Saturday, 15 August 2009

Going Long Treasury Bond and Nikkei, Cautious About Banks

The markets seem to have shrugged off that little selloff, but this week's Commitments of Traders data suggests we may not be completely out of the woods. The numbers are giving some mixed signals, with a bullish trade starting Monday for the Nikkei and positive data for the end of the month in the S&P 500 but bearishness signalled for banks. Also potentially worrisome is a bullish signal Monday for the 30-Year Treasury Bond (meaning the bond yield would fall). Bond prices have a moderately negative long-term correlation with equities and commodities (though that kind of relationship often means less in shorter-time horizons, including for the coming week). I've just updated my latest signals table with the data from Friday's COT report for the markets I'm trading with my COTs Timer system. Some highlights:
- S&P 500: The data in S&P 500 futures and options is looking up. My trading setup for this market goes bullish on the open of Monday, August 24, and the numbers remain hopeful into mid-September. The "smart money" commercial hedgers are super-bullish in their net positioning, while the wrong-way small traders are increasingly negative, as you can see from the data on my latest signals table.

- U.S. banks: More trouble for U.S. banks, if the COT data for the three-month Eurodollar contract is any indication. (That's the interest rate, not the currency.) All three of the groups of traders that make up my setup for the BKX U.S. Bank Index are now in the bearish column. Because of the varying trade delays for the three signals, there's no bearish trade just yet in that setup, which has been in cash since Aug. 10. But it's still not pretty.

- Natural gas: This ridiculously volatile commodity has sold off massively since my trading setup went to cash in early August, making new lows this week. My setup suggests this might not be the end to the correction in natural gas. It's given me a bearish signal with a one-week trade delay - meaning execution on the open of trading on Monday, August 24.

I've also just updated my portfolio page with my bond and Nikkei entries and the current result from my open long crude position. Good luck next week!

19 comments:

Padraic said...

Hi Alex,

I’m glad that someone was able to explain more clearly the query I had about the S&P small traders signal level. I’ll be watching you carefully from now on.
:-)

On the current setup, I’m not so sure that I’d describe the commercial hedgers as ‘super-bullish’. They still have a net short position of 6.1%. Of course this is very bullish relative to their recent positions and I suppose that’s what the system is all about.

Regards and thanks for all your work.

Padraic.

In Debt We Trust said...

Do you think the equity index is positive b/c of opex?

BanfOfAnalysts said...

Banks have had a tremendous rally, which is not really supported by fundamentals but rather by funds flows. Everyone sees Paulson buying banks and feels the need to increase allocation, so banks start running and it becomes a self-fulfilling prophecy. Paulson may be done buying for now though, and many of the large funds are likely full in terms of sector allocation, so even though there's still potential upside it's more limited. On the other hand, fundamentals are looking worse and worse for the back half of 2009. Doug Kass recently expressed bearish views after perfectly timing the upturn, and reports are showing record foreclosure filings that will depress home prices later in 2009. At the same time, you will have the government pulling out of its support of the financial system in October 2009, which could turn out to be very poor timing. Finally, the CDS market has started turning up for banks, but it usually precedes equity markets for some time. All in all, even if there is some more room in financials, it seems to be a good time to lock in some profits and perhaps even start positioning short.

Joseph said...

Hi, are you still long OIL, wondering what your thoughts are.

Alex Roslin said...

Hi In Debt...,

I haven't studied effects of options expiry on the data, so I can't say.

Regards,
Alex

Alex Roslin said...

Hi Joseph,

As you'll see on the latest signals and portfolio pages, yes I am still long crude oil. I will remain so until the cash signal on the open Sept. 7.

Regards,
Alex

Anonymous said...

Hi Alex,

i had been reading your blog for last few days, would like to know if you have some excel spreadsheet for historical data to download.I want to see COT data in excel for last few years.

Appreciate all your efforts and trying to understand it little bit better

thanks
Bill

Alex Roslin said...
This comment has been removed by the author.
Alex Roslin said...

Hi Bill,

That data is available from the CFTC website:

http://www.cftc.gov/
marketreports/
commitmentsoftraders/
cot_historical.html

Regards,
Alex

Anonymous said...

