Original paper
Abstract
We find that innovative efficiency (IE), patents or citations scaled by R&D, is a strong positive predictor of future returns after controlling for firm characteristics and risk. The IE-return relation is associated with the loading on a mispricing factor, and the high Sharpe ratio of the Efficient Minus Inefficient (EMI) portfolio suggests that mispricing plays an important role. Further tests based upon attention and uncertainty proxies suggest that limited attention contributes to the effect. The high weight of the EMI portfolio return in the tangency portfolio suggests that IE captures incremental pricing effects relative to well-known factors.
Link to the presentation slides:Â https://ssrn.com/abstract=3191073.
Keywords:Â innovative efficiency, limited attention
Trading rules
- Target small-cap equities on NYSE, AMEX, and NASDAQ, excluding financial entities, CEFs, ADRs, and REITs.
- Use the NBER patent database for patent-related data.
- Determine innovation productivity using the ratio of patents to R&D expenditures (from two years earlier).
- Consider only equities with a minimum of one approved patent.
- Categorize equities into three segments according to their innovation productivity.
- Go long on high-efficiency stocks and short on low-efficiency stocks.
- Hold the portfolio for 12 months before rebalancing.
Python code
Backtrader
import backtrader as bt
class InnovativeEfficiencyStrategy(bt.Strategy):
def __init__(self):
self.rnd_exp = dict()
def prenext(self):
self.next()
def next(self):
if len(self.datas) < 2:
return
if self.datetime.date(ago=-12).month == self.datetime.date().month:
self.rebalance_portfolio()
def rebalance_portfolio(self):
stocks_with_patents = []
for data in self.datas:
if data._name not in self.rnd_exp:
continue
if data.patents > 0:
stocks_with_patents.append(data)
if not stocks_with_patents:
return
stocks_with_patents.sort(key=lambda data: data.patents / self.rnd_exp[data._name][-2], reverse=True)
third = len(stocks_with_patents) // 3
long_stocks = stocks_with_patents[:third]
short_stocks = stocks_with_patents[-third:]
for data in long_stocks:
self.order_target_percent(data, target=1.0 / len(long_stocks))
for data in short_stocks:
self.order_target_percent(data, target=-1.0 / len(short_stocks))
for data in self.datas:
if data not in long_stocks + short_stocks:
self.order_target_percent(data, target=0.0)
def notify_order(self, order):
if order.status in [order.Submitted, order.Accepted]:
return
if order.status in [order.Completed]:
if order.isbuy():
self.log(f"BUY EXECUTED, {order.executed.price}")
elif order.issell():
self.log(f"SELL EXECUTED, {order.executed.price}")
elif order.status in [order.Canceled, order.Margin, order.Rejected]:
self.log("Order Canceled/Margin/Rejected")
def log(self, txt, dt=None):
dt = dt or self.datetime.date()
print(f"{dt.isoformat()}, {txt}")
def run_strategy():
cerebro = bt.Cerebro()
# Load data and filter stocks
# Add your own data loading and filtering code here
# Add the strategy to the cerebro instance
cerebro.addstrategy(InnovativeEfficiencyStrategy)
# Set the starting cash
cerebro.broker.setcash(100000.0)
# Run the backtest
print("Starting Portfolio Value: %.2f" % cerebro.broker.getvalue())
cerebro.run()
print("Final Portfolio Value: %.2f" % cerebro.broker.getvalue())
if __name__ == "__main__":
run_strategy()Please note that this code is a basic framework for the trading strategy in Backtrader. You will need to implement your own data loading and filtering code to obtain the small-cap stocks from NYSE, AMEX, and NASDAQ while excluding financial firms, CEFs, ADRs, and REITs. Additionally, you will need to incorporate the NBER patent database for patent information.