Friday, 6 July 2007

Holiday Delay

Due to this week's 4th of July holiday, the Commitments of Traders report normally issued Fridays is delayed until Monday. Check back here Monday after 3:30 p.m. for my next update. Have a great weekend!

Wednesday, 4 July 2007

COTs Composite Equity Indicator: Neutral

Hope you're enjoying this slow market week and got some time off. I said I'd be back early this week with some thoughts on the latest Commitments of Traders readings. Sorry I didn't get to it as soon as planned.

Friday's COTs report - based on the previous Tuesday's data - has really moved into strange territory for the equity indexes - a true jumble of bullish and bearish signals. The "dumb money" small traders have struck a historically extreme net long position in S&P 500 futures and options. This gave me a new bearish signal on Friday.

Of the six equity markets in the report, three are now bearish - the S&P 500, Nikkei and S&P 400 Midcap. Meanwhile, the other three - the Dow Jones industrial average, NASDAQ-100 and Russell 2000 - are not only not bearish, they gave renewed bullish signals in the last report.

What can it all possibly mean? To make matters more complicated, my two most robust equity setups in terms of their market-beating profitability confidence levels - the Russell 2000 and the S&P 500 - don't agree with each other.

I think one reason for the discrepancies is probably genuine market uncertainty. So in order to give me a better idea, I created a composite indicator based on the five U.S. equity indexes in the COTs report. It's an average of the standard deviation values of these setups - in other words, how bullish or bearish they all are as a whole. Click here to see the table and chart of this composite indicator going back to 1995:

http://spreadsheets.google.com/pub?key=pfhv09UXvNetFM172ei1ffg

A reading of 1 or more means all five indexes are on average at the historic extremes of bullishness that trigger bullish signals across the board. A reading of 1 or less means the same bearish extremes.

You'll notice the indicator is pretty much neutral at the moment, following a couple of seesaw swings in recent months. It'll be interesting to see how the indicator acts as we move toward the fall.

You'll also see we're maybe not in such strange territory after all. It looks from a casual glance that the five COTs signals don't often swing into the same extremes of bullishness or bearishness. (At least, not based on the signal lines at 1 and -1.)

When the setups did agree, however, they seem to have made righteous calls about historic market junctures. New signals happened three times - a buy in April 1997, a sell in April 1999 and a buy in Jan. 2003. I think this will be an interesting new dataset to fine tune and possibly base setups around. Similar indicators can be created around other sectors of COTs data like energy and the metals. I'll let you know what I find. Stay tuned.

Meanwhile, I invite you to surf over to Kitco.com to read my regular weekly COTs precious metals report.

Friday, 29 June 2007

New Bearish Signal for S&P 500

What a week in the markets - phew! Right up into the close with the wild action. In this week's Commitments of Traders data, the first new bearish signal in a little while in one of the major equity markets - the S&P 500. (Check the explanatory links above for details on how my system trades this. For one thing, there's no trade delay in the S&P 500 trading setup, meaning execution on the next weekly open.)

Meanwhile, renewed bullish signals in some markets, while renewed bearish signals in others. Quite the motley bunch, in fact!

What to make of all this? I'll ponder that over the weekend and get back to you with some deep thoughts early next week. Meanwhile, best wishes for a happy and safe Canada Day to Canadian readers, and happy 4th to American readers.

New Signals*
BULLISH
None

BEARISH
-S&P 500

Renewed Signals**
BULLISH
-10-Year Treasury Yield
-NASDAQ-100
-Dow Jones Industrial Average
-Russell 2000
-Gold
-US Global Investors Funds US Gold Fund, USERX
-S&P/TSE Canadian Gold iUnits ETF, XGD.TO
-Gold Bugs Index, HUI

BEARISH
-Soybean Oil
-Natural Gas***
-S&P/TSE Canadian Energy iUnits ETF, XEG.TO
-Oil Service Holders, OIH
-Semiconductor Index, SOX
-Nikkei Average

