To what extent do you use indicators?

This whole argument reminds me of a conversation I had several years back with a buddy of mine who, at the time, was a pit boss in a craps pit at one of the casinos just off the Vegas strip. I asked him if his casino had any standard operating policy to deal with a player, or an entire table, that got hot for a long period of time and really put a small hurt to the pit's shift numbers (for those who don't know, in most casinos, the floor takes numbers every hour or so and they keep track of how much that pit has made or lost for the entire shift). He kind of snickered and said, "hell ya....open up more tables."

Turns out, he wasn't joking at all. When the laws of uneven distribution cause the laws of averages to take a "random walk" so to speak, what we're really saying is that our sample size is not yet large enough for the laws of averages to have "substantial significance." So the only way to combat uneven distribution is by increasing the sample size. In the casino's case -- place more bets; in a mechanical trader's case -- take more trades. As our sample size continues to approach 1,000,000 the laws of averages will continue to become more and more meaningful (over the entire set, not over any small sub-set).

Now the reason why it appears that your chances for success are greater after a series of losses, or less after a series of wins, is not because they really are (because we know, in fact, that they're not -- they're exactly the same), but because we're allowing ourselves to be influenced by what little we do know about "mean reversion theory," which is an entirely different issue all together. Because of mean reversion theory it is reasonable to expect the laws of averages to evenutally gain significance as our sample size increases, however, its important to note that the operative word is "eventually". I'll say it one more time.... EVENTUALLY.

So in a way, it is sort of true that the most important time to trade IS after a series of losses, but not because our chances of success have increased (if we already had an edge, then we don't need them to), but because we need to increase our sample size in order to give the laws of averages a chance to work to our benefit.
 
Quote from Jachyra:

This whole argument reminds me of a conversation I had several years back with a buddy of mine who, at the time, was a pit boss in a craps pit at one of the casinos just off the Vegas strip. I asked him if his casino had any standard operating policy to deal with a player, or an entire table, that got hot for a long period of time and really put a small hurt to the pit's shift numbers (for those who don't know, in most casinos, the floor takes numbers every hour or so and they keep track of how much that pit has made or lost for the entire shift). He kind of snickered and said, "hell ya....open up more tables."

Turns out, he wasn't joking at all. When the laws of uneven distribution cause the laws of averages to take a "random walk" so to speak, what we're really saying is that our sample size is not yet large enough for the laws of averages to have "substantial significance." So the only way to combat uneven distribution is by increasing the sample size. In the casino's case -- place more bets; in a mechanical trader's case -- take more trades. As our sample size continues to approach 1,000,000 the laws of averages will continue to become more and more meaningful (over the entire set, not over any small sub-set).

Now the reason why it appears that your chances for success are greater after a series of losses, or less after a series of wins, is not because they really are (because we know, in fact, that they're not -- they're exactly the same), but because we're allowing ourselves to be influenced by what little we do know about "mean reversion theory," which is an entirely different issue all together. Because of mean reversion theory it is reasonable to expect the laws of averages to evenutally gain significance as our sample size increases, however, its important to note that the operative word is "eventually". I'll say it one more time.... EVENTUALLY.

So in a way, it is sort of true that the most important time to trade IS after a series of losses, but not because our chances of success have increased (if we already had an edge, then we don't need them to), but because we need to increase our sample size in order to give the laws of averages a chance to work to our benefit.
More...

Now that is a post. Very well said.

BTW it is more apt to talk about "the law of large numbers" rather than "the law of averages"....... as the sample size increases towards infinity, the pecentages more correctly reflect what they actually are.
 
Interesting thread, at first it appears that we are talking about things that are mechanical when we watch indicators, but IMO we are not ... we are talking about emotions.

Somewhere embedded deep in our minds is a fear of direct price action.
This is where our money is lost without excuse.

Indicators are derivatives of price and somehow they soften the sheer brutality of price action and offer false hope and comfort when we trade either with real money or sim.

