Feb

12

Any reader who has not looked at a price chart in the past 90 days please stand up and identify yourself. For that person and that person alone can cast a stone (at technical analysis).

Gary Phillips writes: 

I look at charts all the time, but that's really not the point. For someone who is as truly blessed with the ability to determine causality as yourself, you must realize that charts are not predictive in of themselves.

Larry Williams writes: 

Parts of charts are most definitely predictive. Patterns repeat. And I agree that so much of TA is misleading and based on whims and fancy yet there are parts that really do work.

anonymous writes: 

Ah, the chart debate has returned.

While surely an example of survivor bias, I have witnessed industry greats use charts and technical analysis as part of their speculative arsenal. Of more interest is that these people used their own personally derived versions of these methods and not the versions available at no cost to everyone. I dare say that the creators of well known indicators have ways of using them that they would never reveal (rightly so!).

A few points about charts:

1. At the higher frequency end, in the OTC macro markets, ALL of the chart services are wrong and ALL of the chart services are correct. Each has its own price, so there is no 'right'. This probably doesn't matter to most and doesn't fatally damage the pro chart school.

2. Some market extremes are written out of history for various reasons (regulatory, legal, error, political correctness and vested interest). The move toward full electronic trading might alleviate some of these in future.

3. Commodity prices on charts…. Should we adjust them by inflation? What are we actually looking at? What are we comparing.

4. Equally spaced data? What to do with price action measured in equal intervals (say, for example, 5 minute charts) when the price doesn't change during the period but the recording software has to put a number in there so it averages, uses the last price, the first price of the next period etc…

5. There is a reason why the big quant firms have interesting individuals whose life's passion is ensuring data is clean/ accurate.

6. It is probably a fair point to state that the recording of price information has improved since, say, the 1970's. The tricks now are more to do with latency of its delivery and the subtle recursive methods some providers appear to use to set their lows and highs. As an example, watch EURUSD spot today if you have something approaching Direct Markey Access and if you watch closely enough you may note that the high as printed on your screen (for eg.) sometimes moves higher a few seconds after the price has actually moved lower. A less charitable person than I would suggest it was to ensure all the stops on the banks' electronic platforms could be said to have been done within 'the range' ( whatever that is ). I guess it might just be an optical illusion generated by my mind's inability to accept being stopped at the high. Ha!

SideBar on this last thing– one great method market makers employ to get stops done is to drastically widen their spreads when near stops. ( Much small print allows stops to be done if inside the spread for ' risk management' purposes ). This may go some way to explaining the mystery of the changing highs/lows after the fact….

John Bollinger writes in: 

I don't understand. If charts aren't predictive why in the hell do you all waste your time looking at them? Do you have so much time on your hands that you can engage in frivolous pursuits at work? If you gonna talk the talk, walk the walk. If you think charts aren't helpful, STOP LOOKING AT THEM.

Rocky Humbert writes: 

While I am in agreement with the inestimable Mr. Bollinger that looking at charts has utility, I would be cautious about the term "predictive."

When I go to the doctor's office, her nurse always takes my temperature. My temperature is not so much "predictive," but rather it is informational. In numerous ways, looking at charts are like taking a patient's temperature.

I wish I could claim credit for this insight, but I can't. It's from Bruce Kovner (who I still consider the best trader/investor from a risk-adjusted return perspective of the past 30+ years.)

Ed Stewart writes: 

It seems to me that body temperature is predictive of future temperature change do to homeostasis. The breakout from the range where homeostasis functions is going to be predictive of body temp = ambient temp if there is not a reversal or intervention.

Rocky Humbert replies:

Fair point. But you don't need to take a patient's temperature to know that EVENTUALLY body temperature = ambient temperature.

Keynes figured that out when he wrote that "in the long term, we're all dead." (See: JM Keynes "Tract on Monetary Reform, (1923) Chapter 3)

Kovner's actual quote was in reference to so-called fundamentalists who scoff at charts. He said, "Would you go to a doctor who didn't take a patient's temperature."

Gary Rogan writes: 

You don't need to take the patient's temperature nor to study medicine to know that eventually the body will assume ambient temperature, but there are clearly situations when the current temperature is highly predictive of the timing, barring an intervention. As such, this whole analogy and the corresponding point just don't work.

A more expanded quote by Keynes reads as follows: The long run is a misleading guide to current affairs. In the long run we are all dead. Economists set themselves too easy, too useless a task if in tempestuous seasons they can only tell us that when the storm is past the ocean is flat again. He was in fact arguing for short-term action based on predictions even though in the long run the economy will recover. So it a way it's almost the opposite point to what Rockstergeist indicated he was making.

Craig Mee writes: 

No doubt with the right risk management you can make money trading in many ways, but surely the best outcome is to not leave plenty on the table and have a lot of what ifs in the outcome, together with an ordinary win loss ratio while still banking a healthy return. In the pursuit of excellence, it doesn't seem winning and the above go hand in hand. Though possibly for others this isn't an issue, and probably quite rightly it's all about the bottom line. Hence the saying, "trade the way that you're comfortable with".

Gary Phillips writes: 

Considering the maelstrom of controversy and unchecked emotion the subject elicits, perhaps TA should join sex, politics, and religion on the list of banned subjects for this site.

John Bollinger replies:

Careful, the site will become very quiet as the best part of what is discussed here is technical analysis in one way or another as a survey of the literature will confirm.


Comments

Name

Email

Website

Speak your mind

6 Comments so far

  1. Anekdoten on February 7, 2015 1:49 am

    Charts can most definitely be predictive, and even with a high degree of accuracy, just not all the time as they are populated with periods of uncertainty. Patience is a virtue and most definitely in trading.

    Most don’t even know when or what to look for.

    Truth is those who find no value in charts just failed as chartist, and it’s easier and less embarrasing to say “they don’t work” than “I did not get it”.

    Anek

  2. Jeff Watson on February 7, 2015 5:34 pm

    Exactly how does the past predict the future, and how do you quantify your anticipated predictive results?

  3. Bill Wolfe on February 7, 2015 6:50 pm

    I often wonder if chart naysayers ever took the time to try to learn chart reading, instead of just looking at a chart and proclaiming they don’t work.

    On my site I actually give the rules for reading the chart properly. www.wolfewave.com.

    Bill Wolfe

  4. Arne on February 9, 2015 6:43 am

    Reply to ‘Jeff Watson on February 7, 2015 5:34 pm’

    Was there any way one could’ve predicted whether or not these two people had the intention to walk up those stairs, when the photograph was taken?
    http://previews.123rf.com/images/keithlevit/keithlevit0801/keithlevit080102472/2428693-vancouver-british-columbia-people-walking-towards-stairs-in-vancouver.jpg

  5. Spekulatn on February 12, 2015 1:01 am

    Would someone please pick a chart, any chart and show how it’s predictive?
    Thank you.

  6. Edward Lam on February 12, 2015 5:07 am

    Can I suggest that using charts (pattern recognition) is similar to some versions of algo trading? In both cases the patterns are signals, in both cases you don’t need to know why the signal is a signal, and in both cases you know it works but you don’t know when it will fail until it does…(although why it fails is simple: the Ever Changing Cycles)

Archives

Resources & Links

Search