Existing signals (date of original signal in parentheses)****
BULLISH

-30-Year Treasury Yield (3-Jan-07)
-10-Year Treasury Yield (17-Apr-07)
-S&P 500 (20-Mar-07)
-S&P/TSX Composite (15-Aug-06)
-NASDAQ 100 (27-Mar-07)
-Dow Jones Industrial Average (20-Mar-07)
-Russell 2000 (1-Aug-06)
-Gold (29-May-07)
-US Global Investors Funds US Gold Fund, USERX (12-Jun-07)
-S&P/TSE Canadian Gold iUnits ETF, XGD.TO (22-May-07)
-Gold Bugs Index, HUI (29-May-07)

BEARISH
-13-Week Treasury Bill (27-Feb-07)
-S&P/TSE Canadian Energy iUnits ETF, XEG.TO (3-Apr-07)
-Oil Service Holders, OIH (3-Apr-07)
-NASDAQ Composite (26-Dec-06)
-Semiconductor Index, SOX (20-Mar-07)
-S&P 400 Mid Cap (3-Jan-07)
-Nikkei Average (19-Dec-06)
-Soybean Oil (11-Nov-06)
-Silver (1-May-07)
-Copper (10-Apr-07)
-Canadian Dollar (10-Apr-07)
-U.S. Dollar Index (3-Oct-06)

NEUTRAL
-Crude Oil, Light Sweet (3-Apr-07)***
-Natural Gas (27-Mar-07)***

Notes
* For an explanation of what I do after a new signal, click “How It Works” above.
** A “renewed” signal is when a market is already on a bullish or bearish signal, and traders again register an extreme net trading position in the same direction. I normally ignore renewed signals unless I don't already have a trade on in this market. I haven't studied the profitability of trading on renewed signals.
*** See my special caveats for my Crude Oil and Natural Gas setups (click “The Trading Setups” above and check the table footnotes).
**** The date in parentheses is the date of the COTs report that gave this signal - not the date I would have executed the trade (which can be up to five weeks later). The existing signals are often several months old and are listed here as references, not trading recommendations.

Thursday, 28 June 2007

New 13-Wk T-Bill Yield Setup is Bearish

I just posted results of a really interesting new setup for the 13-Week Treasury Bill Yield on my "Profit/Loss Results" page (click the link to the right). The setup is based on the Commitments of Traders data for the 3-Month Eurodollars.

Specifically, I'm fading the small traders when they hit specific historic extremes of bullishness and bearishness. (There's no COTs data for the T-Bills, but the Eurodollars give great signals for this market.)

The setup flipped to bearish on the T-Bill yield in the COTs report dated Feb. 27, 2007 (with a one-week trade delay) and remains bearish right now. (Bearish on the yield means the yield goes down.) For how this all works, click the explanatory links to the right.

Wednesday, 27 June 2007

Filled!

Bought some Canadian Gold iUnits (XGD) just before the close yesterday to boost my allocation to my maximum for this setup. That's not normally the way I do it, but I have some portfolio rejigging I needed to do.

As part of that process, I also need to sell some of my equities, including the Japan iShares ETF (EWJ) - for which I now have a bearish signal, according to my new Nikkei setup. But today's reversal is keeping me in until a get technical sell signals. (Since switching to COTs-based mechanical trading, I miss trading off the technicals, so this is lots of fun for me.)

I based the XGD trade on an interesting technical system I first heard of on Stephen Vita's excellent AlchemyOfTrading.com website and blog. It's the TD Sequential System named after its developer, Tom DeMark. Stephen uses it a lot, and I've found it's a great way to study overbought and oversold conditions in various timeframes. For the XGD trade, I bought after a TD Seq Countdown buy on the 60-minute chart.

TD Seq is easy to learn and use, and there's lots of info on the web about it or in Tom DeMark's books.

Tuesday, 26 June 2007

Thanks, Hal!

One of the nice things about having a blog is meeting generous and gracious people. I wanted to post a special message to thank one such reader, Hal Read of Arizona. He wrote me early on to share his insights into trading system development and validation. His comments were very helpful as I thought about how to refine my trading setups to make them as robust as possible. Thanks, Hal!

Like the Casinos Do It

Just picked up some UltraShort Oil & Gas ProShares (DUG) as it broke out above this morning's opening range in order to get up to my max. allocation in that area as part of some overall portfolio rebalancing I'm doing.

Also looking at a spot to pick up some more Canadian Gold iUnits (XGD). This bloodbath may lead to a good technical buying opportunity to bring me to my max allocation for this market.

