Can investors time premiums associated with widely used stock/firm fundamental ratios? In their September 2018 paper entitled “It Takes Two to Tango: Fundamental Timing in Stock Market”, Fuwei Jiang, Xinlin Qi, Guohao Tang and Nan Huang use a simple moving average (SMA) trend indicator to time premiums associated with four fundamental stock/firm ratios: book-to-market (BM), earnings-to-price (EP), gross profitability (GP), and return-on-assets (ROA). In calculating these ratios, they lag accounting variables by six months to avoid look-ahead bias. For each ratio, they:
- At the end of each June, rank stocks into tenths (deciles).
- Each day, calculate value-weighted average returns for the deciles with the highest (highest BM, EP, GP, ROA) and lowest (lowest BM, EP, GP, ROA) expected returns and maintain price indexes for these two deciles.
- Each day, hold a long (short) position in the decile with highest (lowest) expected returns only when the decile price index is above (below) its 20-day SMA, indicating an upward (downward) trend. When not holding a decile, hold Treasury bills.
As benchmarks, they each year buy and hold four portfolios that are each long (short) the value-weighted deciles with the highest (lowest) expected returns for one of the fundamental ratios. While focusing on a 20-day SMA, for robustness they also test SMAs of 10, 50, 100 and 200 trading days. While focusing on value weighting, they also look at equal weighting. They run tests on both non-financial Chinese A-share stocks and non-financial U.S. common stocks. Using annual groomed fundamentals data and daily returns for Chinese stocks during January 2001-December 2017 and for U.S. stocks during July 1970-December 2017, and contemporaneous Treasury bill yields, they find that:
- For the Chinese sample:
- Gross annualized value-weighted average returns of buy-and-hold benchmarks range from 7.4% (Sharpe ratio 0.46) for EP to 9.7% (Sharpe ratio 0.64) for BM.
- Gross annualized value-weighted average returns of timed fundamental ratios based on 20-day SMAs range from 34.3% (Sharpe ratio 1.26) to 41.1% for EP (Sharpe ratio 1.35).
- Gross annualized alphas relative to the Fama-French 5-factor (market, size, book-to-market, profitability, investment) model of stock returns exceed raw returns, ranging from 34.7% for BM to 45.5% for EP.
- On a gross basis, while other SMAs substantially beat buy-and-hold, the 20-day SMA is optimal.
- For equal weighting of portfolios, benchmarks are weaker, but timing strategies remain strong, with gross annualized equal-weighted average returns ranging from 40.9% for GP to 43.3% for BM.
- In general, timing strategy turnover increases as SMA lookback interval decreases, such that breakeven long and short side trading frictions across fundamental ratios range from:
- 0.83% to 3.69% using a 200-Day SMA.
- 0.27% to 0.52% using a 10-day SMA.
- Fundamentals timing is most profitable for stocks with high idiosyncratic volatility and low liquidity.
- For the U.S. stock sample:
- Gross annualized value-weighted average returns of buy-and-hold benchmarks range from 4.1% for EP to 8.7% for ROA.
- Gross annualized value-weighted average returns of timed fundamental ratios based on 10-day SMAs range from 16.5% for GP to 28.6% for EP.
- Gross annualized 5-factor alphas are comparable to raw returns.
- On a gross average return basis, the 10-day SMA is optimal. While longer SMAs substantially beat buy-and-hold, gross performance decreases as lookback interval increases.
In summary, evidence indicates that timing stock fundamental ratios using SMAs on associated long and short side indexes may generate attractive performance.
Cautions regarding findings include:
- The sample of Chinese stocks is short for a strategy relying on annual fundamental sorts, especially since the market is young and evolving.
- The breakeven findings must cover portfolio rebalancing frictions and shorting costs. The authors assert, but do not demonstrate, net profitability. Also:
- Shorting may not always be feasible as specified due to lack of shares to borrow.
- As noted, the timing strategy is strongest among the stocks that are likely most costly to trade and most difficult to short.
- Testing multiple fundamental ratios and multiple SMAs on the same sample introduces data snooping bias, such that the most impressive results overstate expectations. The authors feature optimal findings for both Chinese stocks and U.S. stock samples.
- The methodology is beyond the reach of most investors, who would bear fees for delegating to an investment/fund manager.
See also:
