for ticker in portfolio:
weight[ticker] = ((-1)**i) * cap_mrkt_cur_usd[ticker] / mc_sum
# Trade execution
invested: List[str] = [x.Key.Value for x in self.Portfolio if x.Value.Invested]
for symbol in invested:
if symbol not in weight:
self.Liquidate(symbol)
portfolio: List[PortfolioTarget] = [PortfolioTarget(ticker, w * self.percentage_traded) for ticker, w in weight.items() if ticker in data and data[ticker]]
self.SetHoldings(portfolio)
# Crypto network data.
# NOTE: IMPORTANT: Data order must be ascending (datewise)
# Data source: https://coinmetrics.io/community-network-data/
class CryptoNetworkData(PythonData):
_last_update_date: Dict[Symbol, datetime.date] = {}
@staticmethod
def get_last_update_date() -> Dict[Symbol, datetime.date]:
return CryptoNetworkData._last_update_date
def GetSource(self, config: SubscriptionDataConfig, date: datetime, isLiveMode: bool) -> SubscriptionDataSource:
return SubscriptionDataSource(f"data.quantpedia.com/backtesting_data/crypto/{config.Symbol.Value}_network_data.csv", SubscriptionTransportMedium.RemoteFile, FileFormat.Csv)
# File exmaple:
# 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
# 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
def Reader(self, config: SubscriptionDataConfig, line: str, date: datetime, isLiveMode: bool) -> BaseData:
data: CryptoNetworkData = CryptoNetworkData()
data.Symbol = config.Symbol
cols: List[str] = ['CapMrktCurUSD']
try:
if not line[0].isdigit():
header_split = line.split(',')
self.col_index = [header_split.index(x) for x in cols]
return None
split: str = line.split(',')
data.Time = datetime.strptime(split[0], "%Y-%m-%d") + timedelta(days=1)
for i, col in enumerate(cols):
data[col] = float(split[self.col_index[i]])
data.Value = float(split[self.col_index[0]])
if config.Symbol.Value not in CryptoNetworkData._last_update_date:
CryptoNetworkData._last_update_date[config.Symbol.Value] = datetime(1,1,1).date()
if data.Time.date() > CryptoNetworkData._last_update_date[config.Symbol.Value]:
CryptoNetworkData._last_update_date[config.Symbol.Value] = data.Time.date()
except:
return None
return dataVI. Backtest Performance
