Lunar cycle and market moves

[Logique]

I'm not advocating the theory or otherwise, it's just information.

The following is from The Australian - Business, on 23 Dec 2009, link: http://www.theaustralian.com.au/bus...e-lunatic-fringe/story-e6frg91o-1225812989219

"Now, backed with decades of data and behaviour that can no longer be explained by purely rational analysis, the lunar theory has slipped into the mainstream.

In a piece of research that involved 14 of its senior analysts from across five leading financial centres scrutinising data from 32 leading indices over several decades, Macquarie Securities has arrived at a startling discovery: the two days on either side of the new lunar month represent most of the positive returns on equity markets for the next four weeks".

"Using data since 1988 for a wide variety of indices," the report concluded, "it is quite clear that a strong surge in returns can be seen leading into the turn of the (lunar) month."

Next New Moon: 15 Jan 2010.
Next Full Moon: 31 Dec 2009

 

[SmellyTerror]

Ah "many", my favourite weasel word. How many is many, exactly? Are we talking many-number, or many-proportion?

And "higher than average" is always good for a run. About half of most ranges are higher than average. Of all the days of every month, about half are slightly above average in temperature, too. Welcome to stats.

"Of the 32 markets we examined, all showed higher than average returns around the turn of the (lunar) month ... and for many of the markets, the average return for the rest of the month was below, or close to, zero."

...so they'll quote the exact number of markets (32) but not the number with the dramatic result (just "many").

"In many markets, a very clear increase in average returns can be seen leading into the lunar new month."

So how much is that increase?

The results, the analysis said, showed that without exception every market showed a very slight increase in the average return over the new moon period.

Yeah, surprise me. And, way to go news media, this "very slight increase" was earlier described as a "strong surge".

We know there are folk who trade on silly stuff like lunar cycles. Seems to me that this is plenty to explain a "very slight" effect on averages. That, I can believe. Claims of a statistically significant "surge" would need some real numbers to back them up.

 

[Sdajii]

Ah "many", my favourite weasel word. How many is many, exactly? Are we talking many-number, or many-proportion?

And "higher than average" is always good for a run. About half of most ranges are higher than average. Of all the days of every month, about half are slightly above average in temperature, too. Welcome to stats.



...so they'll quote the exact number of markets (32) but not the number with the dramatic result (just "many").



So how much is that increase?



Yeah, surprise me. And, way to go news media, this "very slight increase" was earlier described as a "strong surge".

We know there are folk who trade on silly stuff like lunar cycles. Seems to me that this is plenty to explain a "very slight" effect on averages. That, I can believe. Claims of a statistically significant "surge" would need some real numbers to back them up.

99% of people who claim that statistics can be used to mislead people have either murdered, tortured, dismembered, severely harmed or mildly harmed their very own parents.

Do you honestly want to be a part of that group? Shame on you.

 

[nulla nulla]

99% of people who claim that statistics can be used to mislead people have either murdered, tortured, dismembered, severely harmed or mildly harmed their very own parents.

Do you honestly want to be a part of that group? Shame on you.

"mildly harmed their very own parents" would just about cover everyone that put their mother through a nine month gestation period, give or take a month for the early or late arivals. Then of course there would be the torture a parent goes through trying to deal with teenagers.

 

[Synergy]

The funny thing about markets is that if enough people start believing things like this they will actually become true.

 

[Putty7]

99% of people who claim that statistics can be used to mislead people have either murdered, tortured, dismembered, severely harmed or mildly harmed their very own parents.

Do you honestly want to be a part of that group? Shame on you.

So when Jim Garrison disproved the magic bullet theory used by the Warren Commission in the murder of JFK that was proven by statistsics and tests, he had either murdered, tortured, dismembered, severly harmed or mildly harmed his own parents.

 

[Trader Paul]

I'm not advocating the theory or otherwise, it's just information.

The following is from The Australian - Business, on 23 Dec 2009, link: http://www.theaustralian.com.au/bus...e-lunatic-fringe/story-e6frg91o-1225812989219

"Now, backed with decades of data and behaviour that can no longer be explained by purely rational analysis, the lunar theory has slipped into the mainstream.

In a piece of research that involved 14 of its senior analysts from across five leading financial centres scrutinising data from 32 leading indices over several decades, Macquarie Securities has arrived at a startling discovery: the two days on either side of the new lunar month represent most of the positive returns on equity markets for the next four weeks".

"Using data since 1988 for a wide variety of indices," the report concluded, "it is quite clear that a strong surge in returns can be seen leading into the turn of the (lunar) month."

Next New Moon: 15 Jan 2010.
Next Full Moon: 31 Dec 2009

=====



Hi Logique,

Lunar cycles sometimes trigger other planetary aspects, that may mark
a trend reversal or a significant breakout in particular markets, as well ...

If the new moon is deemed positive, then the full moon may be seen as
POTENTIALLY negative ..... marking the end of a positive move, started
around the new moon, for example.

Sometimes, such negativity is shortlived, as the lunar cycle moves quickly
and any weakness at this time, may be seen as a buying opportunity in
some stocks.

For example, some stocks to watch for weakness this week, going into the
full moon may be:- GER, TPI, TGR, BYR, GWR, III, IRI, CIR ... to name a few.

More later ... with some stocks to watch, at the next new moon, too ...

happy trading

paul



=====

 

[Garpal Gumnut]

To paraphrase the American Declaration of Independence.

