Quick note on the daily charting for longer term trading of the ETFs.
The bull ETF moving averages seem to correspond reasonably well with the same period averages on the SPX chart. The ber funds, on the other hand, vary widely. It appears that the longer the period of the moving average the less correlation it has. This makes using them on the bear fund charts impossible to use. I realized some time ago that trading ETFs really should be based on the underlying index chart and not the ETF chart anyway but this is more important on the daily scale.
On the minute scale the averages are not bad and can be used, loosely, on any chart for comparison but the index should still be the main reference.
This brings me back to the trial trade in the last post. My stop of $56 on SDS was based on monthly pivot points calculated from that chart...I should have been using the index especially as, unlike day trading, a daily chart needs no instantaneous updates to determine trade management.
So, looking at the charts I see that I should have bee stopped out on SDS yesterday, even according to the SPY chart had I bee shorting I would have covered as my stop would have been based on the 200 DMA line and it was breached in both cases as the end of bay...and end of month rush, spiked SPY over a dollar, SDS dropped a bit more than that. That last push would have stopped me out, which would not have been in the plan.
So going back to the SPX chart it peaked at 920. Still short of the 200DMA at 928.60 and short of the recent high of 929.20.
Another consideration is that longer term trades really should be taken on non-leveraged funds as the leverage effect loses money over the long run as a price oscillates from day to day particularly in a sideways market move. Nice straight trending is fine but that is not always the case. In fact, a good long long term plan might be to short a bull fund and hold it as a hedge as, over time, it will continue to drop in peak value until it reaches the point where the fund manager decides to consolidate the units to bring the value back up to a workable level.
Anyone wanting to see this just run a chart with an index and the leveraged bull fund side by side or overlayed. Each successive peak of the fund is slightly lower than the peak of the index, relative to all previous peaks as the fund gets rebalanced each day whereas the index does not. I think that shorting a leveraged bear fund would result in the same effect but I haven't looked closely at it. Had I access to shorting ETFs (no shorting in my registered accounts) I might just short a handful of shares and hold them to see how effective this might be in reality.
I digress.
The point is that my SDS trade would likely be an SH (proshares S&P500 short...non-leveraged) at $66.92. Current closing price is $65.82.
That compares to $55.50 in SDS down from $57.71. Thanks to leverage.
So, we shall see where this goes. I will consider entering the next trade setup for longer term in my TFSA account. I probably will not trade in even lots so I cannot use a true stop loss, but that hardly matters as I intend to be day trading so I can easily monitor the position and close the trade at any time during the day...which also lets me aim for better prices than just the previous day's close. Perhaps even try out a VTSO to enter the trade...I wonder if that will work...
Showing posts with label ETF. Show all posts
Showing posts with label ETF. Show all posts
Saturday, May 30, 2009
Saturday, May 23, 2009
Transferring the TICK to the dailies...EXPERIMENTAL
Well, my longer term outlook was pretty close so far as the S&P500 has slipped a bit. I don't think there is a huge concern for a sudden drop off quite yet but it may be coming. There is not much more good news that can be disseminated that will push the market hard enough to stop the almost inevitable drop.
Here is the last six months on a daily chart for the SPX.
As the value approached the 200dma it lost steam and started a small drop. I have not been watching the S&P500 for long enough to know how it moves so I would not be able to call the short entry based on the chart and the price proximity tot he moving average, so let's take a look at an unorthodox method that I have started using for day trading and apply it to this timeframe and see what happens. It is worth noting that I have not managed to use this successfully in my daytrading as it is still pretty fresh and I am doing some fake trading yet. I tried it on Friday but a tight ranging day requires some different angle than what I am using here. I haven't nailed tight range trading yet.
Here is the chart of the S&P500 with the latest long term trading noted.
The stop line is based solely on the price passing the next monthly pivot point, support or resistance then moving the stop price up to just below that point. One case I used the 50SMA as the pivot point for April dropped and I don't think I would drop my stop setting to suite, just use the next reasonable resistance line.
Now this is the kicker, here is the daily TICK chart that was used to set these trades up...in hindsight right now but I will start tracking as if I did place the trades in "real time"...or as close to that as I can get with daily chart bars.
SMAs are red = 5, blue = 10, green = 15 and yellow = 20...the dotted are somewhere in between and are not really necessary.
