06.04.2023
How did the S&P 500 perform 63 days from its 37th Golden Cross? Read on for the answer.
Part 1: published 01.28.2023.
Part 2: published 05.08.2023.
PART 1: The S&P 500
And Its Provocative Golden Cross
The following research explores what history can teach us about the S&P 500 following a Golden Cross.
THE BIG PICTURE
In the parlance of technical analysis, a Golden Cross is when the 50-day moving average of price crosses above the 200-day moving average of price. As we near the end of January 2023, these Golden Crosses have been a hot topic of conversation. Golden Crosses have happened, or are close to happening, on many charts of individual stocks, ETFs, ratio charts, and indices. Technical analysis folklore generally describes this phenomenon as a bullish indication. Historical testing reveals a Golden Cross precedes bullish price action on some asset classes, ETFs, and individual companies. The results on other charts are no better than a coin-toss. In some cases, a Golden Cross is actually better interpreted as a sell-signal. This dispersion of outcomes is why we need to understand the technical signals symbol-by-symbol.
The chart below is the S&P 500 Index with its 50-day and 200-day simple moving averages. Price has been removed for clarity. After Friday's close, 01/27/2023, the 50-day moving average is just under a half percent, .37%, below its 200-day moving average. This is illustrated by my Golden Cross Oscillator in the bottom pane of the chart. The vertical golden lines flag historical Golden Crosses. To better prepare ourselves for what might happen following the next Golden Cross, the following research examines the statistics of historical returns and drawdowns following them.
S&P 500 Daily. Click to enlarge.
ABOUT THE DATA
The following statistics have been calculated in Microsoft Excel using S&P 500 data courtesy of Optuma. Like most data sets, this one is fraught with uncertainties. The history of the S&P 500 is a bit cloudy. We know Henry Poor formed Poor’s Publishing in the 1860s. Online sources explain, he published both weekly and daily indices. Some sources mention 90 components while others mention 233. The consensus seems to be something to the effect of: Henry Poor started publishing his Composite Index around 1923. It then changed its name to Standard & Poor’s Composite Index sometime around 1941. This is when Poor’s Publishing merged with Standard Statistics. 1957 is likely when the S&P Composite became what we know it as today; A market-capitalization index of large companies with roughly 500 components.
Optuma’s S&P 500 Composite data set begins in January of 1950 with closing prices only. It then adds the high, low, and opening prices in April of 1982. Going back to the start of Optuma’s data set, 1950, the S&P 500 has experienced 36 Golden Crosses, averaging 5 per decade. The chart below shows all of them.
S&P 500 Daily From 1950 w/All Golden Crosses. Click to enlarge.
THE RESEARCH
First, we consider the 21-day period following historical Golden Crosses. We will buy the index at the following day’s open, hold for 21 days, and sell on the following day’s open which is day 22. Below is the distribution of returns and some summary statistics for the prior 36 Golden Crosses. 3 The distribution chart shows us how many times the signal has resulted in a gain or loss of a certain magnitude. For example, the left most bar shows us price has returned between -6% to -7% 1 time, while the bar to its immediate right shows us price has returned between -5% and -6% 2 times
Distribution of Returns and Summary Statistics. 21-day hold. Click to enlarge.
Following the 21-day holding period, the statistics show:
There is a 61% chance of price closing higher and a 39% chance of price closing lower.
The average gain was 1.4%.
The best return was 9.8% following the 1982 signal.
The worst return was a loss of 6.6% following the 1960 signal.
M.A.E. stands for Maximum Adverse Excursion. This statistic helps tell the story of how price traveled during the holding period. The MAE tells us how far price traveled below our entry price over the life of our holding period.
The average MAE over 21 days is a very easy 2.2%. This number however is only an average. Very rarely do we actually see the average number.
The maximum MAE followed the 2015 signal. That resulted in a 10.4% drawdown before ultimately closing with a loss of 5.78%.
31 of the 36 signals, 86%, gave us a drawdown of less than 5%. Next, we consider a holding period of 42-days.
Distribution of Returns and Summary Statistics. 42-day hold. Click to enlarge.
There is a 75% chance of price closing higher and a 25% chance of price closing lower after holding for 42-days. These odds are much better than the 21-day hold.
The average gain of 2.4% is a whole percentage point higher than the 21-day hold.
The best return was a gain of 14.5% in the middle of 2009 off of the bear-market low.
The worst return was a loss of 12.9% in 1969.
