Technical Analysis and Common Sense

[rs7]

Originally posted by Kymar


More generally, believing that price action can in fact be accessible to timely technical analysis and trading is much different from believing that any particular indicators, methods, or systems really are likely to work over any extended period.

Well said!!!
BTW, higher highs and higher lows (and the converse) as mentioned in the post previous to Kymar's are very important IMO. So it is essential to understand TA and what is generally reliable over an "extended period". It is the intraday oscillators that I generally have a problem with insofar as being useful on an intraday basis. But, to each his own.
Good Luck and Kymar, keep 'em coming. I always enjoy your posts more than most!
RS7

 

[Traden4Alpha]

Many of the postings allude to winning with indicator X or discovering the indicator Y does not work. But, an indicator is NOT a trading system. Unfortunately, most of these postings give too few details about exactly how the poster tried to use the indicator. The problem is that trading with an indicator means: 1) choosing a tradable (unless their really is a Holy Grail that works in every market); 2) choosing a particular data sampling period (e.g., 1-minute bars or end-of-day data or ....); 3) choosing some magic values for the parameter(s) of the indicator (i.e., most indicators has one or more parameters that control smoothing, timeframe, or sensitivity); and 4) defining the rules for triggering trades (which may involve confirmations from the price action or other indicators).

Without knowing the answers to the first three factors, one cannot even replicate the indicator trajectories that the trader was looking at. Without the fourth factor, one cannot replicate the trades. This fourth factor is quite complicated and lets a skilled technician find profits were the amateur finds misery. Possible ways of interpreting the indicator to trigger trades include putting on a trade when: the indicator hits some magic value; the indicator goes past some value and then returns back to that value; the indicator crosses its own moving average; the indicator curve parallels or confirms the price pattern; the indicator diverges or disconfirms the price pattern (with further variation depending on whether you watch the lows or highs of both price and indicator values), etc.

Although some books on TA will give popular or suggested values for the parameters and trading rules, few stubbornly insist that theirs is only one right set of parameter values or one right way to generate trading signals from the indicator. The problem is that each of these different approaches is a different use of the indicator. A failure to trade profitability with one application of the indicator is not a failure of the indicator. So if you want to say that an indicator "works" or "doesn't work", then you must describe exact HOW you used it.

BTW, excellent discussion!

-Traden4Alpha

 

[DblArrow]

Originally posted by dotslashfuture
Dblarrow,

I looked at that picture and I just didn't see a good enough correllation to make me want to bet money. For me, it is just too noisy. But that is just my perspective, I'm not presenting it as a fact.

One mans noise is another mans trading oppurtunity!!

I agree with those other comments about noise. Somebody somewhere thought that each price was a place to buy or sell. Some long term, some short term. Not really noise at all - just different trading oppurtunities.

Make 'em pretty, Chris

 

[sempai]

Originally posted by rs7
So yeah, I agree that a ''system" can work for a while. I don't believe anything works for any significant length of time.
I said here before, it is CRUCIAL to recognize patterns in the market. When you see something happening with any regularity, you MUST take advantage of that observation. This is not limited to TA. For example, if you notice that the market always seems to dip or run at a certain time of day, take advantage. If you notice ANTYTHING AT ALL, use that observation. I know a lot of you play poker. What seperates the good players from bad is certainly not the cards. Everyone has an equal chance there. The good players have discipline. The better players notice what we call "tells". If you know what I am talking about, then think of the market giving you "tells" on what it will do. If you can glean repetitive behavior from charts, or oscillators, or moving averages, or relative strength, or volume or any combination, these are the real "tells". As are many other characteristics of the market. Over time, you will begin to recognize more of them and see them better and better. They are what you should be trading on. And they change. As do the "tells" in poker. People wise up, people change their styles, people adapt in poker. So whatever works there also changes. The very best poker players adapt. Sometimes it takes a setback to get them to change their perceptions. But the great ones do.


RS7

Thanks for the post, RS7 - and great thread too. This is something I've been working on lately; being able to recognize different market conditions and then adapting accordingly. Of course, this goes against a lot of the conventional wisdom of having a set plan and sticking to that plan.

I find myself going back and forth between mechanical and discretionary styles of trading and never having much success with either. Lately, I've been working on a more mechanical approach to the market - but the problem with this type of approach is that by the time I've worked out the "system", backtested and then papertraded it for a while, the market is already beginning to change.

Even as I'm papertrading this particular method I'm working on, I can see numerous other opportunities and patterns forming up in the market (what you call 'tells'). Often, they are just a slight modification of what I'm already doing, maybe on a five minute chart instead of the one minute or something like that. I just can't bring myself to trade them, because I don't feel I have a well thought out plan to trade them.

Any suggestions on how to reconcile this conflict between adaptability and having a sound trading plan?

