Friday, 2 October 2009

Financials Could Stabilize or Rebound Next Week: Data

After a tough few days in the markets, next week could look up, according to this afternoon's Commitments of Traders data. My trading setup for the S&P 500 remains bullish. And the data for the bellwether three-month Eurodollar, which gives me signals for the BKX U.S. Bank Index, has taken a bullish turn. I've just updated my latest signals table with the data highlights.
- 30-Year U.S. Treasury: My trading setup for the 30-year bond goes to cash on Monday's open after a three-week bearish signal (meaning the setup thought the yield would go up).

- U.S. banks: My setup for the BKX Bank Index, based on the Eurodollar contract, is in cash again this week - its ninth straight week. But some of the data that makes up that setup is signing a cautiously bullish tune. Now, most of the COT data doesn't correlate very strongly with market prices, but I found that some of the Eurodollar numbers do. The large speculator total open interest, which has a 0.61 correlation with the next week's BKX, has rebounded somewhat after last week's major swoon. Also interesting, the small trader total open interest, which has a moderate 0.41 correlation with the next week-BKX, continues to bounce back from deeply depressed levels. The specific numbers are on the latest signals table.

- S&P 500: The data in this market is looking seriously less bullish, as the commercial hedgers sell down their super-bullish positioning, while the wrong-way small traders get suddenly braver about the rally. The latter are, in fact, this week within a hair of triggering a bearish signal for their trader group, which would have put my setup in the cash column with a three-week delay. But until the setup actually changes its signal, which it hasn't, my current long trade continues. It's sometimes in these end-game periods that the real blow-off gains are made. Plus, the COT data for the S&P 500 correlates very poorly with market prices, so I'm not too worried about this data until an actual new signal.

- Natural gas: My setup goes bearish on natural gas on Monday's open, but just for one week. The setup returns to cash the following week.

Have a great weekend, and see you early next week with a portfolio update.

19 comments:

Anonymous said...

Alex, unless I'm reading this incorrectly, I think I've found a few errors-

In this weeks $BKX you have negative values for all three components (all on follow) yet kept Large Spec Net % bullish. Shouldn't it be bearish like the others?

Also, the S&P 500 small trader is 0.48 this week. Isn't this bullish? Fading it would make it a bearish signal.

All in all, it looks like everything got more bearish from the last report.

In Debt We Trust said...

Watch european financials in particular. There has been a lot of talk about regulatory capital - to -debt ratios being adjusted.

Just recently, the European banks also passed their regulators' stress tests (come on, did you think they would actually be allowed to fail?).

There is also talk about the World Bank swooping in to buy the distressed assets of European banks that made NPLs to E. Europe.

Anonymous said...

COT doesn't work. Couldn't catch this decline which will turn out huge.

x

Anonymous said...

You can't swing trade off COT data. It lags way too far behind. It's only good for long term setups.

You'll miss a significant and profitable part of this downturn , if not all of it. Too much moneyto be left on the table .

Jaywiz

Alex Roslin said...

Hi guys,

You may be interested to know that properly tested, the data actually gives excellent results in backtesting. See my backtesting results table for the details. The signals don't tend to lag. You might be thinking of the popular notions of how the data works that aren't based on any rigorous backtesting. In fact, the signals tend to be early - hence my trade delays for many setups.

Regards,
Alex

Alex Roslin said...

Hi Anonymous,

Thanks for your message. The table is correct. The large spec net position didn't fall enough to go bearish. It must hit the signal line. (See the value in the Signal Line column.)

In the S&P 500, the small traders getting more bullish is actually a bearish sign. But they didn't get bearish enough to hit the signal line and reverse their currently bullish signal. Also, as I explained in my post, the S&P 500 COT data has very little correlation with next-week market prices, so I don't read much into the week-to-week fluctuations in the S&P 500 data. The Eurodollar COT data, on the other hand, has stronger correlations, as explained in my post.

Regards,
Alex

Anonymous said...

Hi Alex, thanks for your response. I missed the "line" required to turn the signal- was only looking at positive/negative values.

Thanks for all your hard work.

keiz said...

Hi Alex, was curious what price series you use for the 30-yr bond? Do you use the futures price, bond ETF, or some other time series? Thnx

Alex Roslin said...

Hi Keiz,

It was the cash price as given by the CRB (symbol US).

Regards,
Alex

Roger said...

Your setup was for a bearish trade on natural gas beginning Oct 6 but your portfolio does not show a short position. What changed?

Thanks for your reply.

carl said...

Hi Alex,
Why is it that gold net% large spec last week was 1.64, thats under the treshhold of 1,9 but still gave a signal?
Greetings Appi

Alex Roslin said...

Hi Appi,

That signal had already turned bearish, so it didn't matter that it crossed the signal line again.

Regards,
Alex

Alex Roslin said...

Hi Roger,

Just an oversight when I did that post. I just amended it to add that short position.

Regards,
Alex

Anonymous said...

Hi Alex,

It looks like Gold is on a run, beating the bearish signal. Is your stop for the 2X leverage 5.32% = 10.6% for your entry price for HGD.TO? It looks that it may hit this today or tomorrow...

Thx.
Pete

Alex Roslin said...

Hi Pete,

Yeah, these things happen. I'm actually using HBD. (For those who don't know, HBD is based on gold bullion, as opposed to HGD, which is based on gold stocks. I find the trade tends to work best when it is as close as possible to the underlying market that was backtested.)

My entry was $11.75, so my stop is $10.50. If the stop is triggered, I usually look for a trade in the opposite direction, the logic being that the move is historically unusual against the data, so it could keep going.

Regards,
Alex

carl said...

Hi Alex,
Thanx for quick reply. I understood that large spec %net =cash if between 1.9 and -1. Or is it that it gives a signal when it crosses one of the 2 treshholds and the signal for this specific trader group is then BUY or SELL?
Appi

Alex Roslin said...

Hi Carl,

Thanks for your message. The large spec net position is either long or short. It doesn't go to cash. Its signal switches when the opposing signal line is crossed. The overall setup is, however, be in cash when both signals that make it up are not in agreement. Hope that explains things better. I just clarified this on my "How It Works" page in case others are confused.

Regards,
Alex

Anonymous said...

Hi Alex,

Thanks for the clarification with HGD.TO. So if your stops kicks in for HBD.TO, then your getting HBU.TO as the opposite move?

Rgds.
Pete

Alex Roslin said...

Hi Pete,

Since it's a discretionary trade, I'd consider anything in the precious metals area that was breaking out technically. In fact, however, I just did go long HBU as it broke out above weekly and daily TDST resistance lines.

Regards,
Alex