The average MAE is a drawdown of 2.8%.
The maximum MAE was a loss of 13.34% following the signal of 1969, which was also the worst return resulting in a loss of 12.91%.
29 of the 36 signals, 80%, gave us a drawdown of less than 5%.
Finally, we consider a holding period of 63-days.
Distribution of Returns and Summary Statistics. 63-day hold. Click to enlarge.
Holding for 63-days following a Golden Cross increases our odds of a positive return to the largest of the three holding periods. The returns is positive 78% of the time. That leaves a 22% chance of price closing lower after 63-days.
The average gain increases to 4.2%, larger than the 21 and 42-day holding periods.
The largest gain was an impressive 19.7% in just 63-days. This followed the mid-2009 signal off of the bear market low.
The worst return was a loss of 12.1% following the 1990 signal.
The average MAE is a drawdown of 3.6%.
The maximum MAE was a loss of 13.7% following the 1990 signal, which was also the worst return resulting in a loss of 12.1%. Also noteworthy was the 13.3% MAE following the late 1969 signal. That resulted in a loss of 8.11%
27 of the 36 signals, 75%, gave us a drawdown of less than 5%.
IMPORTANT TAKEAWAYS
Controlling risk is job number one, so the final chart gives us a closer look at the risks associated with getting long the S&P 500 on the day following a Golden Cross and holding for the next 63-days.
Distribution of MAEs and Summary Statistics. 63-day hold. Click to enlarge.
Looking at the 63-day distribution of returns, several charts above, and looking at the 63-day distribution of MAEs, directly above, we learned the following:
The average gain after a holding period of 63-days was 4.2%, and the signal was positive roughly 78% of the time.
The average MAE is 3.56%. Impressively, 27 of the 36 Golden Crosses, so 75% of them, had drawdowns of less than 5% below of our entry price. 23 of the 36 signals, so almost 64%, had a drawdown less than the 3.56% average MAE.
While these stats are impressive, the optimism is worth tempering by noting that 9 of the 36 signals, 25%, had more significant drawdowns. Those include 4 drawdowns ranging from losing 6 10% to 14%. Also, at the end of the 63-day holding period following the 1990 signal, the 63-day hold resulted in a loss of more than 12%.
Hopefully, understanding the historical statistics of Golden Crosses on the S&P 500 will help anchor our expectations for when the next Golden Cross triggers on the S&P 500. Please reach out with any and all feedback and comments. I would love to know your thoughts.
PART 2: The S&P 500
And Its Provocative Golden Cross - UPDATE
4 trading days after the original memo was published, the S&P 500 gave us our 37th Golden Cross in 73 years. How did the market perform?
THE EXPECTATIONS
Our research determined, 63-days after the S&P 500’s Golden Cross, price closed higher 78% of the time or 28 of 36 occurrences. The average return was 4.2%. The average drawdown from our entry price was - 3.6%
Chart 1: S&P 500 Daily. Click to enlarge.
THE RESULTS
Chart 2: S&P 500 Daily & The 37th Golden Cross Through The 63-Day Hold.
Click To Enlarge.
Our 37th signal was not great. In fact, its performance was the 30th worst of all 37 signals. After 63-days the S&P 500 closed down 1.17%, well below our average gain of 4.2%. The drawdown was almost 8%, also worse than our average of 3.6%.
The next chart shows the average of the prior 36 Golden Crosses in black, and the 63-day performance of the 37th in blue.
Results of the 37th Golden Cross & The Average Performance. Click to enlarge.
IMPORTANT TAKEAWAYS
The latest Golden Cross signal generated a return that was less than average with a drawdown that was larger than average. The takeaway is a reminder. The market does not care about our expectations, researched conclusions, or opinions. We must take our signals time and time again, and we must control our risk. Controlling our risk affords us the ability to keep taking those signals. We are looking to capture the large and infrequent gains found underneath the right-tails of return distributions. In order to capture these large and infrequent returns, we must accept that we will have losses, and specifically more losses, than gains. However, the statistics show the probabilities are on our side even if many times we lose money. Perhaps 63-days after the S&P 500’s next Golden Cross, price will have gained 14%, similar to the 1982 signal, or almost 20%, similar to the 2009 signal. Keep taking those signals.
As always, thank you for reading. I’d love to hear from you. E-mail me: Louis@eastcoastcharts.com
This article is for educational and informational purposes only. The author may or may not have a position in the securities mentioned. Read our full disclaimer here.
Comments