The 30 / 50 crossover needs to be confirmed by some upward (or downward for shorting) trending no matter what the price chart looks like. I have run these in backtests by marking the minute TICK chart and only afterwards referring to the price chart. Using the TICK data for exiting the trade is not as good as using the stops, so far as I can tell now. Interesting to note, and something I didn't expect, as the TICK indicated it was time to get short I may still be in the long position...depending on how tight my stop was set to the 1/2 R1 line. According to my chart I would still be in. Using ETFs is an advantage here as I can enter a regular one, SPY, SSO, to play the long trade and use a short ETF like SDS for the downside. Simultaneous long and short positions which are not possible in a single issue...although I have multiple accounts so I could manage this if I REALLY wanted to.
I think that the TICK trending is a very strong indicator for the S&P index and comes very close to being a leading indicator. I am working on a method of using this in conjunction with some sort of sector rotation to be in the ETFs that are moving the most consistently. That may take some time to get going though. One project at a time for now.
In order to track these trades I would, for now, use the previous day close price as the next day entry price. So for the trade entry on the 19th I will use the 18th closing price of $57.71. My stop will be just above the monthly R1 point...taking this over to SDS, well, this would be not as easy to manage as the price and averages for bear funds are not the same as they are for bull funds, leveraged or otherwise. Shorts are easier. I'd say the stop might be about $56.00 for now.
Enter = $57.71 ( I might have set a lower limit order to get in expecting a down day though)
Stop = $56.00
Last close = $60.54
For curiosity I checked the gain from SPY for the long trade.
March 11th entry at $72.17
Current stop at about $88.10
Nice tidy little paper gain. I may run some of these in one of my accounts to see how it goes...only I have to wait for the next good setup before placing a trade...the whole reason, or most of it, for me going back to day trading, I didn't have the patience to wait for a setup to setup on a daily chart. Maybe use part lots and not worry about the lack of stop loss orders for the smaller risk profile as a result. Just enough to prove a live system.
Jeff.
Here is the last six months on a daily chart for the SPX.
As the value approached the 200dma it lost steam and started a small drop. I have not been watching the S&P500 for long enough to know how it moves so I would not be able to call the short entry based on the chart and the price proximity tot he moving average, so let's take a look at an unorthodox method that I have started using for day trading and apply it to this timeframe and see what happens. It is worth noting that I have not managed to use this successfully in my daytrading as it is still pretty fresh and I am doing some fake trading yet. I tried it on Friday but a tight ranging day requires some different angle than what I am using here. I haven't nailed tight range trading yet.
Here is the chart of the S&P500 with the latest long term trading noted.
The stop line is based solely on the price passing the next monthly pivot point, support or resistance then moving the stop price up to just below that point. One case I used the 50SMA as the pivot point for April dropped and I don't think I would drop my stop setting to suite, just use the next reasonable resistance line.
Now this is the kicker, here is the daily TICK chart that was used to set these trades up...in hindsight right now but I will start tracking as if I did place the trades in "real time"...or as close to that as I can get with daily chart bars.
SMAs are red = 5, blue = 10, green = 15 and yellow = 20...the dotted are somewhere in between and are not really necessary.
The 30 / 50 crossover needs to be confirmed by some upward (or downward for shorting) trending no matter what the price chart looks like. I have run these in backtests by marking the minute TICK chart and only afterwards referring to the price chart. Using the TICK data for exiting the trade is not as good as using the stops, so far as I can tell now. Interesting to note, and something I didn't expect, as the TICK indicated it was time to get short I may still be in the long position...depending on how tight my stop was set to the 1/2 R1 line. According to my chart I would still be in. Using ETFs is an advantage here as I can enter a regular one, SPY, SSO, to play the long trade and use a short ETF like SDS for the downside. Simultaneous long and short positions which are not possible in a single issue...although I have multiple accounts so I could manage this if I REALLY wanted to.
I think that the TICK trending is a very strong indicator for the S&P index and comes very close to being a leading indicator. I am working on a method of using this in conjunction with some sort of sector rotation to be in the ETFs that are moving the most consistently. That may take some time to get going though. One project at a time for now.
In order to track these trades I would, for now, use the previous day close price as the next day entry price. So for the trade entry on the 19th I will use the 18th closing price of $57.71. My stop will be just above the monthly R1 point...taking this over to SDS, well, this would be not as easy to manage as the price and averages for bear funds are not the same as they are for bull funds, leveraged or otherwise. Shorts are easier. I'd say the stop might be about $56.00 for now.
