Tech Analysis of ETFs??

Thank you everyone for all the response. It appears to be a near consensus that a chart is a chart regardless of the subject. It still seems counter-intuitive to me.

To use japanese candlestick as an example:
a given setup depicting a tug of war between the bulls and bears within a given single stock could be stumbled upon with a random basket of stocks that produce an ETF, and represent nothing of the sentiment attributed to the candlestick setup, merely coincidental math. So how could the subsequent result of a such a setup actually move in a predictable way? (Assume just for the purpose of this illustration that japanese candlestick analysis is reliable)

If ETF and Stock charts do in fact behave in the same way, then it is interesting to note that much of the discussion within Technical Analysis literature attributes chart behavior to occurrences and sentiments that are unique to and possible only within the context of single stocks. The purported *rationale* does not translate to the context of an ETF because the ETF is an average of logic and emotions across many contexts. Presumably this would make the sentiment ascribed to support or resistance, for example, an irrelavant concept to attribute to an ETF both in the intention and the nomencalture.

No?
 
I mentioned this concept of ETFs vs Single Stock to my wife. She has a PHD in a statistics field and is able to relate the concept to the protocols of deriving information from disimilar data sets. She said it is laughable to arrive at a similar conclusion when comparing apparently similar charts, one from across a pure single source, and one from across a random set of multiple sources.

In other words, to compare the charting actions of Cisco to an ETF divided in thirds of a Chinese shipping company, a Mexican telecom, and a Canadian mine makes absolutely no sense. The data of the ETF in this extreme example would basically define random.

The above example is no doubt extreme. A set of all American stocks might better align the charting activity of the different stocks by reacting to Wall Street overall, but the randomness persists, it is just a little less so. The point here is to highlight the presence of the dirty data.

Perhaps it is actually true that both single stock and ETF charts can be interpreted in the same way, but when I try to work backwards and deconstruct how this could be I am at a loss.
 
Quote from canslim:

I mentioned this concept of ETFs vs Single Stock to my wife. She has a PHD in a statistics field and is able to relate the concept to the protocols of deriving information from disimilar data sets. She said it is laughable to arrive at a similar conclusion when comparing apparently similar charts, one from across a pure single source, and one from across a random set of multiple sources.

In other words, to compare the charting actions of Cisco to an ETF divided in thirds of a Chinese shipping company, a Mexican telecom, and a Canadian mine makes absolutely no sense. The data of the ETF in this extreme example would basically define random.

The above example is no doubt extreme. A set of all American stocks might better align the charting activity of the different stocks by reacting to Wall Street overall, but the randomness persists, it is just a little less so. The point here is to highlight the presence of the dirty data.

Perhaps it is actually true that both single stock and ETF charts can be interpreted in the same way, but when I try to work backwards and deconstruct how this could be I am at a loss.
More...
I think you lost your way somewhere in the woods. Your original observations were simple and related to how to make money using an ETF. This other stuff is just mental gyrations and will take you dangerously into an analysis paralysis mode, very nasty basis for trading.

If you are going to use TA, find a way to use TA that makes you money. Regardless of what others say, it CAN be done, and consistently so, and forget about all the "why" and worry of "randomness" stuff. Leave that to non-trading analysts and statisticians (no disrespect intended to your wife).

Traders need actionable information, a narrow enough focus to pull the trigger in AND out of a trade, and sufficient common sense to not let our all too human emotions interfere.

If you "believe" in TA, YES, you do interpret TA on a CSCO or ETF or penny stock the same way, but be very careful that you use sufficiently uncorrelated TA indicators so that your TA approach, especially for thinly traded instruments, is robust. I'll repeat what I mentioned before <blockquote><b>You've answered your own question. That the underlying stocks don't necessary track the ETF is the most important thing to remember if you trade stocks based on indices or ETFs. People get people into trouble when they conclude that "the market" (or in this case, an ETF) is strong, so I'll buy a particular individual stock in anticipation that it will rise with the tide. The TA you did on the ETF applies only to that ETF, not the underlying issues. </b></blockquote> From this perspective, your observation about CSCO and the chino-shipping-mining ETF of course makes sense, even common sense. But that does not negate the use of the underlying math. The math (ie, TA) isn't at fault, but the illogical deductions of the analyst are.

Finally, regarding just what TA to use. It makes for analysis paralysis to have 5 panes of Moving Average based indicators under a chart. Try a mix of moving average (use one of the following: RSI, MACD, Stochastic, CCI or variations on them), volatility (ie, range based such as ADR, Bollinger, Kelter, or Donchian Bands), and volume-based indicators. Once you master a simple combination of only three or four indicators, and you find it suits your temperment, only then expand into more exotic ideas.