The Lunar effect on all creatures on earth, and its effect on behaviour....

We hold these truths to be self-evident

Chartists have been aware of this for centuries and Paul's comments should be well heeded by the funnymentalists on this forum.

gg

 

[GoodCall]

=====


If the new moon is deemed positive, then the full moon may be seen as
POTENTIALLY negative ..... marking the end of a positive move, started
around the new moon, for example.

What deems a new moon positive?

 

[Wysiwyg]

What deems a new moon positive?

May, could, if the market moves up.

 

[MRC & Co]

What deems a new moon positive?

Good call.

 

[ThingyMajiggy]

What deems a new moon positive?

Well...I'd say if they forgot to wipe, its a negative new moon. So therefore...

 

[howardbandy]

Greetings all --

In my book, "Quantitative Trading Systems", pages 159 through 164, I explore the phase of the moon as a market indicator and give the AmiBroker code you can use to test the relationship yourself.

Assume the new moon occurs at 0 and again at 100 on a clock that runs from 0 to 100. That means a full moon will occur at 50, and so forth.

I used 1/1/1995 through 1/1/2005 as the in-sample period, and found there is a reasonably strong seasonality over the in-sample period. The system buys the market (each of 513 active securities) about the time of the moon's third quarter, reverses to short at about full moon, reverses back to long at the next third quarter, and so forth -- two trades each lunar month. In terms of the 0 to 100 clock, the system is long when the moon is anywhere in the 79 to 59 range (wrapping around zero), and short when the moon is in the 60 to 78 range. The median annual gain was about 20%.

As with all trading systems, the proof is in the out-of-sample results. Using 1/1/2005 through 1/1/2007 as out-of-sample, the relationship completely fell apart. About 80 of the 100 time slots reversed. Keeping the same system as was discovered using the in-sample data, the out-of-sample results were a median annual gain of about -11% (a loss of 11% per year).

As always, do your own research.

Thanks for listening,
Howard

 

[Garpal Gumnut]

Lunar cycles are an obvious trading tool.

Women are lunar, they vary in their temperament and reaction to males on a daily basis, predictable though on a monthly chart.

The trade in DNA is the oldest market going, so I fail to see how anyone can argue that it doesn't apply to stocks.

That is why I am a long term trader and eschew minute or thirty second entires, now that I am more mature in my data collection.

gg

 

[Sdajii]

"mildly harmed their very own parents" would just about cover everyone that put their mother through a nine month gestation period, give or take a month for the early or late arivals. Then of course there would be the torture a parent goes through trying to deal with teenagers.

Yeah, which is what made my statement true... although analysing it any further than the point where you see the joke kills the humour.

 

[Logique]

Healthy scepticism is fine by me.
Law enforcement authorities and hospital casualty departments could tell a story on the effects of a full moon on human psychology. So it's not such a big stretch to believe that it could affect the psychology of market players.

BTW New Years Eve is a blue moon this year (second full moon in a calendar month). Won't happen again on a New Years Eve until 2028. Take care out there folks.

 

[SmellyTerror]

Law enforcement authorities and hospital casualty departments could tell a story on the effects of a full moon on human psychology.

It's a myth.

Quite apart from my own experience (I'm in emergency management) which satisfies me that there is no correlation, there are plenty of debunkings all over the net. For example, near the top of a google search for "lunar effect": http://www.skepdic.com/fullmoon.html - note especially the references at the end.

 

[Timmy]

It's a myth.

Quite apart from my own experience (I'm in emergency management) which satisfies me that there is no correlation, there are plenty of debunkings all over the net. For example, near the top of a google search for "lunar effect": http://www.skepdic.com/fullmoon.html - note especially the references at the end.

Nice reference thanks SmellyT.

 

[Whiskers]

As with all trading systems, the proof is in the out-of-sample results. Using 1/1/2005 through 1/1/2007 as out-of-sample, the relationship completely fell apart. About 80 of the 100 time slots reversed. Keeping the same system as was discovered using the in-sample data, the out-of-sample results were a median annual gain of about -11% (a loss of 11% per year).

As always, do your own research.

Thanks for listening,
Howard

Hi Howard, I have not read your book, but based on your post your in-sample period was 10 yrs and your out-of-sample was only 2yrs.

I'd be interested to see the results of an out-of-sample period equal to your in-sample period. I think the longer period will always give more consistant/reliable results... and also I think cycle analysis should keep in mind the old cliche, compare apples with apples. In that sense I believe the time frame the test was done in is of significant importance arguably more important than the 'micro' lunar cycle.

To clarify, given there is a lunar cycle that even the scientific community recognises has an effect on plants and other animals... what about other cycles that interact with the lunar cycle. There are sun cycles (solar flares and their effect on weather, well documented by the late Indigo Jones), other planets and most importantly I refer particularly to cycles of human behaviour based on our own individuality, eg the time we were born and how all these other cycles had suttle influences in framing our genetic makeup.

I'm trying to formulate a program to do this, however so far I've not found a way to program interpretations that need to be made and weighed against the various cycles and influences.

The nearest I've heard to describe such a program is our 'intuition'... but due to our human frailties we often don't even recognise, let alone take notice of our intuition.

 

[jbocker]

What?
Pure Lunacy.
Only a true lunatic would believe that there is a correlation.
(Oh Please dont sic the werewolves on me!)