Speaking of drawdowns, this can be one of the toughest parts of using mechanical systems. It's often not the system that fails, but the trader's discipline.

To that end, I just read something interesting in Way of the Turtle by Curtis Faith - a great book on trading and developing trading systems that a member of the Canadian Society of Technical Analysts was kind enough to recommend. Faith likens the inevitable drawdowns to the wins a casino must occasionally pay to gamblers.

"Casino owners do not care about the losses they incur because such losses only encourage their gambling clientele," Faith writes. "For owners, losses are just the cost of doing business; they know they will come out ahead over the long run."

Same thing with losses in a trading system, he says. "They are the cost of doing business rather than an indication of a trading error or a bad decision. To approach losses in this way, we had to know that the method by which the losses were incurred would pay out over the long run." That seems to be the key.

And to that end, I've just gotten a copy of Robert Pardo's classic Design, Testing, and Optimization of Trading Systems. I hope to review my system some more based on Pardo's techniques. I'll let you know how it goes.

Friday, 22 June 2007

Thumbs Up for Equities, Down for Energy

Today's Commitments of Traders has generally bullish news for equity markets, if the three renewed bullish signals I've gotten are any indication.

But there were more bearish tidings for energy, with several renewed bearish signals in this sector. No new signals to report this week.

Below is my first series of signals using my new list of trading setups based solely on the COTs reports. I have greater confidence in these setups because I was able to statistically validate almost all of them at a high level of confidence, both in terms of profitability and against the underlying markets I'm trying to beat.

In many cases, the new setups are more profitable than the old ones. In a few cases, I had to make small sacrifices in terms of profitability to get a setup in which I can have more confidence about the future results matching past profits. The problem with some of the old setups was that the small number of trades made it impossible to say with confidence whether the profitability was due to excessive curve-fitting. Now, I've managed to find setups with enough trades to be statistically validated.

Most of the existing signals remain the same, but there are a few notable changes. All of my gold-related setups are now in the bullish column, having just flipped from bearish. As well, you'll notice that my new setups for the NASDAQ Composite, Semiconductor Index (SOX) and the Nikkei are now in the bearish column.

The changes come as I've switched which group of traders I follow for these markets. In the Nikkei, for example, my old setup followed the commercials, who have been highly bullish the Nikkei, but the new one fades the small traders, who've also been highly bullish. Go figure! It's a bit of surprise for me because I've got a profitable position in the Japan iShares ETF (EWJ) as a result of the earlier signals. I'm going to find a good technical spot to unload it.

Note, however, that my setup for the NASDAQ-100 is still bullish, so I'm holding on to my Ultra QQQ ProShares ETF (QLD), which tracks that index.

New Signals*
BULLISH
None

BEARISH
None

Renewed Signals**
BULLISH
-Dow Jones Industrial Average
-NASDAQ 100
-Russell 2000

BEARISH
-Soybean Oil
-Natural Gas***
-Crude Oil***
-S&P/TSE Canadian Energy iUnits ETF, XEG.TO
-Oil Service Holders, OIH
-Semiconductor Index, SOX

Existing signals (date of original signal in parentheses)****
BULLISH

-30-Year Treasury Yield (3-Jan-07)
-10-Year Treasury Yield (17-Apr-07)
-S&P 500 (20-Mar-07)
-S&P/TSX Composite (15-Aug-06)
-NASDAQ 100 (27-Mar-07)
-Dow Jones Industrial Average (20-Mar-07)
-Russell 2000 (1-Aug-06)
-Gold (29-May-07)
-US Global Investors Funds US Gold Fund, USERX (12-Jun-07)
-S&P/TSE Canadian Gold iUnits ETF, XGD.TO (22-May-07)
-Gold Bugs Index, HUI (29-May-07)

BEARISH
-S&P/TSE Canadian Energy iUnits ETF, XEG.TO (3-Apr-07)
-Oil Service Holders, OIH (3-Apr-07)
-NASDAQ Composite (26-Dec-06)
-Semiconductor Index, SOX (20-Mar-07)
-S&P 400 Mid Cap (3-Jan-07)
-Nikkei Average (19-Dec-06)
-Soybean Oil (11-Nov-06)
-Silver (1-May-07)
-Copper (10-Apr-07)
-Canadian Dollar (10-Apr-07)
-U.S. Dollar Index (3-Oct-06)

