This article presents 7 indicators showing logical and historical links with stock market returns. Formulas and backtests are disclosed, there is no black box. They are arguably among the best long-term timing indicators, and they are even better used together.
SEP
The simplified equity risk premium (hereafter SEP) aims at measuring the difference between the expected annual return of a stock index and a safe bond yield. SEP is inspired by the Fed model.
It is defined as: SEP = (E/P) - GS10, where:
- E is the aggregate estimate earnings of the US large cap stock index S&P 500. Some implementations of the Fed model look forward by using projected future EPS. Our "E" reflects the current state of the economy.
- P is the monthly price of S&P 500, defined as the average of daily closing prices.
- GS10 is the 10-Year Treasury Constant Maturity Rate (GS10 is the name of the data series in Saint Louis Fed's database)
On the 1st day of month "m", we can make decisions using SEP(m-1), calculated from the data of the month ending the day before.
I will show market timing test results based on monthly decisions. Indicators are observed on the 1st day of every month. Every indicator is tested by calculating the performance of an investment in the S&P 500 (VOO, or IVV, SPY) with a market timing strategy going gradually out of the market during the month of a bearish signal. Gradualness is simulated using the average of daily closing prices as monthly price. Using smoothed monthly prices lowers sensitivity to short-term moves. There is no risk to design a model unwillingly curve-fitted to monthly opening days. It is also more realistic for fund managers who cannot make a big move on a single day. The following tests simulate going to cash on a bearish signal. This is rarely the best strategy. Opening or increasing hedging positions incurs lower trading costs when positions are numerous or not very liquid. It also keeps dividends coming when there are some.
The next tables show simulation with bearish signals when the 3-month simple moving average of SEP (3mma) is below both the 2-year average (24mma) and the 5-year average (60mma), and bullish otherwise.
The metrics are:
- CAGR: the annualized return in percentage points.
- Ddmax: the maximum drawdown depth also in percentage.
- DLmax: the maximum duration in months.
- MAR: a risk-adjusted performance ratio defined as MAR = CAGR/Ddmax.
- The first column gives the starting year, the end date is always 1/1/2019.
For all tables, benchmark data are repeated in italic to facilitate comparisons (S&P 500, buy and hold).
Since CAGR MAR Ddmax DLmax CAGR MAR Ddmax DLmax 1993 7.12 0.14 50.82 80 7.21 0.47 15.22 31 1956 6.68 0.13 50.82 89 6.73 0.17 38.92 74 1913 5.45 0.06 84.76 299 5.88 0.11 53.40 345 1876 4.55 0.05 84.76 299 4.74 0.09 53.40 345
Chart since 1993:
Our SEP indicator improves the drawdown and MAR ratio on all studied intervals.
SPX moving average cross-over
As it is used in this study, SPX is a monthly data series calculated as the average of daily closing prices during the month. On the 1st day of month "m", we can make decisions using the value for month "m-1", noted SPX(m-1).
Two indicators are tested below. The first one is bearish when the stock index is below its 10-month simple moving average (hereafter named 10mma) and bullish otherwise. The second one is bearish when the 3-month simple moving average (3mma) is below the 12-month simple moving average (12mma) and bullish otherwise. A combination is also tested: bearish when both bearish conditions are met and bullish otherwise. We use the same assumptions and metrics than in SEP simulations and the benchmark is in italic.
