Saturday 28 February 2009

Small Makeover for Gold Setup

My trading setup for gold based on the Commitments of Traders reports has gotten a small makeover. The new specs and results are all posted on my latest signals table. The change is that I've tweaked the parameter values for the large spec total open interest setup. The upper signal line has gone from 0.2 standard deviations above the moving average to 0.25. The lower signal line went from 0.6 standard deviations below the average to 0.65. Not a big different, but those values produced slightly better results in my walk-around testing - i.e., my check on the backtested performance of neighbouring setups that have slightly altered parameter values. A robust setup should have neighbouring setups that also performed well. Taking a look at them is an important way to reduce the risk of curve-fitting and improve the odds of good performance in actual trading. Right now, the new setup is in cash (like the old one). Now that I've finalized my testing on the S&P 500 and gold - at long last! - I'm planning to get to work on new setups for a couple of markets that aren't strongly correlated to the first two: the Nikkei and natural gas.

Friday 27 February 2009

S&P 500 Looking Gloomy

I've been busy most of the day with some virus problems, so apologies for an abbreviated Commitments of Traders update today. My two trading setups based on this weekly derivatives data are unchanged: bearish for the S&P 500 for next week and cash for gold.

My S&P 500 setup will be going to cash on the open of trading on Monday, March 9. That signal will last all of one week, then revert to bearish for at least two more weeks... and possibly more. Three weeks out is all the signal can see. The data for the S&P 500 is uniformly bearish, except for that brief respite in a week's time. In this afternoon's COT data, the commercial traders are 0.47 standard deviations below the moving average I use for their signal. And the small traders are still far too exuberant - their net percentage-of-open-interest position 1.48 standard deviations above the average. That's down from 1.61 standard deviations the week before - but still far from falling into bearish territory. Unfortunately, the little guy doesn't seem to have lost enough money yet.

In gold, still no resolution in the standoff in my trading setup. The large specs have dropped their net position as a percentage of the total interest two weeks in a row. They now sit 1.02 standard deviations above the average. That is bullish. These guys are usually wrongly positioned at market turns. However, the large spec total open interest is still alarmingly high. It's at 1.07 standard deviations above the average. We need to see that come down dramatically at least to 0.6 standard deviations below the average to flip the entire setup back to bullish. So the overall setup remains in cash for a fourth straight week. It could mean there needs to be a pullback before gold will continue on its march to $3,000 or whatever the gold bugs are predicting these days.

Hope you have a good weekend. And be sure to check out the updated post on my portfolio page.

TAGS: S&P 500, SPX, gold, COT, Commitments of Traders, derivatives, Black Swans, market timing, trading system development, CFTC, Commodity Futures Trading Commission, COTs Timer, Monte Carlo, out-of-sample testing, walk-around testing

Friday 20 February 2009

S&P 500 Bearish Two More Weeks

Okay, this isn't funny any more. I just checked the BKX Bank Index. It's down 50 percent YTD. How is that even possible? The mind reels. I hope you fared okay this last week. The Commitments of Traders report from this afternoon doesn't give much hope things will get better soon. At least not the way I read it. This weekly government data on futures and options holdings gives me no new signals for next week. So my existing bearish signal for the S&P 500 and cash for gold are still in effect. (See my latest signals table for more details.) Some other highlights:

- Odds are we're not at a bottom yet, according to my read of the latest S&P 500 COT data. The small traders in S&P 500 futures and options are showing, yet again, why everyone calls them the "dumb money." While the S&P 500 fell another 48 points from this week's open to the close, the small traders significantly stepped up their net long position as a percentage of the total open interest. They went from being 1.02 standard deviations above the moving average to 1.61 standard deviations above. I have a feeling we're not going to see a bottom until these unfortunate folks lose a lot more money. The sad reality is they first hit an extreme of bullishness the week of Nov. 4 expecting the usual Christmas rally, coupled with some Obamania. That week opened at 969. This week closed at 770. That's down 20.5 percent. You'd think the bear market would have sunk in by now. Not according to these numbers.

The "smart money" commercial traders got a little bit bullish last week, but reversed course quickly and this week are back to very bearish. Their net percentage-of-open-interest position is 0.49 standard deviations below the average - down from 0.21 standard deviations above the average last Friday. That short-lived blip of optimism gives me a new delayed signal for two weeks out. My setup will go to cash for the week of March 9 - because of the two signals disagreeing with each other temporarily. It will then go back to bearish the following week.

- My gold setup is in cash for a third week in a row. The wrong-way large speculator crowd's net position is not at any extreme of bulllishness yet. In fact, it fell this week to 1.22 standard deviations above the average, from last week's 1.37 standard deviations above. But the large speculator total open interest (long plus short positions) remains highly elevated, which gives me a bearish signal. So the setup as a whole is in cash because the two signals don't agree. Incidentally, I'm re-examining my gold setups a little right now because I've identified a few possible setups that might be slightly more robust. I am quite satisfied with the existing one, but I am going back to study it against some other possible setups to make sure I'm using the very best one possible, based on my latest appreciation for things statistical.

Hope you have a good weekend. See you early next week with an update of my portfolio page and maybe more news on that new gold setup.

