Does including a measure of asset valuation as a qualifier improve the performance of intrinsic (absolute or time series) momentum? In their October 2019 paper entitled “Carry and Time-Series Momentum: A Match Made in Heaven”, Marat Molyboga, Junkai Qian and Chaohua He investigate modification of an intrinsic momentum strategy as applied to futures using the sign of the basis (difference between nearest and next-nearest futures prices) for four asset classes: equity indexes (12 series), fixed income (18 series), currencies (7 series) and commodities (28 series). Their benchmark intrinsic momentum strategy is long (short) assets with positive (negative) returns over the last 12 months, with either: (1) equal allocations to assets, or (2) dynamic allocations that each month target 40% annualized volatility for each contract series. The modified strategy limits long (short) positions to assets with positive (negative) prior-month basis. They account for frictions due to portfolio rebalancing and rolling of contracts using cost estimates from a prior study. They focus on Sharpe ratio to assess strategy performance. Using monthly returns for 65 relatively liquid futures contract series during January 1975 through December 2016, they find that:
- Benchmark intrinsic momentum is profitable on the long and short sides for each asset class, except for short sides within equity indexes and fixed income.
- The modified intrinsic momentum strategy generates net annualized Sharpe ratio 0.43 (0.73) for equal weighting (dynamic volatility targeting) of assets, compared to 0.25 (0.56) for the benchmark strategy.
- The modified strategy improves both long and short sides of the overall portfolio and is robust to extending the lookback interval for basis calculation to 12 months.
- Improvement in intrinsic momentum strategy performance from the basis filter concentrates during the first half of NBER (U.S.) economic contractions. With dynamic volatility targeting:
- During economic expansions (contractions) the modified strategy has net annualized Sharpe ratio 0.68 (1.05), compared to 0.56 (0.61)for the benchmark strategy.
- During the first (second) half of U.S. economic contractions, the modified strategy has net annualized Sharpe ratio 2.22 (0.30), compared to 0.90 (0.38) for the benchmark strategy.
- Findings based on Sortino ratio rather than Sharpe ratio are similar.
In summary, evidence indicates that adding a basis qualifier to an intrinsic momentum strategy applied to futures improves risk-adjusted performance, with improvement concentrating during the first half of U.S. economic contractions.
Cautions regarding findings include:
- Some investors may bear frictions higher than those assumed.
- The study does not look at subperiods (except for trading frictions) to determine whether profitability persists into recent years.
- Exploiting the strong performance of the modified strategy during the first half of U.S. economic contractions requires accurately predicting the timing and length of these contractions.
- The strategies described may be beyond the reach of some investors (particularly in terms of number of positions), who would bear fees for delegating to a fund manager.
See “Intrinsic Momentum Investing” for a summary of research motivating the dynamic volatility targeting approach used above.
