The data shows the wrong-way small traders in S&P 500 futures and options at a remarkable 2.02 standard deviations below the moving average I use to study their relative positioning. That's their most bearish in over a year. The small traders have steadily reduced their relative net positioning for seven straight weeks as a percentage of the total interest - during almost the entire rise of the market off its mid-July low. Meanwhile, the commercial hedgers, who tend to be correctly positioned at market turns, have gotten even more bullish than they were last week.
All this bodes very nicely for my bullish signal for the S&P 500 that takes effect on Monday's open of trading. I'll hold that trade for at least three weeks due to the trade delays within this setup. Some other highlights from the new data:
- Banks: Mixed news for U.S. financials. Large speculator net positioning in the three-month Eurodollar contract has completely reversed course after a four-week decline as a percentage of the total open interest. And that is bearish. Today's data shows it bumping back up, which has turned their signal bullish. However, the small trader total open interest continues to fall off a cliff, and that is bearish. My setup for the U.S. BKX Bank Index remains in cash for a third week.
- Crude oil: My trading setup for crude oil - presently bullish - stays so for two more weeks, then goes to cash. Friday's data shows the small trader net position collapsing as a percentage of the total open interest. Their signal has now gone to bearish. So if the commercial hedgers, which are already on a bearish signal, remain that way inclined a few more weeks, we could get actually a bearish call from this setup in eight weeks' time. (The small trader signal has an eight-week trade delay before execution.)
- Natural gas: My setup for natural gas goes bearish on Monday's open. Gas has already gotten destroyed for more than two weeks. (I bought a discretionary short position last week as it broke down.) My COT signal will last two weeks, after which time the setup will go either to cash or bullish. Today's data saw the small trader total open interest rally nicely, which is bullish in this market according to my backtesting. But that signal has a two-week trade delay, so it won't impact my setup right away.
Be sure to tune in early next week for my portfolio update. Have a good weekend, and rabbit feet for next week.
24 comments:
I'm curious if you have tried your system with the ES COT data instead of SPX and/or done any comparisons? It's interesting because the SPX COT data is pretty much the opposite of the ES COT data. In the past I have always used the SPX data but my research has found that commercials are moving from the SPX to the ES. Over the past year ES open interest has increased 8% while SPX open interest has decreased 25%.
Your SPX analysis is that commercials are buying and small traders are selling. I can see this in the net position numbers (long - short). However using the ES data, it's the opposite. The commercials are selling and the small traders are buying.
And if we combine the ES & SPX (account for the contract size difference), the result is similar to the ES result (commercials short and small traders long).
I'd love to hear your thoughts on this. Thanks in advance.
Hi Michael,
Thanks for your message. The mini data hasn't given as robust results in the testing I've done. Mind you, I haven't re-tested it with my current methodology. So much to do, so little time!
The difference in positioning could be due to differences in the value of the two contracts. They're not the same kinds of traders necessarily. Combining the data between the two contracts could be problematic because of that.
Regards,
Alex
What I'm saying is they are the same traders. Here's another statistic, over the past year, ES commercial open interest has risen 130% while SPX commercial open interest has declined 83%. This shows the commercials are moving from SPX to ES.
In the ES, commercial net position is 143.17% of open interest. In SPX it's 134.02%. One year ago these percentages were 118.63% & 120.92%. This tells me the percent of OI due to commercials is greater in ES and it has become even more greater over the past year.
in the past the ES results might not have been good because the ES had less commercial traders and more speculators. But that is changing. Unfortunately I'm not sure how to account for this for backtesting purposes.
Finally, for combining the two I multiply the ES net position by 5 or divide the SPX net position by 5, based on the $50/$250 multiplier difference. It takes 5 ES contracts to equal 1 SPX contract.
According to my equity curve of your system, it's currently in a drawdown. My hunch is that the drawdown could continue if the ES commercials end up being on the right side.
The bears had it coming to them:
http://debtsofanation.blogspot.com/2009/08/debts-of-world-
where-did-bears-go.html
The markets will be flooded w/fresh liquidity next week due to 2 POMO's scheduled.
As you may know, the Fed routinely sprinkles free cash to the primary dealers through buying treasuries from them on a routine basis.