Hello

quick question on your signal for the long bond. Do you have any indication if that signal goes beyond this week? If so how long?

Am curious because that would seem to conflict with your S and P bullish signal for August 24. Does that create a conflict in your system?

thanks

Alex Roslin said...

Hi Anonymous,

The Treasury signal will go to cash or bearish after three weeks being long. I can't tell which at this point. There's no conflict with any other setup. Each setup stands on its own.

Regards,
Alex

In Debt We Trust said...

Alex, do you think the time to go long treasuries is based upon POMO Fed buying?

The longer dated the maturity, the higher the likelihood of DXY getting trashed in intra-day trading and equities rising.

Bonds, which normally trade inverse to equities, can also rise on such days.

http://www.newyorkfed.org/markets/operation_schedule.html

carl said...

Hi Alex,
We started to look at historic data of S&P500 and compared it to your signals. And yes on 1st glance you are right a lot of the time...!!! It seems that you are more right on BUY signals then on SELL, depending on how to calculate that(right/wrong, or how much right/wrong) More scientiffic data to follow. Question: the historic data from cot has to be turned around upside down in excell, for the formulas to work. Do you know how? My friend the math professor uses MATLAB and hasn't got a clue about excell ( to simple).... Anyway it getting more intrigueing all the time.
Greetings, Appi

Alex Roslin said...

Hi In Debt...,

It's an interesting question, but you've stumped me. It's hard to read whys into the data. That's a whole other field of research, though not an unworthy one. You could try to correlate all sorts of factors with the COT data. You could then, I suppose, develop setups based on those other factors too - or combine them with the COT data. Too much to do and too little time!

Regards,
Alex

Alex Roslin said...

Hi Appi,

Yes, generally the SPX signal until end of 2007 made more of its money long than short, and the confidence interval for profitability is a little higher than for beating the market (which tends to reflects the short side more). I think that's probably largely a reflection of the bullish tilt in the dataset from 1995. Using detrended price data - which is critical in development of a trading system - the profit on the short side is actually slightly higher than from the long side.

And in other markets, particularly Nikkei, which were more in a trading range in the 1995-2007 period, the confidence interval is actually higher on the market-beating side than on the profit-making side.

As for reversing the order of the data in Excel, use the "Sort" function. It is invaluable within Excel. Microsoft has a very good Excel tutorial site, including formulas for complex financial indicators:

http://office.microsoft.com/
en-us/excel/default.aspx

Regards,
Alex

carl said...

Hi Alex,
Thanx for the tip. I got confused a bit looking at the crude and uso-etf data, because in the cot-history there is missing february untill april 2000 period. Funny.
Amazing to see your system works. I guess you must have faith sometimes when signals are counter intuitive or contrarian to newspaper economic news. But in the end it is supply and demand that make the prices and that is done by traders. Something to have faith in, yes. Question: how can it be that american traders would have an influence on the nikkei-index. Wouldn't we have to look at the japanese cot-data? Aren't they much bigger? Or is it the psychological aspect of american traders being right(small traders) or wrong (large specs)?
Thanx for having patience with me..:)
Appi

Alex Roslin said...

Hi Appi,

It's not clear if the COT data is influencing prices directly or merely reflecting and prognosticating trends. The fact that the data gives signals for the Nikkei suggests the latter.

Regards,
Alex

Joseph said...

OIL is having trouble going over $75. Not sure if being Long into Sept 7th is wise. Whats your thoughts about this or I assume will have better knowledge after todays close. USO broke out as well, but OIL as mentioned just keeps bouncing at $75. Whats your view of the OIL $$ price for the Seek of Sept 4-7th

Thanks
Joseph

Alex Roslin said...

Hi Joseph,

Thanks for your question. Nothing has changed in my outlook. Even today's data won't really change anything for that setup because of its trade delays lasting well into September.

I also don't have any outlook for the crude price in September - except to say that past data patterns suggest that in similar situations historically, there was a better-than-odds probability of the crude price being higher at the end of the trade than it was when I started the position.

Regards,
Alex