Hmmmmm..... could you be more specific on what an "old indicator" might be? Are we talking W.D Gann's Law of Twelve or Square of Nine from 80 years ago or J. Wells Wilder's RSI from 30 years ago? I'd be happy to provide a list of about 30 that do not work when used in objective rule based trading systems that do not involve human judgement, intuition, inspiration or decision making (objective statistical validity in predicting future market behaviour). Now if you are talking human based judgement I know this trader in Denver who just rubs a magic rabbit foot to get turns in the S&P.

TICK. TRIN.

Take ol' J. Wells and stop right there. Work the ol' ATR, ADX/DMI.

Help Lane's (so-called) Stochastic with a little volume-weighting, and you've got Wilder's RSI-with-a twist: Quong/Soudak's Mondy Flow Index. Primo!

With those 5, you've got volume, trend, bound-oscillators *and* intellectual incest.

(Who, really, could ask for more?)

Except RSI has no predictive ability when you statistically look at a few dozen instruments over a few decades. Over the long term is is essentially a random number in term of predicting future market characteristics. I'd stick with the rabbit's foot.

I would seriously advise you not to misapply any technical indicator.

NWR (Not worth a reply)

There are no indicators with predictive value or capability and never will be any. Bull markets tend to create such notions. Technical analysis is more art than science. Data processing power enables us to track incredibly large data sets on our own desktops. Of course it is easy to confuse the power, breadth and speed capability with the mistaken belief that predictive models can be built and applied. Amazingly access to the hundreds of sophisticated indicators seems to produce more paralysis by analysis and misguided assumptions than old style tape and chart reading. Basic blocking and tackling in the tradition of Magee and Edwards coupled with a decent grasp of the prevailing market mood and psychology continues to produce more gratifying results. Data processing simply allows us to cover more ground by automating the sorting process. Technical analysis is an art which resists conversion into pure science despite our herculean efforts.

Larry,

I would agree about the low value of indicators because as I mentioned all they do is blur and summarize the market..... a bit like trying to drive your cat at night wearing dirty smeared sunglasses. Also for 99% of the folks data processing is rather problematic. However, it is possible to use software to make predictions about the market successfully. It is very difficult and limited in terms of what can be predicted and for how long into the future.

Those with some ability in this endeavour are not likely to confirm this or post a how-to guide in a popular forum, either. So for all practical purposes in terms of general discussion we can conclude that it does not exist.

Restraint from using the term predictive is a lifetime habit. Maybe it is the compliance officer fragments embedded in my skull from a past portion of the life cycle. I do enjoy the witchcraft aspect of technical analysis and am admittedly captivated by the price pattern tendencies that seem to regularly emerge. Found your remarks entertaining as well as ringing quite true. Thanks for the pragmatic views. The market arena is increasingly awash in gibberish posing as scientific wisdom. As the bull market ages, more and more people are violating one of the cardinal rules of the business; never confuse brains with a bull market.

Well, so much for Algorithmic Trading, then.

Larry,

You are very observant in that a bull market does create illusions of great grandeur in many.. To really talk about predictive capacity we have be able to have a model that can function in all types of market conditions and in terms of daily bars covers several decades. Alternatively the model knows when to say "I don't have a clue" and turn itself off. Nothing predictive can deal with Black Swans. And to be clear when I user the term model or predictive I am talking about things that come out of Black Boxes and not the mind of some pontificating trading guru or algo programmer.

i will say with a specific combination of the common indicators you can create a sound strategy. check my market replay . http://ttraders.blogspot.com/2015/03/esf-030315-market-replay-explained.html

All strategies aside, this comes down to personal choice. Most traders operate in their own comfort space and tend to do reasonably well as long as they don't venture out and use someone else's strategy without understanding it.

In an uncertain market Algorithms tend to become dangerous as they don't pick up what mood the market is in. I agree the bull market does create its own illusions and caution is needed when you're using an Algorithm in such times.

If you spend enough time watching something, a pattern will emerge and you will see the pattern change depending on the mood of the market. It's all about the investment of your time into developing the understanding of the pattern (art) for the specific product.

How about the ultimate indicator - price and if you want to get fancy, throw in volume and for those non purists, I guess you might consider time (but I see that as sketchy at best)

I love these . . . discussions. G. Fredrick, I would love to have dinner with you and pick your brain. Rarely do I enjoy the posts of people with such market clarity. I know Scott and Larry would enjoy the conversation and the company as well. Scott and I have had this discussion a few times.