If you think about, we are all encouraged to lead derivative lives so others may profit, so what is the difference.

Once I became aware of this, my trading took on a new dimension and took me away from the land of odds, the land of % accuracy, the land of ratios and into a realm where I am only concerned with what I see right now.
I see the price coiled at s&r lines and I watch the bid/ask being hammered or not. The price is going to BO, but is it going to fake first or not. These are the only things that concern me.
MA, RSI,CCi can offer nothing of use, in fact they get in the way and serve to confuse. This is not an air conditioned B52 at 40,00 feet. No, this is knife fighting in the trenches.

To learn this I turned off all indicators, colours etc and forced myself to watch only the price in black and white.

I remember in the beginnig my mind was screaming to do anything else but watch the screen. Gradually it began to relax and enjoy the ride., but it took a while before I would trust myself sufficiently to add in more information to the screen.

In a small way like a recovering addict contemplating their first social drink ... can I trust myself.

So what was the problem.
I have a long history in business and sports and I am/was addicted to winning through the process of right and wrong.

By watching price only in B&W my mind knew that if I was wrong there would be not way that I could spin the disaster back into success. In essence it was trying to protect me by applying old successful behaviour to this new venture.
What I had failed to do when I began the journey of trading was retrain and integrate parts of my mind for a totally different world with new sets of behaviour.
In other words, a part of me entered trading leaving a big part of me behind.

Anyway, make of these comments what you will, that is the smorgasborg of life.

In answer to the question of this thread .....
I watch only volume impacting on price.
No indicators.
 
Quote from KPCURRENCY:

BTW it is more apt to talk about "the law of large numbers" rather than "the law of averages"....... as the sample size increases towards infinity, the pecentages more correctly reflect what they actually are.
More...

You are absolutely correct. Technically speaking, the "law of averages" is really just a lay term for a concept supported by the "law of large numbers." Likewise the "law of distribution" typically refers to some sort of probability distribution function (of which there are many).

Also note that these concepts apply strictly to random events, which may or may not apply to the market (depending on what camp you happen to be in).
 
Quote from IronFist:

I'm falling back into the indicator trap. I keep trying to find that magical combination, like:

"ok, when x indicator falls below 100, then look for a negative crossover of y indicator, but it only counts if the red line on z indicator is below the green line..."

Then I find 5 good trades that DIDN'T follow my rules, which doesn't concern me (missing a good opportunity is better than taking a poor one).

Then I find some trades that matched my "system," but they didn't work (backtesting by eyeballing the charts), and I justify it as "oh, well blah blah blah, so I wouldn't have taken that one."

Then I find a sweet one that works, and I think I'm a brilliant genius.

But the weird part is that I LOVE pouring over the charts, looking at "ok, it would have been profitable to enter here, so what were the indicators doing at this time?" and trying to find patterns and systems that involve weird combinations of indicators and stuff that no one else would ever think to look for. In fact, I'll load a new indicator on my chart that I've never heard of, study it for a while to try and figure out how it works (or determine patterns), and THEN do a google search and figure out how it's supposed to work. Sometimes I'm completely off, and I think my way works better than the "conventional" way.

Do you use any indicators at all? Which ones? If a trade looks good, but your indicator(s) doesn't support it, will you still enter? It reassures me when I enter a trade, and then my indicator confirms it... but by that point it's too late, anyway.

Could you trade successfully using ONLY indicator signals? I mean using a wrote system based entirely on indicators (ie. enter when x, y, and z happen and exit when a, b, and c happen) that a robot could follow and have it be profitable?

Or is this what they call "searching for the holy grail" and I should just give up?
More...

Perhaps this information may be of some use in becoming less of an indicator trader:

http://www.tacticaltradingmethod.com/indicator-trading.html

Some interesting thoughts at this site.