On these days, the market tends to go up as the DXY sinks (to accomodate Bernanke's $ debasement) and equities roar as PDs use that freshly minted cash to speculate in stock futures.
http://www.newyorkfed.org/markets/operation_schedule.html
Hi,
Seeing the changes in natural gas positions last week (short 8,376 and long 28,811)
for the commercial traders, and observing that NG has fallen badly last couple of days dont you think that this will affect NG more on the upside. Dont you think that the commercial number are even higher as we speak, regarding the fall last week?
Dont you think the commercial traders will carry the price higher next two weeks? Its 322,981 against 196,084, and maybe even more on the long side.
thoughts?
Hi Alex,
Should the Eurodollar Large Speculator Total OI number be 1.08 instead of minus 1.08?
Should the Oil Small Trader Net % be minus 1.32 instead of 1.32?
Regards,
Padraic
Hi Padraic,
No, those numbers are correct as indicated.
Regards,
Alex
Hi Anonymous,
My setup for natural gas doesn't involve the commercial traders, so they're positioning isn't relevant to me. The two groups of traders that setup is based on did also, however, get more bullish on Friday. But backtesting shows their signals are most effectively traded with the trade delays indicated for this setup. So historically, price didn't move right away after such changes in positioning.
Regards,
Alex
Hi Alex,
Sorry to bother you again.
Why is the Net position Small Trader OIL Signal now bold red Bearish if the COT position is 1.32?
Aren't +ve signals bullish?
Regards,
Padraic
Hi Padraic,
Oops - you were right in your earlier message. It should have said -1.32. I actually reopened my spreadsheet and checked it, but I missed have been sleep computing this morning or something. Thanks for asking again. I've now corrected it.
Best regards,
Alex
Hi Michael,
Thanks for your message. Your analysis of the data is interesting but doesn't capture the whole picture. For example, you fail to note that the mini commercial total open interest also collapsed after it peaked in Dec. 2008, falling by nearly half by March. It's still well off where it was last year.
It's also useful to look back further than just one year, as is vital in backtesting trading systems. If you do that, you'll notice that the commercial total open interest in the larger SPX contract had doubled between June 2008 and its peak last Dec. 2008. So even though it's true it's fallen since then, it fell from bubble highs.
Even more importantly in my eyes, my research suggests the commercial total open interest in the larger SPX contract does not give very reliable trading information in general. I didn't get a single potential trading setup out of the total commercial open interest during my backtesting.
That might not be the case in the mini contract. I just don't know until I do some backtesting. But a year's data doesn't mean anything, in my opinion.
One other point is that the commercial traders in the mini contract really are not the same as those in the SPX contract, which is five times bigger. The last time I checked, the CFTC's reporting levels for both contracts are the same. That means the commercial category in the mini contract includes a number of traders who would be considered small traders (i.e. the non-reportable category) in the larger contract.
Finally, regarding the current drawdown in my crude oil trading setup, that has nothing to do with the SPX setup. Completely different data. Some trades just don't work out. That's the nature of trading. Take a look at my backtesting results table to see that none of my setups has a 100% win record. Far from it. That's why I use risk-control measures. As well, that trade isn't over yet. It could very well end up in the black.
Regards,
Alex
Hi again Michael,
One other thought occurred to me: It's true there's been a very long-term growth in the commercial total open interest in the mini S&P 500 contract. That actually started pretty much at the beginning of the data in 1997. That, however, hasn't been reflected either in a long-term decline in the commercial total open interest in the larger S&P 500 contract. Neither has it been reflected in any kind of deterioration of trading setup backtesting results in more recent years. The setup I'm using for the S&P 500 has results in the 2003-2008 period that match those in the 1995-2003 timeframe.
All that said, I still think it's worth doing some testing for setups in the mini data. I just don't think any of these trends impacts my S&P 500 setup very much.
Regards,
Alex
Hi Alex, thanks for responding. BTW I didn't mention Crude, I was talking about your SPX system being in drawdown.