IMHO - Predictive analytics are a waste of time. The ultimate indicator is price . . . period. There is nothing more accurate than seeing the flow of price of filled orders. Add volume because it is another perfect product of that price action and you have all of the necessary components of the most beautiful reactionary environment one needs to read any individual market. Reactive analytics . . . RAH!

So we should all be thankful that there will always be people in this industry that disagree. I personally love all of the people that spend their lives trying to create algorithms or that analyze and group indicators to predict the direction of markets, stock or commodities. They create a method or system that will work for a day, week or month and then when it implodes they start all over again. I love economists. I love quants. I love all of the people that make what I love doing an option. As long as individuals are determined to analyze variable data to try to solve problems where the outcomes only benefit the specifics of the data and time analyzed, we all benefit. Their analysis only adds to the fuel our actions needs to burn.

I didn't realise we were including the obvious, I completely agree the number 1 indicator is price and following quickly is Volume.

Here is an example of how charting works, on Friday before the release of the new job numbers I posted a chart showing the price points for gold, should it start to fall. The points were 1190, 1182 and 1162. I then placed my buy order at 1164 (usually the lowest support doesn't always reach otherwise there is risk of it falling further, and for 1 day it's hard to fall more than a certain percentage).

Now, I did get lucky and the job numbers came in as they did, gold took a dive and I got my buy. I was reasonably pleased to be able to achieve that but I would not have been able to do it by guessing.

Now, here is another factor. Anyone that is experienced who wanted to opt in for the option of watching the price movement and then deciding what time to buy would have probably been able to pick a price similar to my price without doing the charting. However this is very dangerous, let me explain.

I have been trading for over a decade and I know if you really love trading, you will get sucked into it whether you like it or not. You just love what you do and it makes you get involved.

If I was watching the gold price move down, It is a fair assumption that I would have purchased when gold got down to 1170, I would have said to myself it's down over 2% and it's time to buy.

Charting does eliminate some of that need to jump in.

In summary, every trader has their best way to trade the market, I personally use a combination of both charting to determine my price point, and sometimes the price and volume (when I’m watching the market live). Don't forget not all people that trade sit in front of the screen and trade. The people that trade offline do all their research before hand, place their orders and then wait for the events to occur.

As William said, I love these discussions. In a discussion there is never right or wrong, everyone has their valid opinion. Talking and discussing something that we all have a passion for and gaining other people's opinions is real treasure.

Thanks!

@Daniel - Nicely put. I post the obvious because you would be surprised at the people that feel price/volume isn't a viable indicator for determining direction.

William,

how do you distinguish one volume signal as being better qualified than the other? Like OHLC, there are only so many combinations, like:

Hi Vol on longer than an ATR bar;

Hi Vol on shorter than an ATR bar;

Low Vol on longer than an ATR bar;

Low Vol on shorter than an ATR bar.

Unless convergence/divergence relative to price turns makes a better case, I'm not sure if Hi or Low Volume even on an ATR bar has any greater significance; but I could be wrong. :)

John,

Not entirely. RSI hasn't worked (defined as having objective statistical predictive power for future market conditions) since a few years after J.Wells Wilder invented it. That statement is limited to the 20 or 30 major market instruments and the 20 years I tested it on.

Hi,

An indicator is exactly what is says it is, an indicator.

When driving a car and you want to take a right (or left) turn, you switch on your "indicator". That does not mean it is safe to take the turn, you still have to watch out for other drivers, pedestrians, bikers ...

That being said, imho, an indicator is a tool you use in your decision making process. It is linked (when automated) to entry/exit rules and possibly money management.

To answer the question of this post, "Yes", we have created a consolidation indicator in 2002 that tells us when to be sidelined, when to enter a long (or short) position and when to exit the trade. It has a 70 % + success ratio on stock market indices and a 50 % + success ratio on FX (currencies) which is considered the most unpredictable of all markets.

The indicator does not predict. It just signals when a market moves from a consolidation (sideways trading period) to a trend, will point the direction to take (long or short) and stop loss rules are integrated into the model (not into the indicator) to avoid disaster.

The indicator will not predict a black swan event as this is by definition unpredictable.

What is seen however, is that large market moves are often being signaled in the price behaviour before the (mostly down) move occurs.

On a more general level. The more an indicator is known, the greater the chance it will lose validity. Why would anyone disclose a successful indicator anyway ?

kind regards

Marc

@Frank

Using volume as a signal is "piling on" to the already variable aspect of a typical chart. In other words, it would be adding to the inconsistency already embedded on a typical chart made up of tick, minute, range or other imbalanced bar type.