NEUTRAL
-Crude Oil, Light Sweet (3-Apr-07)***
-Natural Gas (27-Mar-07)***

Notes
* For an explanation of what I do after a new signal, click “How It Works” to the right.
** A “renewed” signal is when a market is already on a bullish or bearish signal, and traders again register an extreme net trading position in the same direction. I normally ignore renewed signals unless I don't already have a trade on in this market. I haven't studied the profitability of trading on renewed signals.
*** See my special caveats for my Crude Oil and Natural Gas setups (click “Profit/Loss Results” to the right and check the footnotes).
**** The date in parentheses is the date of the COTs report that gave this signal - not the date I would have executed the trade (which can be up to four weeks later). For details on how I trade this system, including trade delays and portfolio allocations, click on "How It Works" and “Profit/Loss Results” to the right. These "existing signals," which are mostly several months old, are listed here as a reference, not a trading recommendation.

Thursday, 21 June 2007

Update: New Trading Setups!

I've just posted the results for my revised list of trading setups. Surf to the "Profit/Loss Results" page in the right-hand column and click the link provided. Make sure to check out the explanatory notes on Sheet 2.

I'll post signals for these new setups starting with tomorrow's Commitments of Traders reports. See you Friday!

Update: New Setups Almost Ready

My apologies for the lack of Commitments of Traders updates this week! I've been busy refining a new series of trading setups for the markets I'm following. I'm quite excited about the results. I've found high-profit, statistically validated setups in almost every market. Some of the setups will remain the same, but most will change a little - or a lot in a few cases.

The importance of validation is to get a confidence level of how past results may translate into future performance. I'm also applying a second test to see how confident we can be that the setup will beat the underlying market.

I should have the details ready by the time of tomorrow's COTs report. Thanks for your patience.

Sorry also to those who wrote asking for a sample spreadsheet for my S&P 500 trading setup. The setup has slightly changed, so I wanted to finalize it before sending it out. If you'd like the new spreadsheet when it's ready, email me at aroslin1@yahoo.ca.

Saturday, 16 June 2007

Update: New Setups Turn Bullish on Gold, HUI, USERX Gold Fund

I've been busily refining my COTs-based trading setups - looking for ones that are profitable, beat the markets and can be validated statistically. I guess my timing was good because I've just found validated setups for gold and the USERX Gold Fund - and they've both just given new "bullish" signals, like my XGD gold ETF setup (see post below).

Unfortunately, I'm having some trouble updating my "Profit/Loss Results" page with all the details, so I'll report them right here.

My new gold setup gave a "bullish" signal in the Commitments of Traders report dated May 29, with a trade delay of zero (meaning I would have executed my buy on the open Monday, June 4). This setup had a maximum drawdown of 9 percent (meaning my maximum portfolio allocation would be 22 percent). The setup had 16 trades (12 wins and four losses), with an average profit of 9.4 percent and an overall profit of 352 percent since the first trade, compared to 174 percent buying and holding gold itself.

My new USERX setup gave a "bullish" signal in the latest COTs report, also with no trade delay (meaning my buy should be executed for the open Monday, June 18) like the XGD trade (see post below). This setup had a maximum drawdown of 28 percent (meaning my maximum portfolio allocation would be 7 percent). There were 17 trades (13 wins and four losses), with an average profit of 32.5 percent and an overall profit of 2,743 percent since the first trade, compared to a 25-percent loss for the USERX Gold Fund.

I also got a "bullish" signal for a new setup for the HUI Gold Bugs Index in the May 29 COTs report, but this setup is only validated at a 90-percent confidence level (meaning that 90 percent of the time, the average trade is expected to be profitable) - which is less than the 95-percent confidence I want to see in my trading setups.

All three of these new setups are based on trading the same side as the commercial traders when their net percentage-of-open-interest position in the combined futures-and-options COTs data hits specific historic extremes that in the past have led to superior profits.

I'll update the new results on the "Profit/Loss Results" page as soon as I work out the glitches.

