U.S. Stocks and All-Time Highs
Earlier this week, I noted that the U.S. stock market – represented by the benchmark S&P 500 Index – was just a shade off its all-time high price. I was excited by this as research shows that when stocks make all-time highs (which are typically called as such when they are at least 5-year highs), they are more likely to rise further over the following days than they are to fall, and to rise on average by more than they fall. This can be used to create a profitable trading edge.
A 5-year high just means the highest price which has been reached over the past 5 years.
The price of the S&P 500 Index did go on to make that new high, just under the 2950 level. So, it was a development in the market which was worth paying attention to.
I mentioned in my previous blog post that when the Index closed at a 5-year high over the last 45 years or so, it tended to rise over the following week and went up over this period in about 54% of cases. I ran that back test nearly three years ago, so I thought it was worth updating it and tweaking it. I also wanted to add volatility into the mix, to see how useful it can be as a trade filter when going long of the S&P 500 Index.
In this back test, the test is conducted from 1976 until approximately one week ago, which covers the last 42 years. Transaction costs such as spreads and commissions are not included in the results outlined below.
If the Index closes at a 5-year high, and is bought and held for one week, it went up on 56.23% of trades, and on average rose by 0.04% over the week. If held for one month, it went up on 61.28% of trades, and on average rose by 0.43% over the month.
What if we add a volatility filter? Let’s say that we only took the trades where the entry day’s range was greater than the average range of the previous 15 days, using the Average True Range (ATR) indicator as a measuring tool. This improves the overall results slightly, raising the weekly win percentage to 57.99%, and the average win to 0.10% over a week. The monthly hold result was just a little worse, with a win percentage of 61.25% and average monthly rise of 0.40%. So, it seems that above-average volatility improves the shorter-term result, but also causes a deterioration in the longer-term result.
In one of my forthcoming posts, I will outline how the back test applied to 50-day highs compares to these 5-year highs.