--------------

Now on the other hand is the indicator fascination people like the following:

http://www.easysignal.com/About.htm
http://www.easysignal.com/What_Is_KwikPoP.htm

"At the risk of sounding arrogant...we know our indicators work. We know our indicators will improve your trading because the signals are clean and easy to follow."
---------------

Good luck! Lots of people out there willing to sell you something when it comes to trading......why.....much easier to sell stuff than trade.
 
Quote from Jachyra:

You are absolutely correct. Technically speaking, the "law of averages" is really just a lay term for a concept supported by the "law of large numbers." Likewise the "law of distribution" typically refers to some sort of probability distribution function (of which there are many).

Also note that these concepts apply strictly to random events, which may or may not apply to the market (depending on what camp you happen to be in).
More...

Technically, that is not true. The law of large numbers allows for periods of tend. While trends can't be quantified or predicted mathematically, their existence in nonetheless accounted for. That is, the law states that for some period x, the "coin flip" can be heads 15 out 20 times. This is a trend. Yet as the sample size move towards infinity, the distribution will more correctly reflect actual probabilities. So it may seem random at 50/50, but certainly we can get there with periods of trend.

It is the mathmatics of trend that is not yet completely understood. That trends do exist, however, is not so much a debate.
 
Quote from KPCURRENCY:

Technically, that is not true. The law of large numbers allows for periods of tend. While trends can't be quantified or predicted mathematically, their existence in nonetheless accounted for. That is, the law states that for some period x, the "coin flip" can be heads 15 out 20 times. This is a trend. Yet as the sample size move towards infinity, the distribution will more correctly reflect actual probabilities. So it may seem random at 50/50, but certainly we can get there with periods of trend.

It is the mathmatics of trend that is not yet completely understood. That trends do exist, however, is not so much a debate.
More...

I think you probabably misunderstood what I was saying. Either way, it doesn't really matter because it looks like we killed this thread.
 
Most indicators don't work and those that sometimes work, work because people think that they do.

However, as a future trader I do use technical indicators and without them I would have a harder time trading.

First and beyond all, geometrics and volume.

Of lesser importance but still very useful.

- Keltner Channels
- Bollinger Bands
- 13 EMA and 20 EMA

And last but not least Pivots. Pivots are extremely useful because unlike the above mentioned indicators they try to predict the future instead of telling the past.

I have recently killed MACD, RSI and several other COMMON indicators that were actually hurting my trading. In fact, I have found the new screen space provided by their elimination to be much more useful :)
 
Interesting discussion and I found myself in the very same position as IronFist...

What has been most helpful to me is a program called StrataSearch...I am in no way associated with the company,but I have found it to be the best program for "answering" the questions regarding indicators and backtesting...

www.stratasearch.com

Essentially it is a program that comes with 2500 trading rules as well as the ability to create indicators and trading rules..It also creates sectors and industries(equal weighted) and things like A/D,breadth studies,New Highs/New Lows,RS vs industry can be incorporated.The program will search thru the rules,on selected stocks,sectors or markets and come up with the "best" permutations and combinations.Optimisation is an option as well as ranking/selecting systems(rules) dependent upon ones criteria,i.e Sharpe,Max DD,recovery ratios,etc..
Extensive stats reports are generated as well as Monte Carlo Simulations....
One can run various systems all at once,long or short and on differnt markets,stocks or sectors..And automated walk foward analysis can be performed as well.....

With all this said,I have learned that combining systems is probably the most efficient way to emply indicators and systems.As SS can produce "optimal" combinations,you may be VERY suprised as to what works and what doesnt.Much of our beliefs and what we have read,simply is not the way to go.And as I have learned the hard way,and verified with WFA in SS,what works today may not work tommorow...Hence,the multisystem approach..
As one poster noted,all we are trying to do is shift the probability of success slightly in our favor,and that is what the proper use of indicators should do..
 
You must log in or register to reply here.
Back
Top