Trading the sun

[Gringotts Bank]

http://www-personal.umich.edu/~shumway/papers.dir/weather.pdf

Since Melbourne is about to end an insanely hot and unprecedented 11 day sunny streak in 3 days time, this might be worth a read.

 

[Trembling Hand]

LOL GB you ever get the feeling you are going mad?

Will the market returns be different in Chicago compared to LA although its the same market

 

[skc]

I trust/hope you are just taking the pi$$ out of the paper / weather right?

 

[Gringotts Bank]

LOL GB you ever get the feeling you are going mad?

Will the market returns be different in Chicago compared to LA although its the same market

Heat can cause one to go troppo.

Good question, but in the results section (after the references) you'll see that New York was the only US city included. I realize people from all over the world trade on the NYSE, but they were just looking for sun:return correlation, no more than that. A better study would be to look at an averaged world weather sunniness and compare that to the MSCI world index.

 

[Trembling Hand]

Heat can cause one to go troppo.

Good question, but in the results section (after the references) you'll see that New York was the only US city included. I realize people from all over the world trade on the NYSE, but they were just looking for sun:return correlation, no more than that. A better study would be to look at an averaged world weather sunniness and compare that to the MSCI world index.

OMFG you are serious!

 

[Gringotts Bank]

OMFG you are serious!

I've bet everything I have, leveraged x100 that the market will go down after Melbourne cools in 3 days. What's wrong with that?

 

[sinner]

heaps of research on this by respected names, the RBS paper on moon cycles (also tested by marketsci) being the most famous.

http://www.mcoscillator.com/learning_center/weekly_chart/are_traders_really_just_driven_by_the_sun/

by changes in solar activity. I have some trouble accepting it myself. But every once in a while, it can be worthwhile to take a look at a wild idea just to see if there is any merit to it. The evidence here seems to be strong enough to demonstrate that there is something going on there, even if it is not good enough to figure out how to make money off of it.

I recommend googling "mcoscillator sunspots" if you wish to see his other research on the topic.

Also, this

http://mpra.ub.uni-muenchen.de/40271/

EDIT: Moon stuff
http://www.markettiming.nl/img/image/file/RBS_MoonTrading13Jun10.pdf
http://marketsci.wordpress.com/2009/05/12/the-lunar-cycle-and-the-stock-market/
http://www.mta.org/eweb/docs/1998DowAward.pdf

 

[skyQuake]

Correlation does not imply causation

 

[Gringotts Bank]

heaps of research on this by respected names, the RBS paper on moon cycles (also tested by marketsci) being the most famous.

http://www.mcoscillator.com/learning_center/weekly_chart/are_traders_really_just_driven_by_the_sun/


I recommend googling "mcoscillator sunspots" if you wish to see his other research on the topic.

Also, this

http://mpra.ub.uni-muenchen.de/40271/

Yes the sun spot stuff was interesting. My favourite non-traditional study is the one in my signature.

 

[Gringotts Bank]

Correlation does not imply causation

Of course that's true.

Causation would be almost impossible to determine in any method of analysis, wouldn't you say? I think correlation is the best anyone can do.

 

[sinner]

Correlation does not imply causation

I am not sure there is an implication of causality here, only that observed returns around cyclical points are higher/lower than returns starting at the other end of the cycle. If returns were spread out equally across the cycle lambda then wouldn't you say there is no relationship? But since returns are divided between cycle highs and lows there is some relationship.

 

[skc]

Fair enough. I also note that Suncorp (ASX:SUN) has vastly outperformed Cloud Peak Energy Inc (NYSE: CLD)... further supporting the notion that trading the sun is a valid strategy.

 

[sinner]

Fair enough. I also note that Suncorp (ASX:SUN) has vastly outperformed Cloud Peak Energy Inc (NYSE: CLD)... further supporting the notion that trading the sun is a valid strategy.

View attachment 51277

I've never put money on any astrological bet, but I am consistently intriuged by return profiles which differ based on astrological cycles, as if you don't believe in the astrology then it seems pretty inexplicable as to how those return profiles can exist.