Enter = $57.71 ( I might have set a lower limit order to get in expecting a down day though)
Stop = $56.00
Last close = $60.54
For curiosity I checked the gain from SPY for the long trade.
March 11th entry at $72.17
Current stop at about $88.10
Nice tidy little paper gain. I may run some of these in one of my accounts to see how it goes...only I have to wait for the next good setup before placing a trade...the whole reason, or most of it, for me going back to day trading, I didn't have the patience to wait for a setup to setup on a daily chart. Maybe use part lots and not worry about the lack of stop loss orders for the smaller risk profile as a result. Just enough to prove a live system.
Jeff.
Saturday, February 14, 2009
The ETF angle for medium term trading
I see a ton of interest in ETFs lately and specifically oil related ETFs, Horizons BetaPro NYMEX Crude Oil Bull and Bear Plus ETFs, HOU and HOD respectively. The particular ETF or index has not really interested me unless they are in a certain price range, have ample intraday volatility, enough liquidity and have an index that I can track in real time...this all for day trading purposes. So in this regard I really don't care too much which one I am using.
Having said that this particular interest caught my attention and leads me to check into ETFs a little closer.
Link to my post about ETFs from my daytrading blog.
Another link with more particulars and other links to Horizons BetaPro ETFs.
I took a table from the Horizons site, exported it into a spreadsheet and ran a few numbers to get a feel for how the ETFs seem to behave. I figured out some ratios to see a few relationships. Now these may not mean much as a single day snapshot so I may have to track them for a bit but they are interesting none the less.
RELATIVE VOLUME (Relative to the outstanding shares)
To get an idea of the interest in a particular fund I divide the volume for the day by the outstanding shares. I figure that the closer to 1 this number is the higher the trading interest, this is relevant more to compare the bear vs the bull of the same ETF than comparing anything between the various ETFs, although higher interest ratios can lead to at least better liquidity. Interestingly the bear crude oil ETF, HOD, was ahead as it traded more than the number of shares in circulation by a ratio of 1.57:1.
Top five in relative volume:
1.57 - Crude Oil Bear (HOD)
0.48 - Natural Gas Bear (HND)
0.43 - Crude Oil Bull (HOU)
0.31 - Global Gold Bear (HGD)
0.18 - S&P 500 series Bear (HSD)
I'm not sure if this really means anything but the rankings were interesting to note.
PERCENTAGE CHANGE CORRELATION
This shows that the inverse funds track well together...when one goes up the other goes down a corresponding percentage. I think that this is a very important factor in trading these funds as a bear that wildly varies in the inverse relationship from the bull creates an unpredictable result when trying to trade them based on the index that they are supposed to be following...they cannot be tracking well and with less correlation, the worse they are at that.
I used the percentage price change of the last day between the bear and bull of the same fund. I divide the increase of one by the decrease of the other then also calculate the decrease by the increase. If they are tracking well I would expect both numbers to be close to 1. They were not all, so to make the results easier to see quickly I just use the difference between the two calculations. The closer to zero this number is the closer the overall correlation.
The top five in percentage change correlation:
0.02 - TSX 60 series (HXU/HXD)
0.07 - Global Gold (HGU/HGD)
0.08 - S&P 500 series (HSU/HSD)
0.13 - Financial series (HFU/HFD)
0.29 - Crude Oil (HOU/HOD)
MEAN VARIANCE
The last number that I noted was the Mean Variance or how far away from the mean of the two funds, bear and bull, the current prices were as a percentage of the mean. This sort of falls into the category of Reversion to the Mean as a method of determining the probability of a trade.
I think, but have not thought long on this, that the mean might represent the underlying index if it were represented as a number relating to the two derivative funds and charted over time. In theory I would expect that the funds would track equally on either side of this number...but they won't in practise I am sure. So it may serve as a snapshot indication of the current state of the price only. I would like to see a large variance to consider a larger likely price change to return to the mean...which should happen at some point.
The top five in wide mean variance:
88% - Natural Gas (HNU/HND)
80% - Financials (HFU/HFD)
76% - Crude Oil (HOU/HOD)
70% - S&P 500 (HSU/HSD)
69% - NASDAQ 100 (HQU/HQD)
So, after playing with these numbers it is interesting to see who has the best showing in the top five lists, I suppose that I could have easily guessed this answer without doing any of these calculations. Having said that, it suggests that perhaps my spreadsheet may be useful in the future when it is not quite so "cut and dried".