TAGS: S&P 500, SPX, gold, COT, Commitments of Traders, derivatives, Black Swans, market timing, trading system development, CFTC, Commodity Futures Trading Commission, COTs Timer, Monte Carlo, out-of-sample testing, walk-around testing

Friday 13 February 2009

S&P 500 Bearish, Data Mixed for Gold

No new signals this week for my trading setups for the S&P 500 and gold based on the Commitments of Traders reports, but the data is saying some interesting things. This afternoon's new data keeps both setups in their existing signals: bearish for the S&P 500 and cash for gold. (Be sure to read my post from this morning about my new S&P 500 setup.)

- For the S&P 500, the new setup is based on trading alongside the commercial traders when they hit extremes of futures and options positioning - and fading the small traders. The commercial traders have dramatically reduced their net short position as a percentage of the total open interest since its bearish extreme in late November, when it hit 2.8 standard deviations below the moving average I use for that signal. The commercials have jumped from being 0.5 standard deviations below the average last week to 0.2 standard deviations above the average now. That signal has now gone bullish (with a trade delay of three weeks).

But my other signal for this setup - based on trading opposite to the small traders - remains bearish. That's because these guys remain one standard deviation above the moving average for that setup and are still far from working off the bullish excess in their positioning that they've displayed since the week of the U.S. presidential election, when they suddenly shot up to 1.96 standard deviations above the average - a huge jump from 0.2 standard deviations below that average the week previous. Until the small traders get more burned, this signal won't go to bullish. At best, the setup will go to cash in three weeks' time due to the commercial signal not agreeing with the small traders' call.

- For gold, the large speculators, whom this setup fades, have steadily built up a net long positioning that is starting to approach extreme territory, but we're not there yet. That signal is now 1.4 standard deviations above the average, but needs to hit 1.9 standard deviations for it to flip the signal to bearish. However, the other signal for this setup - based on fading the large spec total open interest (long plus short positioning) - reveals that the large spec total positioning has been at a bullish extreme for several weeks - 0.9 standard deviations above the average this week. Thus, this signal is bearish. So the combined setup is in cash for a second week this week. No joy.

Hope you did okay this week and have a good weekend and Happy Valentine's Day. See you next week, and be sure to check my portfolio page with an update.

TAGS: S&P 500, SPX, gold, COT, Commitments of Traders, derivatives, Black Swans, market timing, trading system development, CFTC, Commodity Futures Trading Commission, COTs Timer, Monte Carlo, out-of-sample testing, walk-around testing

New S&P 500 Setup Bearish

Sorry I missed my promised post of earlier this week updating my take on the last Commitments of Traders data. I was busy finalizing another look at my S&P 500 trading setups based on this weekly government data that tells us how big traders are positioned in the futures and options markets. I was reviewing my best setups based on slightly rebalanced weightings for the indicators I'm using to choose the best ones. I've given a little more weight to my "walk-around" test results (i.e., how "neighbouring" setups do, with slightly altered parametre values) and to my Monte Carlo results. Turns out that those measures tend to correlate nicely with how setups did during the market disaster of the last year.

My new S&P 500 setup's specs are all now posted on my latest signals table, and you can see it for yourself on my DIY sample spreadsheet page, including the equity curves for the setup and the S&P 500. Unlike my last S&P 500 setup, which lost about 14 percent in 2008, this one was down 7.6 percent - not great, but not as bad as the market's 40.9 percent. (Those numbers are based on weekly open prices and, for my old setup, don't take into account my stops or Black Swan rule, which would have slightly reduced that loss.) With this kind of system, it's nearly impossible to design it so it will win each and every year. In fact, doing so deliberately would have a good chance of creating a non-robust system with poor future results. So one bad year doesn't worry me, especially if the setup beat the market by a wide margin. You might be wondering why I don't go backwards and look for systems that did well in 2008. Again, that kind of exercize runs the risk of cherry-picking a setup that's useless in real-life trading. In fact, the leading contender setups I looked at that were profitable last year tended to be inferior in various robustness measures.

Meanwhile, this year, my new S&P 500 setup is up 3.5 percent so far (as of Thursday's close). And the setup's compound annual growth of 17.5 percent was 72 percent greater than the market's between 2003 and 2007 (including a 0.2-percent trade friction per trade for commissions and slippage). Far more importantly to my mind, its robustness scores in walk-around, out-of-sample and Monte Carlo testing are all very strong.

What does the setup say now? It has been bearish since Dec. 1 and remains so right now - with little sign of the data being anywhere close to turning bulllish. The latest bearish signal is up 5.4 percent as of yesterday's close. You might recall the setup I was using until now was also bearish until this week, when it went to cash for a single week, and then was going back to bearish. So either way, I'd be bearish going into next week's open. Tune back in here for another update from this afternoon's COT data. Good luck today.

Saturday 7 February 2009

S&P 500, Gold Both to Cash

My latest signals table is now updated based on Friday's Commitments of Traders data. Both of my setups - for the S&P 500 and gold - are in cash as of next Monday. I'll post a more detailed update on the data early next week - along with a portfolio update.