PDA

View Full Version : Gold



manucastle
10-25-2011, 10:33 AM
Can anyone explain the large move up in the Gold price just now?

Trev

nickola.pazderic
10-25-2011, 11:43 AM
Erik Swartz (http://www.marketanthropology.com/p/about_9790.html) noted today that $SPX and $GLD are developing and inverse relationship not seen since early August.

Despite the fact that I knew this, trade ideas and news flashes (e.g., European ministers will/won't...) have yet to gel into the perfect trader's mind; thus, I only stood by as Gold, NUGT and SLV broke out heavily to the upside on the news that some meeting was canceled-- a sure sign of discord, etc.. Any of the three trades could have served as an excellent hedge to the Robot trade.

I assume the trade wizards caught this, if only as an E-mini short.

Now, as I look again, GDX, GLD, and SLV are running from DR1 to DR2.

Again, I add: the big Indian festival Diwali is tomorrow-- I believe.
And, many traders everywhere are anticipating QE* on both continents, rendering gold as the most certain of hard assets.

grems8544
10-25-2011, 12:07 PM
My own personal belief of the recent correlation of gold to the equity markets is that gold purchases are being leveraged. When equity markets are doing well, leverage on, when equity markets are doing poorly, leverage off. Coupled with the basing/oversold levels of gold, a gradual move back into gold cannot be discounted.

Further, I can see that the 20d trend line on GLD has moved nearly horizontal, from a downward trend, and this indicates a lack of continued selling or the beginning of stealthy buying. In fact, if you look closely, GLD EV actually crossed ABOVE the 20d trend line, which is bullish for gold in general.

Finally, we can see that GLD is making higher lows, so this portends (but we have not realized) higher highs.

Time will tell.

Regards,

pgd

grems8544
10-25-2011, 12:28 PM
When evaluating correlation, time frame for developing the correlation is important.

Your link to Erik Swartz made me want to run a quick test, the results of which are below. Basically, I did a correlation study of the major indexes, IWM and it's related ETFs, GDX and it's related ETFs, and GLD. The ebb and flow of the correlations will be of interest to some.

Bottom line, GDX and GLD are highly correlated within the markets on this present bull leg (since early October). Gold was uncorrelated until mid September. This obviously has implications on the performance of the robot call on GDX.

This data is generated using a an exponentially weighted moving average with a weight factor of 0.8, which gives a full decay in about 30 days or so.

Regards,

pgd

11093

nickola.pazderic
10-25-2011, 01:12 PM
Thanks Paul, very interesting indeed.

The comparison between August and today in GDX/GLD shows, once again, that flux is the order of things. Moroever, we can no longer assume that GLD and GDX are not correlated.

I'm losing about 1.33% of my total risk portfolio on the GDX trade. Pascal gives a big hint that it might be called off by the open tomorrow. I'll trade another day. Steady as she goes...

There is a psychological dimension to this, which I cannot escape. It is this: while taking a robot position, I become myself a robot. I become one because the entries that it asks me to take are usually outside of my imagination, and I believe there is overwhelming and increasingly backtestable evidence to show they trade better than me.

All that said, I think I can improve on the robot in some ways. I learned a lesson from today. Namely: as the market opened in favor of our GDX position and thereafter, I kept an eye on GLD. It stayed about neutral. It did not go positive or negative. When the position began to deteriorate in GDX, a hedge of GLD,SLV or NUGT suggested themselves. I should develop an internal robot that might allow me to hedge against the robot trade when the forces seem aligned against it.

EB
10-25-2011, 01:27 PM
All that said, I think I can improve on the robot in some ways. I learned a lesson from today. Namely: as the market opened in favor of our GDX position and thereafter, I kept an eye on GLD. It stayed about neutral. It did not go positive or negative. When the position began to deteriorate in GDX, a hedge of GLD,SLV or NUGT suggested themselves. I should develop an internal robot that might allow me to hedge against the robot trade when the forces seem aligned against it.

Yes, thanks Paul.

@Nickola: Just keep in mind that by altering the robot in any way, including the addition of hedges, creates an entirely new system that has not been backtested. You might very well outperform based on the addition of systematic or discretionary changes, but any security you feel from the backtested edges will be illusory.

nickola.pazderic
10-25-2011, 02:32 PM
Should I deviate from the robot, I no longer participate in its outcomes. That is, I'm no longer fully robotic.

This is true. But it poses a serious question: can one really be robotic?

Perhaps. Perhaps not and remain a sentient creature.

I would offer a third choice: use the robot as a teaching tool and as a feedback tool.