<a href="http://stockcharts.com/education/Overview/index.html">Here is a simple tutorial on the various types of stock analysis</a>. Again, keep things simple and don't invite over-analysis into your mind.

best, amg
 
Quote from canslim:

Is it correct to conclude that technical analysis of individual securities is a completely different science than for ETFs?
More...

In trading and charting you will find significant differences. The most apparent is short term volatility.
 
Quote from amg:


If you are going to use TA, find a way to use TA that makes you money.
More...

My observation is that there are more systems that will make you money on individual stocks than on an ETF. Now that's just MY observation. I know some people do just great with ETFs. But what I've noticed is that - again this is imho - it is harder to find systems that work well with an ETF...
 
shoeshineboy, I'd have to agree that in my experience, directly trading an index ETF "long term" (ie, swings greater than a week in length), while doable and profitable, didn't make me as much money as directly trading a stock.

To get to that point, however, I'd still do a top down approach, ie, look at a strong ETF of interest, probably using FA -- fundamental analysis-- then TA, then drill down into its components and select stocks that correlate well with the ETF, then use specific TA signals on the final stock selections.

The pitfall to avoid is to "assume" anything.
 
your wife's initial logic SOUNDS reasonable, but falls apart upon analysis

the reason that the same GENERAL TA principles apply to ETF's and individual stocks (or commodities) is that ETF's represent proxies for supply/demand just like any product in a 2 way auction market does - whether it's corn, gold, MSFT, UTH, etc.

also, i am sure she doesn nto know (but you should) that most of a random stocks' movement is attributable to the market as a whole, a smaller part is the sector it is in , and the smallest part (again, for an AVERAGE stock) is the actual stock itself

if there is massive demand for stocks IN GENERAL, and for TRANSPORT stocks in particular, then a transport stock will be more likely to go up than a utilities stock, ceteris paribus

similarly, laws of supply and demand apply to IYT just like they apply to an individual stock

INSTITUTIONS buy stocks in baskets .. also, since people TRADE etf's, the laws of supply and demand are just as relevant.

when people want to buy stocks (institutions and individuals), they buy stocks qua stocks, but they also buy baskets of stocks, like SPY or S&P futures.

so, of COURSE TA applies

knowing statistics without knowing how to apply them to a 2 way auction market is completely useless, as her analysis shows
 
Glance at the topic "beta" and you can reformulate the relationship of stocks, ETF's and indexes.

Then there is the "making money" factor.

The prioities for making money do vary with the approach.

In this thread fear and risk dominate so far.

Fear comes primarily from lack of knowledge and subsequently skill and risk comes from adding risk to an approach for some reasons by choice.

ETF's are part of markets as instruments as a way of capturing and pooling money. Does a trader need to have his capital captured?? What is the basic cost he pays for this submission?
 
Quote from amg:

shoeshineboy, I'd have to agree that in my experience, directly trading an index ETF "long term" (ie, swings greater than a week in length), while doable and profitable, didn't make me as much money as directly trading a stock.

To get to that point, however, I'd still do a top down approach, ie, look at a strong ETF of interest, probably using FA -- fundamental analysis-- then TA, then drill down into its components and select stocks that correlate well with the ETF, then use specific TA signals on the final stock selections.

The pitfall to avoid is to "assume" anything.
More...

You mean TA of an ETF? If you had success with that, my hat is off to you. I've never had any luck with it and I think it's cuzz of going head-to-head with the big boys...

There's another great difficulty with ETFs in my opinion from what I've seen: what ever works for five years does not work for the next five years and for the five years after that, etc. Large markets delight in pulling the carpet out from under you just when you think you have them figured out. Now I'm sure that's not true - I know some people do very with ETFs, but I'm just saying what I've observed. I've never found a system that I thought would be reliable on ETFs across the decades if you will...

And while I'm on a roll, here's another issue with ETFs: they are much more difficult to backtest because the sample size is so small usually. How can you feel comfortable, for example, if you backtest the SPY and Q's for the last ten years. Two issues won't give you enough backtested trades (in swing trading at least) to giveyou any comfort level.

Again, I know people do well swing trading ETF's, but I'm just pointing out there are definite obstacles to overcome...
 
TA is a very broad statement. To say "I've had no luck with TA" or "I've had success with TA" really doesn't mean much. No offense, but there are so many variables in those statements.
 
You must log in or register to reply here.
Back
Top