Bearish signal: SPX(m-1) < 10mma:
Since CAGR MAR Ddmax DLmax CAGR MAR Ddmax DLmax 1993 7.12 0.14 50.82 80 7.60 0.53 14.42 38 1956 6.68 0.13 50.82 89 5.79 0.21 27.11 76 1913 5.46 0.06 84.76 299 5.91 0.11 53.65 184 1872 4.37 0.05 84.76 299 4.86 0.09 53.65 184
Bearish signal: 3mma < 12mma:
Since CAGR MAR Ddmax DLmax CAGR MAR Ddmax DLmax 1993 7.12 0.14 50.82 80 7.65 0.54 14.07 40 1956 6.68 0.13 50.82 89 5.40 0.20 26.84 72 1913 5.46 0.06 84.76 299 5.71 0.15 38.71 81 1872 4.37 0.05 84.76 299 4.83 0.12 38.71 141
Bearish signal: SPX (m-1)< 10mma and 3mma < 12mma:
Since CAGR MAR Ddmax DLmax CAGR MAR Ddmax DLmax 1993 7.12 0.14 50.82 80 8.30 0.64 13.04 39 1956 6.68 0.13 50.82 89 6.10 0.22 28.30 72 1913 5.46 0.06 84.76 299 6.15 0.12 52.97 99 1872 4.37 0.05 84.76 299 5.05 0.10 52.97 166
Chart since 1993:
The three SPX indicators improve the risk-adjusted performance and reduce drawdown depth and length on all studied intervals. They generally improve the annualized return but result in lagging the benchmark since 1956.
CAB
The Chemical Activity Barometer, hereafter named CAB, is the result of proprietary information and calculation by the American Chemistry Council. It is designed as a leading indicator based on chemical activity rather than an indicator of chemical activity. CAB has been published since 1948, and the data series has been calculated backward to start in 1912. The next chart shows some coincidence between downturns in CAB (red) and recessions spotted by the National Bureau of Economic Research (grey).
CAB chart by the American Chemistry Council
CAB is published monthly close to the end of the month. On the first day of every month, we can use the value published a few days before for the previous month. If we are in month "m", I name it CAB(m-1). However, because revisions are substantial and frequent for the most recent reading, I prefer ignoring it and using the most recent value already revised once, which is for the prior month: CAB(m-2). Subsequent secondary revisions are smaller and much less likely to change the trend.
Two signals are tested below. The first one is bearish when CAB value is below its 1-year average (12mma) and bullish otherwise. The second one is bullish when CAB value went down in a 6-month period, bullish otherwise.
The next tables show market timing test results based on monthly decisions with the same assumptions and metrics as previously. The benchmark is in italic.
Bearish signal: CAB(m-2) < 12mma:
Since CAGR MAR Ddmax DLmax CAGR MAR Ddmax DLmax 1993 7.12 0.14 50.82 80 8.12 0.29 28.04 50 1956 6.68 0.13 50.82 89 6.41 0.23 28.04 78 1913 5.46 0.06 84.76 299 6.45 0.14 46.13 183
Bearish signal: CAB(m-2) < CAB(m-8), meaning the 6-month momentum is negative:
Since CAGR MAR Ddmax DLmax CAGR MAR Ddmax DLmax 1993 7.12 0.14 50.82 80 7.47 0.29 25.93 23 1956 6.68 0.13 50.82 89 6.00 0.22 26.84 78 1913 5.46 0.06 84.76 299 6.64 0.17 39.04 173
The next chart compares the 2 indicators and the benchmark (S&P 500, noted SPX) since 1993:
Both indicators based on CAB improve the MAR ratio and reduce the drawdown on all intervals. They lag the benchmark regarding CAGR since 1956. They outperform on shorter and longer intervals.
Because it is long, this article had to be split. Here ends the first part. In the second and last part, we will continue with indicators based on unemployment, retail sales and a bulk shipping index.
For the month of November, 6 out of the 7 indicators presented here were bullish. The monthly update will be posted in next week-end after the unemployment rate for November is released. If you don't have the time to calculate all these indicators yourself (all formulas have been disclosed above), it may be an opportunity to start a free trial.
This article is original, but it includes charts and short excerpts from the book Market Timing For Long-Term Investors, which is an 80+ page research report with selected and discarded indicators, the percentage of false signals, the impact of data revisions, robustness tests and the best way to use them together.
It is some of the leading indicators we follow in Quantitative Risk & Value. We combine indicators in an innovative way because we think no indicator is good enough to make “risk on/risk off” decisions. I will publish research results about other indicators and occasional updates in free-access articles. However, timely updates are posted in private.Get started with a two-week free trial and see how QRV can improve your investing decisions.