Individuals building a chart environment (GUI) do not see the value of building it the same way one would build a house. No contractor would ever build any structure where each individual piece of the foundation is drastically imbalanced or not level. Yet individuals building systems or methods using GUI's repeatedly construct the foundations of their hard work on bars overflowing with variable information forcing them to be imbalanced or not level.

I feel building algorithms from GUI's to be invaluable. The GUI to me acts as a microscope that I can adjust to fine tune the algorithm. To me consistency is critical and you can't get consistency from variable bar data. That is simple common sense from a physics standpoint. You can't find a cure for a virus using dirty lab equipment.

@William, I get the overall concept of the need for consistency. That, however, why mention price at all due to its variable nature? Volume is also variable, but less volatile perhaps. So is averaging the sequence of price over time the best or only way to contain that variance so that it appears more stable? In other words, how can it become clean for sufficient use without losing its responsiveness?

@Frank, I can build a chart where there is absolutely "zero" variability in the weight of the bars on the chart. That mean the only movement of the bars is solely based on pure price rising and falling. That is pure untainted price action. That is the sheer definition of pure price flow and is perfectly consistent, day in and day out regardless of the symbol or stock it is applied to..

The price of a completed buy or sell order is NEVER a variable. Once a transaction takes place that price is written in stone. It never changes! Volume is only a variable if you build a chart that isn't based in it. Averaging, as a physical function in market physics is worthless because it isn't precise. Something is either accurate or it isn't.

The endless story of indicators and their "predictive power". Why has no one ever imagined discussing "predictive power" of road signs, light boxes or, even better, IKEA assembly instructions?

Hi Alex,

Yes indeed, this topic has been covered a zillion times, but still people are looking for something to hold on to, to base their decisions on (Central banks have destroyed all asset valuations, so that is no longer valid as an element in the decision making process). So, let's change "predictive" into "probability". One can calculate the historic (mind the word 'historic') probability of a positive outcome and adjust money management to live another day. One must avoid having a false sense of security. Sailing blindly on an indicator is an accident waiting to happen. And talking of accidents, if you run through a red traffic light on busy crossroads, you can calculate to probability of being killed in an accident and so it may be unwise to do so. I will not elaborate on Ikea assembly instructions ..:-)

kind regards and godspeed

Hi Marc,

in my humble opinion the point here is that neither indicators nor traffic lights have any probability by themselves. It is only the trading rule (or driving rule) that may be statistically confirmed or unconfirmed. See the difference? I bet you do, as you wisely added "on busy crossroads" to your example with the red traduce light. It's more to the context where indicators (road signs) are used to determine the probability of a certain outcome of a certain trading rule. An analogy here works again amazingly well: for example, if you always drive through a red light on crossroads during night time then you may have quite an impressive statistics of positive outcomes, especially if you mostly drive in the countryside.

Indeed Alex, you are right and i stand corrected. Not the indicator in itself but the attached trading rule can be given a probability measure. I did mean to say that but not being a native english speaker, I did not get the message across.

By the way, Alex, you start very politely with "in my humble opinion", but some "humble opinions" carry more weight than others as they are more educated opinions. Yours is such.

At least publicly available ones are not consistently profitable on their own. They they had been, they wouldn't have hit the public domain. You can still use them for limited purposes. They are not trading systems by themselves.

@ Bharath, I know that a voice in the internet telling you that ALL predictive indicators have limited if no consistent predictive value, doesn't carry a lot of weight. Those of us that have spent decades verifying that very fact are simply trying to save you the same pain and wasting of time learning what we had to endure learning that fact.

@Alex, I absolutely agree with you regarding the predictive value of indicators. I'm curious as to your reasoning for dismissing the logical value of reactive indicators though. Each time I reference "reactive analytics" you answer directing your comments toward what we agree on that predictive indicators are worthless. You so eloquently reference "signs". Does not EVERY "road sign" warrant a reaction? Does not reading the IKEA instructions warrant a reaction? Does not EVERY crossroads warrant a reaction, regardless of traffic or time of day? EVERY action in the market is met with an equal and opposite reaction. This is common sense. ANY & ALL trading built on imbalanced or inconsistent environments is like building any structure on an unstable foundation, it is only a matter of time and it WILL crash. This is why 99.9% of trading methods degrade and stop working over time. Some far quicker than other because of the time environment they are constructed on.