Crude Oil. Both the bear and bull funds made the high relative volume list and placed in the change correlation and wide mean variance lists as well.
I was surprised to see the S&P 500 show on all three lists as well. I think I may have a closer look at these fund offerings to see if they match my other normal criteria.
Do note that these are not recommendations for trading any of these funds even though I will probably check them out closer and perhaps even do some trading on them myself...afterall, that is the only real way to prove an idea one way or the other.
Jeff.
Having said that this particular interest caught my attention and leads me to check into ETFs a little closer.
Link to my post about ETFs from my daytrading blog.
Another link with more particulars and other links to Horizons BetaPro ETFs.
I took a table from the Horizons site, exported it into a spreadsheet and ran a few numbers to get a feel for how the ETFs seem to behave. I figured out some ratios to see a few relationships. Now these may not mean much as a single day snapshot so I may have to track them for a bit but they are interesting none the less.
RELATIVE VOLUME (Relative to the outstanding shares)
To get an idea of the interest in a particular fund I divide the volume for the day by the outstanding shares. I figure that the closer to 1 this number is the higher the trading interest, this is relevant more to compare the bear vs the bull of the same ETF than comparing anything between the various ETFs, although higher interest ratios can lead to at least better liquidity. Interestingly the bear crude oil ETF, HOD, was ahead as it traded more than the number of shares in circulation by a ratio of 1.57:1.
Top five in relative volume:
1.57 - Crude Oil Bear (HOD)
0.48 - Natural Gas Bear (HND)
0.43 - Crude Oil Bull (HOU)
0.31 - Global Gold Bear (HGD)
0.18 - S&P 500 series Bear (HSD)
I'm not sure if this really means anything but the rankings were interesting to note.
PERCENTAGE CHANGE CORRELATION
This shows that the inverse funds track well together...when one goes up the other goes down a corresponding percentage. I think that this is a very important factor in trading these funds as a bear that wildly varies in the inverse relationship from the bull creates an unpredictable result when trying to trade them based on the index that they are supposed to be following...they cannot be tracking well and with less correlation, the worse they are at that.
I used the percentage price change of the last day between the bear and bull of the same fund. I divide the increase of one by the decrease of the other then also calculate the decrease by the increase. If they are tracking well I would expect both numbers to be close to 1. They were not all, so to make the results easier to see quickly I just use the difference between the two calculations. The closer to zero this number is the closer the overall correlation.
The top five in percentage change correlation:
0.02 - TSX 60 series (HXU/HXD)
0.07 - Global Gold (HGU/HGD)
0.08 - S&P 500 series (HSU/HSD)
0.13 - Financial series (HFU/HFD)
0.29 - Crude Oil (HOU/HOD)
MEAN VARIANCE
The last number that I noted was the Mean Variance or how far away from the mean of the two funds, bear and bull, the current prices were as a percentage of the mean. This sort of falls into the category of Reversion to the Mean as a method of determining the probability of a trade.
I think, but have not thought long on this, that the mean might represent the underlying index if it were represented as a number relating to the two derivative funds and charted over time. In theory I would expect that the funds would track equally on either side of this number...but they won't in practise I am sure. So it may serve as a snapshot indication of the current state of the price only. I would like to see a large variance to consider a larger likely price change to return to the mean...which should happen at some point.
The top five in wide mean variance:
88% - Natural Gas (HNU/HND)
80% - Financials (HFU/HFD)
76% - Crude Oil (HOU/HOD)
70% - S&P 500 (HSU/HSD)
69% - NASDAQ 100 (HQU/HQD)
So, after playing with these numbers it is interesting to see who has the best showing in the top five lists, I suppose that I could have easily guessed this answer without doing any of these calculations. Having said that, it suggests that perhaps my spreadsheet may be useful in the future when it is not quite so "cut and dried".
Crude Oil. Both the bear and bull funds made the high relative volume list and placed in the change correlation and wide mean variance lists as well.
I was surprised to see the S&P 500 show on all three lists as well. I think I may have a closer look at these fund offerings to see if they match my other normal criteria.
Do note that these are not recommendations for trading any of these funds even though I will probably check them out closer and perhaps even do some trading on them myself...afterall, that is the only real way to prove an idea one way or the other.
Jeff.
Subscribe to:
Posts (Atom)