The problem I encountered today emerged from my recent tendency to "read the tape" as a method by which I understand real time market activity. I've found that all my artful charts do not work as well as simply folloing a price/volume chart with a Christmas tree of lights on the side that display the prices and changes of stocks and commodities and currencies that interest me. E.g.,

11094

They interest me because I perceive corrrelation. Since I watch them all day over several weeks, I develop an intuition about what is happening. Frankly, via small stake trades, I'm learning that my intuition is profitable. Moreover, I want to add that my intuition is a form of reasoning based on pattern recognition. (I regret that, unlike many of the wizards here, I'm unable to translate my intuitions into mathematical language. I am, however, at work on narratives that give my intuitions greater integrity).

What is perhaps superior about my intuitions over those of the robot is that my intutions are evolving real time. I adjust to feedback on a moment by moment basis. My chief draw back remains emotion. Emotion can cloud and inspire thought and action. They are unpredictable. This is why the robot is valuable. While it cannot adapt itself to every short term change in the market (e.g., GLD and GDX's recent correlation), it does provide an opportunity to act without emotion. It sets a standard of objectivity that I now take to be a real time teacher.

The question of whether or not I can improve upon the robot is a quantifiable question. But the degree to which I learn from the robot is, it seems to me, a more crucial question.

roberto.giusto
10-25-2011, 02:41 PM
You might very well outperform based on the addition of systematic or discretionary changes, but any security you feel from the backtested edges will be illusory.

One may feel the need to add discretionary trading to the robot whenever unhappy of its last trade if closed at a loss.

Only to discover that "discretionary fear" caps the robot's gains on its next profitable trade.

I use money management depending on the strenght of the signal, place the new order or update its stops before the market opens each day.

And then I close the robot's account and never open it during the day.

Whenever I have "intraday" doubts, I just watch the robot's equity line and my own: this is enough for me to decide that I won't meddle with it. ;-)

EB
10-25-2011, 02:42 PM
They interest me because I perceive corrrelation. Since I watch them all day over several weeks, I develop an intuition about what is happening. Frankly, via small stake trades, I'm learning that my intuition is profitable. Moreover, I want to add that my intuition is a form of reasoning based on pattern recognition. (I regret that, unlike many of the wizards here, I'm unable to translate my intuitions into mathematical language. I am, however, at work on narratives that give my intuitions greater integrity).

It looks like you're embarking on learning the Wyckoff method of trading. It can be very rewarding as you internalize thousands of nuances and patterns. It does take serious screen time and dedication because the correlations themselves change due to fundamentals, and also as the market transitions from bull to bear and back on various time frames. The activity is particularly relevant as major markets reach inflection points. Most of what happens in between is noise. That's why many who read the tape put up a time & sales on their main trading instrument with a block filter (e.g., only show trades of 10 or more in the ES) and watch it around projected inflection points in time (news) and price (important areas).

nickola.pazderic
10-25-2011, 02:51 PM
EB--

I neglected to note in the previous post that I'm learning very much from everyone here.

I should also note that today, in the first 45 minutes of trading, it appeared that the market tipped its hand, as it were. It was a very exciting moment, as millions/billions sloshed into the precious metals space and treasuries were bid up again. Restrained, I missed out on a very profitable move.

Time and sales with a filter...This is something I will look into.

Many thanks,

EB
10-25-2011, 02:54 PM
Time and sales with a filter...This is something I will look into.


I should add that in a fast market I find it very difficult to read T&S because I'm looking at the change in velocity and trying to see if the big orders are coming in around inflection points. It's been a bit better lately as ATR has contracted (at least in my market, the ES).

manucastle
10-25-2011, 03:19 PM
Yes, thanks Paul.

@Nickola: Just keep in mind that by altering the robot in any way, including the addition of hedges, creates an entirely new system that has not been backtested. You might very well outperform based on the addition of systematic or discretionary changes, but any security you feel from the backtested edges will be illusory.

EB

I take it from your comments that trading ES provides even better returns than mechanically trading the Robots . Your long term returns must be phenomenal. :O)

Trev

EB
10-25-2011, 04:16 PM
EB

I take it from your comments that trading ES provides even better returns than mechanically trading the Robots . Your long term returns must be phenomenal. :O)

Trev
I go through rough patches like every one else, and bear markets have been especially tricky for me. The draw back to discretionary trading is it doesn't necessarily scale. Moving up in size takes significant psychic capital, whereas a robot is only theoretically constrained by liquidity concerns (unlikely to be reached by most). Actually, one of my long term goals is to produce an intraday eMini robot.

manucastle
10-25-2011, 06:04 PM
I go through rough patches like every one else, and bear markets have been especially tricky for me. The draw back to discretionary trading is it doesn't necessarily scale. Moving up in size takes significant psychic capital, whereas a robot is only theoretically constrained by liquidity concerns (unlikely to be reached by most). Actually, one of my long term goals is to produce an intraday eMini robot.

Sounds good and I hope you can keep us posted !

Trev