Friday 13 November 2009

No End for Market Grind

The market sausage grinder kept churning this week. Hope you kept yourself intact. Unfortunately, the Commitments of Traders report this Friday doesn't really help to clarify which way things will break from this long trading range we're stuck in. I've just updated my latest signals table with some new numbers based on that COT report. Two new signals: bearish for natural gas with a week's delay (for the open of trading on Nov. 23) and cash for my 30-year Treasury bond setup on Monday's open (Nov. 16). Some other highlights:
- S&P 500: Not much to report here. This trading setup is in its third week of being bearish and shows no signs of changing course. The wrong-way small traders are still highly bullish in their net positioning as a percentage of the total open interest, while the "smart money" commercial hedgers are still seriously gothic.

- U.S. banks: My setup for the BKX U.S. Bank Index enters its second week of being bullish. The total open interest for the large speculators and small traders - both of which correlate well with BKX prices the following week - is down somewhat this week. But the change is small enough it's hard to know what to read into it. I haven't studied the correlation data enough to know how trade-worthy small or large changes week-to-week changes are. (It's on my massive to-do list!) What I do know is the trading setup - which does have robust backtested results - remains bullish, so that's the real key for me.

- Gold: My trading setup for gold goes into its third week of being bearish. As I mentioned in a post Wednesday, I got stopped out of this position that morning and was looking for a possible chance to put on a discretionary long trade. I didn't see anything worth jumping at. Gold stocks and silver aren't making new highs like bullion is, which is probably a cautionary sign. The gold data this week is also giving mixed signals. The large spec net position, which has a 62-percent correlation with gold prices, has dropped a fair bit, as you can see on my latest signals table. Meanwhile, the large spec total open interest, which has a 77-percent correlation with next week's gold price, has shot up. What does it mean? A good chance of more trading grind next week.

Have a good weekend, and be sure to check back in early next week for an update of my portfolio page. Apologies for missing my update this week. Bad cold. Oink. Anybody got some truffles?


7 comments:

In Debt We Trust said...

Btw, Barron's is pumping Exxon Mobil and Fedex in weekend article.

So, maybe those sectors will rise in the next week as a rush of retail (small money) traders try to plug the gap between valuation and advertising.

More importantly, as a journalist, what do you think of such front page stories? Isn't it a bit disingenuous of them to come out w/articles like that when their financial relationship is probably murky (e.g. essentially paid advertisement)?

Alex Roslin said...

Hi In Debt...,

I can't comment directly on this particular front page of Barron's, but I do know of traders who fade whatever Barron's recommends. As well - not speaking specifically about Barron's - in general, financial journalists can definitely have questionable conflict issues regarding stories.

I used to read Barron's pretty religiously, but I find it pretty useless for my personal trading at this point. At best, there's a small amount of entertainment value in a tiny fraction of their stories.

Regards,
Alex

jesuscheung said...

hi COTs Timer, your recent signals are working against you- bearish SP, bullish nekki. Maybe the trade delay is larger than expected?

Alex Roslin said...

Hi Jesuscheung,

Please see my FAQs page for discussion of this kind of situation.

Regards,
Alex

jesuscheung said...

hi Alex, i am sure your system works in the long run!
Have you think about maybe improving your system by taking into account this year's data? As I understand your system is based on data before 2007.
great blog!

Alex Roslin said...

Hi Jesuscheung,

Yes, I plan to start updating the setups in coming months with the new data. There was just under 13 years of data originally for most markets. That should provide good signals for one-eighth to one-quarter of that time going forward - 19 to 38 months. If the system is designed well, the current setups should still be among the top ones when they're retested on the new data.

Regards,
Alex

Alex Roslin said...

Just a small correction - the current setups were tested on data up to the end of 2007.

Regards,
Alex