You're correct that the emini commercials contains commercials that would be in the small trader category on the larger contract. This is why we can't just fade the small traders, there are commercials hiding in there. If anything this should be an argument for the emini, because it keeps the commercials in the commercial category and makes the small traders "dumber". :)
I'm just brainstorming in all this, anything can happen in a year and it could very well be that next year commercials are back to the SPX. But my hunch is the pit trading is dying out. Everything is going electronic.
I have your SPX system programmed into Tradestation. I'm going to run some tests with the ES and let you know what I find.
Hi Michael,
Thanks. Let me know what you find. It's certainly true trading is changing. The question is how does that impact the patterns and signals in this kind of data, if at all. My SPX setups this year aren't in a drawdown. The two setups I've used since last December (I switched part-way through the year) had four trades since December, not including the current trade I just started. Three were profitable and one wasn't. The results were: +0.5%, +17.6, -8.9% and +3.2%. (These were the results after I divided them in half because they were all 200% leveraged trades.)
The current setup is up 7% as of this week's open of trading YTD. It was in cash about 40% of the time and saw 23% less volatility than the market.
Regards,
Alex
Hi ALex
Wondering if the OIL brakout on Resistance - UGA $36 XLE $53 USO $39 was a fakeout. They all brk thru and I was pretty happy for that buy OIL took a great kick today and Im not sure why? Maybe the banks caused a flip of some cash to financials. I know on the TSX banks flew high today on BMO earnings and the rest of them report this week.
In short, are you still long OIL to Sept 4/7th ??
Hi Joseph,
Thanks for your message. I am still long. I don't second-guess my signals. That said, I'm the only that backtested them, so I have full confidence in them. Other people need to rely on their own judgement and systems.
Technically speaking, I don't believe USO or HOU (crude ETF in Toronto) have broken out on the daily charts. They've been testing resistance all this month pretty much, but a breakout in my opinion would be at the TDST line at $38.87 for USO, for example. I add a small buffer derived from the recent Average True Range - in this case $1.25. So I wouldn't buy it for technical reasons if I was trading on the daily chart until it hit $40.13. Not a recommendation. That's just how I do it.
Regards,
Alex
Thanks Alex, well OIL is bumping against its weekly PP at around $71 OIL just seems to have an issue busting thru $75... volume was low on the sell off in my view so lets see. I do rely on your data amongst my own data and other sources. Just OIL has bumped against $75 3 times now, wheres the breakout bro : )
Hi Joseph,
Just checking the charts. Probably not today!
Alex
well it seems the banks on the TSX will do well this week as BMO and CIBC have reported good earnings. BNS, National and TD are this Thurs/Fri. So my guess is Mon Aug 31st thru to the 1st week of SEPT OIL will rise, for now Im sure folks are moving some money out of Energy into banks... Crazy if you ask me but Banks are UP and OIL indexs are down, not that much down compared to the actual OIL,,,, USO etc..
OIL at 70.50 needs to hold
Alex:
I noticed that the open interest in many equity futures is decreasing rather quickly. The large SPX and Nasdaq100 have had dramatic falls.
Have you done any studies on OI and how it reflects the market? I presume a declining OI means that market participants feel that the market is at a fair value.
Hi Aviat,
I have studied total open interest for trading signals in the S&P 500 and the other markets I'm covering. In some cases, that's a useful indicator, but in the S&P 500 it wasn't the best one. The small trader total open interest gave the potentially most robust historic signals, but not as good as the ones I ended up choosing for my setup. Whether total open interest of the small traders should be faded or not depends entirely on the signal parameters (i.e. standard deviation and moving average period).
Take care,
Alex
I have heard that declining open interest near a rollover can precede a trend change. I have no idea if it's true, I'm curious if anyone has any information about that.
Hi Michael,
It's an interesting question. I haven't studied total open interest of all trader groups or rollovers. But in the case of individual trader groups, the only one in which the total open interest had even limited use for trading signals was the small traders. And in that case, everything depends on the parameter values. One set can be faded, while another set should be traded alongside. So I'd be skeptical of a blanket statement like that unless the person who made the claim comes up with some actual serious backtesting (including stuff like detrended price data, Monte Carlo testing, walk around testing, out-of-sample testing). I don't mean just eyeballing a chart. Otherwise, you might as well roll a dice.
Take care,
Alex
Post a Comment