It's not like people are doing tests and finding there is no (or a statistically insignificant versus random and control signals) relationship between various cycles and market returns, so what's the basis for your derision?

 

[Trembling Hand]

so what's the basis for your derision?

The more I see people chasing the key to crack the code the more I feel like retreating to my own cave. One where I can go about improving my trading skills and leave the 2 bob hopefuls to fanciful dreams of finding the easy way to one day becoming profitable.

 

[sinner]

The more I see people chasing the key to crack the code the more I feel like retreating to my own cave. One where I can go about improving my trading skills and leave the 2 bob hopefuls to fanciful dreams of finding the easy way to one day becoming profitable.

Shame you see it like that TH, because all I saw was GB sharing a research paper (incidentally one which I haven't read previously) and getting slammed for it.

What could have been an interesting discussion, as usual for this forum when it comes to technicals, ended up a boring fight. At least in this case, equity curves have been provided.

"It is the mark of an educated mind to be able to entertain a thought without accepting it."

 

[Trembling Hand]

"It is the mark of an educated mind to be able to entertain a thought without accepting it."

Problem is though Sinner that there is a plethora of "research papers" on all sorts of edges. Until one goes about actually forward testing they mean nothing, just like every other two bob backtest. Forward test is the traders 'peer review". This one clearly fails the common sense test. What does sunny days in NYC got to do with what the futs traders are doing/thinking in cloudy Chicago?

 

[chops_a_must]

Problem is though Sinner that there is a plethora of "research papers" on all sorts of edges. Until one goes about actually forward testing they mean nothing, just like every other two bob backtest. Forward test is the traders 'peer review". This one clearly fails the common sense test. What does sunny days in NYC got to do with what the futs traders are doing/thinking in cloudy Chicago?

I reckon could it actually make sense.

Sunny weather is associated with less depression and greater wellbeing.

Therefore resulting in more optimism.

 

[Trembling Hand]

I reckon could it actually make sense.

How can the sun shinning in a city where the majority of traders DO NOT live have an effect on them?

 

[sinner]

Problem is though Sinner that there is a plethora of "research papers" on all sorts of edges. Until one goes about actually forward testing they mean nothing, just like every other two bob backtest. Forward test is the traders 'peer review".

No disagreement from me here.

This one clearly fails the common sense test. What does sunny days in NYC got to do with what the futs traders are doing/thinking in cloudy Chicago?

I'm not sure if you bothered to read the paper. The main finding is that the explanatory power of sunshine/cloudiness for returns is relatively low, r^2 = 0.02 ...(to be expected) despite being statistically significant. So it's not like there is huge returns to be had buying the NYSE composite on sunny NY mornings. The authors of the paper are more interested in behavioural aspects of the findings, such as

But we think the main practical implication of our findings is somewhat less direct. Our results suggest that investors can benefit from becoming aware of their moods, in order to avoid mood-based errors in their judgments and trades.

and, having already pointed out the well understood link between mood and sunshine:

Our findings also suggest some broader implications for asset pricing. Sunshine is just one of many influences on mood. In confirming the effect of mood on asset prices, this study suggests that other mood effects may be important. For example, as discussed in Section I, negative moods tend to stimulate effort at careful analysis, whereas positive moods are associated with less critical and more receptive information processing.

 

[Gringotts Bank]

How can the sun shinning in a city where the majority of traders DO NOT live have an effect on them?

The "how" is nice to know, but not essential.

If my grass grows more quickly when BHP is dropping, and I can identify a strong correlation, and I can watch that same pattern play out in real trading, then that's a valid short system. It will make money. I don't need to know how, even if it seems ridiculous to me. The system will continue to perform as long as it does. Maybe at some point in the future, such a "grass growing BHP short sell system" will stop making money.

But when you compare this to something you'd consider "reasonable" like an open range breakout trading system, there's no difference. You don't know why it works (even though you have theories, there's no proof at all). And at some point, it will stop working, and once again, you won't know why or how.