December 30, 2013 |

 My apologies in advance for a seemingly strange piece of research.

Recently a Speclister posted a link to a site which inferred considerable success in trading various markets on the basis of solar and lunar events. We have all seen these for decades. There are lots of charts that seemingly draw the connection between full and new moons, sunspots, geomagnetic radiation and of course the financial markets. I myself found nothing in the way of serious data that would make me want to trade on that basis, but the site exuded so much confidence that it was hard to dismiss out of hand.

The site like many in the genre spends a lot of space arguing WHY. You know, humans are mostly water and Earth's tides are controlled by solar and lunar gravitation, so why not humans. Personally I don't care what the reason is, as long as a reason exists and the data is non-random. In this case I am going to assume that a reason exists, but is not discernible. So the answer was for me to take a look at the data with our research tools.

My period of study was from January 1, 2005 through December 27, 2013. That could always be enlarged if some worthwhile results were forthcoming. As a benchmark equity asset I used SPY, as it included dividend yield and was a real and tradable market.

Over the period SPY achieved a 7.4 percent compound annual rate of return (CAROR) while experiencing a 60.83 percent maximum drawdown (DD). Thus the return to risk ratio (R/R) was 0.12. Full statistics and a chart are here.

The site made some strong claims about the value of the full and new moon dates, so my first look was there. To look at solar influences I would need a significant number of cycles and they are approximately 11 years each. First I bracketed the half-month on either side of the full moon, and the same with regards to the new moon. With regards to the full moon, you would buy SPY at the first quarter and hold for the half-month through the full moon, selling at the third quarter. When you were out of the market you were in cash, earning nothing. Thus the following constitute programs in which you are only invested for half the possible time:

Full Moon Bracketing:           2.1% CAROR,    36% DD,     0.06 R/R
New Moon Bracketing:        5.19% CAROR,    47.98% DD,     0.11 R/R

This agreed with the site in that longs would favor the new moon. But if the full and new events corresponded to troughs and peaks, we had to look at equity growth between the events. This also constituted investing for only half the possible time.

New to Full (waxing):        9.82% CAROR,    46.08% DD,     0.21 R/R
Full to New (waning):        -2.2% CAROR,    41.17% DD,     -0.05 R/R

These results would suggest that equity prices tend to trough at the full moon and peak at the new moon, exactly as conveyed by the website.

Links to stats: 






Steve Ellison writes:

To what does the t score of 3.46 refer, and how significant is it given multiple comparisons (you tested 4 subsets of data, and one looked pretty good)?


WordPress database error: [Table './dailyspeculations_com_@002d_dailywordpress/wp_comments' is marked as crashed and last (automatic?) repair failed]
SELECT * FROM wp_comments WHERE comment_post_ID = '8947' AND comment_approved = '1' ORDER BY comment_date




Speak your mind


Resources & Links