Friday, 15 June 2007

"Bullish" Signal for Canadian Gold ETF

A new bullish signal came in this week's Commitments of Traders reports. The "smart money" commercial traders have flipped to a historic extreme of bullishness in gold, giving a "bullish" signal for Canada's Gold iUnits exchange-traded fund (symbol XGD). The commercial traders have been on a "bearish" signal for XGD and the rest of the gold sector since January and February.

Check the "Profit/Loss Results" and "How It Works" links to the right if you're interested in how I trade new signals for this and my other setups. For example, my XGD setup calls for the trade to be executed on the next weekly open (that is, Monday, June 18).

Please keep in mind, though, that this setup's results are based on only six past trades. (That's because the data on XGD goes back only to March 2001.) That means this setup can't yet be validated statistically. In other words, I can't give any confidence level on its future performance. My other gold-related setups came just shy of giving similar "bullish" signals this week, but nothing's been triggered so far, so they remain "bearish" until further notice.

The new XGD bullish signal does, however, look timely since XGD today had a go at a breakout from its recent downtrend. Check here for more analysis of the latest COTs release next Monday or Tuesday.

New Signals*
BULLISH
-S&P/TSE Canadian Gold iUnits ETF, symbol: XGD.TO

BEARISH
None

Renewed Signals**
BULLISH
-TSX

BEARISH
-Soybean Oil
-Natural Gas***
-Crude Oil***

Existing signals (date of original signal in parentheses)****
BULLISH

-S&P/TSE Canadian Gold iUnits ETF, symbol: XGD.TO (12-Jun-07)
-30-Year Treasury Yield (3-Jan-07)
-10-Year Treasury Yield (20-Mar-07)
-S&P 500 (20-Mar-07)
-NASDAQ (27-Mar-07)
-Semiconductor Index, symbol: SOX (30-Mar-07)
-Dow Jones Industrial Average (20-Dec-05)
-Russell 2000 (25-Mar-03)
-Nikkei (3-Feb-04)
-TSX (20-Mar-07)

BEARISH
-S&P/TSE Canadian Energy iUnits ETF, symbol: XEG.TO (3-Apr-07)
-Oil Service Holders ETF, symbol: OIH (18-Apr-06)
-Soybean Oil (2-May-06)
-US Global Investors Funds US Gold Fund, USERX (30-Jan-07)
-Gold Bugs Index, HUI (13-Feb-07)
-Canadian Dollar (10-Apr-07)
-U.S. Dollar Index (3-Oct-06)
-Gold (13-Feb-07)
-Copper (10-Apr-07)

NEUTRAL
-S&P 400 Midcap (3-Jan-07)
-Crude Oil, Light Sweet (3-Apr-07)***
-Natural Gas (27-Mar-07)***
-Silver (21-Nov-06)

Notes
* For an explanation of what I do after a new signal, click “How It Works” to the right.
** A “renewed” signal is when a market is already on a bullish or bearish signal, and traders again register an extreme net trading position in the same direction. I normally ignore renewed signals unless I don't already have a trade on in this market. I haven't studied the profitability of trading on renewed signals.
*** See my special caveats for my Crude Oil and Natural Gas setups (click “Profit/Loss Results” to the right and check the footnotes).
**** The date in parentheses is the date of the COTs report that gave this signal - not the date I executed the trade (which can be up to four weeks later). For details on how I trade this system, including trade delays and portfolio allocations, click on "How It Works" and “Profit/Loss Results” to the right. These "existing signals," which are mostly several months old, are listed here as a reference, not a trading recommendation.

Thursday, 14 June 2007

Yields Have Topped, According to New 10-Year Treasury Setup

I've just posted the results of a new improved trading setup for the 10-year Treasury note. It's based on trading the same side as the small traders, not the usual practice of fading this group.

I've validated this setup at the 95-percent confidence level. (Ninety-five percent of the time, the average trade is expected to be profitable given a normal distribution of results.)

Also interesting is the fact the setup just gave a "bearish" signal for the 10-year yield - meaning yields may have just topped. The signal came in the COTs report dated May 22. My trade delay for this setup is one week, meaning the trade was to be executed for the open of trading on Monday, June 4.