 

[rs7]

Originally posted by sempai


Even as I'm papertrading this particular method I'm working on, I can see numerous other opportunities and patterns forming up in the market (what you call 'tells'). Often, they are just a slight modification of what I'm already doing, maybe on a five minute chart instead of the one minute or something like that. I just can't bring myself to trade them, because I don't feel I have a well thought out plan to trade them.

Any suggestions on how to reconcile this conflict between adaptability and having a sound trading plan?

I wish I could give you a simple answer. I think "trial and error" is about as best as I can give. Of course I realize this can be an expensive way to learn, so if you can just paper trade maybe that will work. The problem with paper trading is it just doesn't give a real picture because you never know what kinds of fills you will get. You need to simulate horrible fills and anything else you can think of that can and will go wrong.
Depending on your account size (buying power), maybe you can try very small positions to experiment with different strategies. This is what I have always done. I like to build up a level of confidence slowly in new approaches.
Good Luck,

RS7

 

[Traden4Alpha]

As rs7 points out, paper trading is different from real trading. Leaving aside the massive subjective issues of having real money at stake in real trading, the issue of fills also means that paper trading (simulations and backtesting) does not always give an accurate picture of how a trading system will behave. That is why I advocate both trading experiments and dual trading (both paper trading and real trading at the same time). The key is document the discrepancy between the actual trade and the paper trade to get a better understanding of the ways the fill occur for the instruments that you trade. Different experiments might look at how orders near or better than NBBO execute, what market orders do at the open, etc. I even do occasional experiments of submitting similar orders of different sizes just to see how size impacts the fill.

These experiments are often frustrating, but always instructive (the vagaries of fills, spreads, and slippage both randomize actual performance and undercut it). Yes, the experiments do consume some capital thru both commission costs and occasional under-performance. But I would rather get a good feel for how the market really acts than to assume that the market will behave exactly like it does in simulations or backtesting.

Parallel trading also helps you see if emotions are getting the better of you. By paper trading with discipline (very easy with no money involved) and real trading with emotion (Uh Oh!) you quickly see the impact of emotional decisions on P&L. You can learn about the true magnitude of losses from emotional trading. You might discover that even your winning "emotional" trades win less than a more disciplined approach. And, who knows, you may even find that your "gut" is actually better than your rules (lucky bastard!).

Good trading to all,

-Traden4Alpha

 

[CalTrader]

My take on classical technical analysis is that it is part of an overall trading strategy. Our models dont rely on classical technical indicators alone: we use a series of weights that represent market sentiment and behaviour in addition to purely classic technical indicators and these are adjusted manually each day. This extra review works better (for us) than models that are completely automated.

 

[mike s]

Here's a couple of quotes from The New Market Wizards that seem appropriate for this thread. It's Schwager interviewing Faulkner. Pg. 429.

Schwager: All these technical methods are based on price. In effect, they're all different-colored glasses for looking at price. Proponents of RSI and Stochastics (two popular overbought/oversold indicators) would see price patterns filtered through these price-derived series. Gann analysis enthusiasts would see the price patterns throught Gann-based interpretation. In these cases and others, traders accumulate experience on price patterns--albeit from different perspectives. Some of the methodologies employed, however, are probably totally worthless. It's simply that instead of looking at prices through clear glass, traders who use these methods are looking at prices through different-colored tints. The method, or tint shade, is a matter of individual preference. To extend the analogy, I would compare these methods to nonprescription sunglasses: they change the view but don't necessarily improve the vision. The bottom line is that these methods seem to work only because the people who use them have developed some sort of intuitive experience about price.

Faulkner: That actually fits pretty well with my own view. People need to have a perceptual filter that matches the way they think. The appropriate perceptual filter for a trader has more to do with how well it fits a trader's mental strategy, his mode of thinking and decision making, than how well it accounts for market activity. When a person gets to know any perceptual filter deeply, it helps develop his or her intuition. There's no substitute for experience.

 

[trader99]

Originally posted by mike s

Faulkner: That actually fits pretty well with my own view. People need to have a perceptual filter that matches the way they think. The appropriate perceptual filter for a trader has more to do with how well it fits a trader's mental strategy, his mode of thinking and decision making, than how well it accounts for market activity. When a person gets to know any perceptual filter deeply, it helps develop his or her intuition. There's no substitute for experience.

I think Faulkner's quote get to the VERY HEART of the matter in trading. It almost doesn't matter what indicators/technical chart patterns/T&S/L2, etc. technique you used as long as it helps you interpret price actions and get an intuition about things.

That's why traders who just buy blackbox trading systems have a hard time realize the returns of the backtested system(in additions to slippage, transaction cost, market impact, etc.)

So, even people who uses supposedly "unscientific" approach like astrology can in theory can still make $ if they use it consistently and with proper risk management.

hehe.


99

 

[Buy1Sell2]

Quote from DblArrow:



An indicator based on some mathamatical calculation of price will always remain the same. It will always react to the market in the same manner. Perhaps price action will change, traders perception of the market may change, or the length and breadth of trends, but the indicator will react to that change.

Very very good observations!