Objective research and reviews to aid investing decisions
 Looking for information on debt consolidation loans? Want to know more about reverse mortgages? |
When the market close is strong (weak), does it indicate pent-up buying (selling) demand? Should a trader follow the trend of the close the next day, position for a reversal or look for a better indicator? To investigate, we compare the position of the daily close for a broad market index relative to same-day high and low to the next-day return for the index. We also compare the five-day and ten-day average relative closes to the index return for the next five and ten days, respectively. Using daily high, low and close levels of the S&P 500 index for the period 7/15/83 (the earliest without obvious errors available) through 8/15/07 (6077 trading days), we find that:
The following figure shows summary statistics for the position of the daily closing level of the S&P 500 index relative to its daily high and low for the entire sample period. On average, the close is 55% above (45% below) the daily low (high). The standard deviation of the relative close is 14%.
How does the value of the relative close relate to the return for the index the next day?
The following scatter plot relates the daily return on the S&P 500 index to the prior-day relative close. There is no obvious relationship between the two series. The Pearson correlation for the entire sample is 0.01 and the R-squared statistic is 0.00, supporting the observation that there is no relationship.
Despite lack of visual evidence, might there be ranges of prior-day relative close that are predictive? Specifically, might very weak closes relate reliably to next-day returns?
The next chart is an ordered version of the prior scatter plot, truncated to focus on those days for which the S&P 500 index closes within 5% of its low for the day (534 instances) and arranged from lowest relative close (0%) to 5%. Again, there is no obvious relationship between relative close and next-day return, even when the index closes roughly at the low for the day. The Pearson correlation for the subsample is -0.01 and the R-squared statistic is still 0.00, supporting the observation that there is no relationship.
Might the average relative close over the past five or ten trading days be predictive for the index return over the next five or ten trading days?
The next scatter plot relates the five-day return on the S&P 500 index to the average relative close over the past five days. For this plot, we use data for every fifth trading day (1225 observations) to ensure independence of sample points. The trend line indicates a small negative relationship between the two series. In other words, the smaller the average relative close over the past five days, the greater the return for the index the next five days. The Pearson correlation for this sample is -0.11. However, the R-squared statistic is only 0.01, suggesting that past relative close explains only 1% of future returns.
Might very weak average relative closes relate to five-day returns?
The next chart is an ordered version of the prior scatter plot, truncated to focus on those five-day intervals for which the average relative close for the S&P 500 index is less than 34% (93 instances) and arranged from lowest average (7%) to 34%. Again, there is no obvious relationship between relative close and next-day return. The Pearson correlation for this subsample is -0.18 and the R-squared statistic is 0.03, indicating a somewhat stronger (but still very weak) negative relationship than for the entire sample.
Might the average relative close over the past ten trading days be predictive for returns over the next ten trading days?
The weak negative relationship between past relative closes and future returns for five-day intervals is even weaker for ten-day intervals. The Pearson correlation between the the ten-day return on the S&P 500 index and the average relative close over the past ten days (606 independent observations) is -0.05 and the R-squared statistic is 0.00. When the average relative close for the S&P 500 index over the past ten days is less than 41% (45 instances), the Pearson correlation is -0.18 and the R-squared statistic is 0.01.
In summary, relatively weak and strong daily closes say nothing about next-day returns. Weak closes over the past several days are very slightly (untradably) predictive of above-average returns over the next few days.
See Blog Synthesis: Some Trading Indicators for analyses of the usefulness of other technical indicators.