Update: I Feel Validated... Somewhat

We all like to be validated. I'd like to give you an update on what I've been up to in terms of validating the COTs Timer system. As I've mentioned before, this trading system is a work-in-progress. That's one of the reasions I've urged caution about acting on my signals and recently switched from outright "buys" and "sells" to "bullish" or "bearish."

If you've been a regular visitor, you've seen that I'm often revising the setups and looking for ways to validate the system in general. The last point is important because of the small number of trades for many of the setups and the fact that the Commitments of Traders combined futures-and-options data goes back only to 1995. The question is whether past profits tell us anything about future results.

Happily, statistics give us a few ways to test this question and validate trading systems. One is the Student's t-test. It is often used to give confidence levels for trading systems and other data like survey results. Ideally, we would like a trading system to be profitable at a 95- or 99-percent confidence level. (In other words, 95 or 99 percent of the time, the system's average trade is profitable.)

A key variable is the number of trades generated by a trading setup. The more trades, the more confident we can be the future results will be profitable. Because of the small number of trades, I've found many of my setups are profitable only at the 90-percent confidence level at this point. A few are valid at the 95- or even 99-percent levels. Some still simply have too few trades to have a confidence level.

Thus, a setup needs at least 10 trades to have a confidence level, based on the number of trading variables in my system. Of if I were to drop my stop rule - which counts as two variables - a setup needs at least eight trades. Click the "Profit/Loss Results" link to the right to see the number of trades in each setup, keeping in mind that I haven't updated some of them in a while. (The "Last Updated" column tells you when.)

I plan to post some specifics for each setup soon and to keep testing the system in other ways. My apologies for not getting to this sooner. This testing has been extremely time-consuming, but it's also been an extremely valuable and educational process that's helped me find better setups in a couple of markets. I hope to post those details soon, too.

Many thanks to the readers who wrote in about these questions and were so generous with their time and sharing their insights. Your further input is welcome.

Tuesday, 12 June 2007

"Big Money in the Main Movements"

Last week's market craziness continues as we speak. The uptrends in the major equity indexes since March appear to have broken (even if not decisively just yet).

Wall St. and regular folks alike are all wondering where it will stop. Will the recent lows hold? Are rising bond yields going to break the market's back? Is this 1987 all over again?

One answer may lie in the Commitments of Traders reports. Last Friday’s report - based on data as of last Tuesday, June 5 - showed the “smart money” commercial hedgers still heavily net long the S&P 500 index. Meanwhile, the "dumb money" large speculators - so called because they tend to get it wrong at market turns - are still heavily net short the NASDAQ, a bullish sign.

For other indexes, the picture is essentially neutral. Overall, the latest report hasn't overturned any of the bullish signals that my system generated before for equities. Same thing for the bearish signals in the metals and energy. That would change only if traders register new historic extreme net positions in the other direction. So for me and my positions, that means the existing signals are still good.

All this market turbulence does raise another question I also posed in a report I just did for Kitco.com, the bullion website: Should we worry about weekly fluctuations in the COTs data? I don’t believe so. Anyone trying to find correlations between weekly changes in the data and market prices will quickly come to a surprising conclusion: There is virtually no correlation.

This might be puzzling to readers of the financial media, which widely reports on the COTs reports’ weekly changes in the number of contracts in various markets.

But such short-term changes have little value for trading purposes, as the lack of correlations shows. Last week, for example, the COTs data was fairly upbeat for precious metals, as the commercial traders slightly reduced their net short positions in gold and silver. We might have expected prices to rise. Instead, there was a rout.

So are the COTs reports worthless? Analysts have been befuddled about the data for years. It has to mean something, they thought. A breakthrough came when some clever analysts suggested it’s best to follow the commercials when they reach historic extremes in their net percentage-of-open-interest positions. I think that’s the right approach.

My own twist on it is that I found it was best to fade - or trade opposite to - the large speculators and small traders in some markets. Also, I think it’s important to backtest and validate the specific historic extremes that have led to the most consistently profitable, least-volatile trading signals in each market.

That’s why I don’t get too excited any more about short-term changes in the COTs data. Edwin Lefèvre said something along similar lines in Reminiscences of a Stock Operator. (Yes, here's another quote from this great read.) “The big money was not in the individual fluctuations but in the main movements… I want to tell you this: It never was my thinking that made the big money for me. It always was my sitting. Got that? My sitting tight!”