Practicalities and Philosophies of Successful Trading by Goldenmember

Beginner Questions
64

[B]Charting Part 1
[/B]
What value do charts have then? They provide you with old and incomplete information about the following:

  1. They inform you where some speculators have traded long or short (note they do not tell you how many speculators or what their intentions are)
  2. Since prices are mean reverting in the long term, they give you a framework for calculating that mean/median/mode and the possible deviation

Of course, 1) is the concept behind support and resistance and 2) is the concept behind indicators

Then you might ask - if they can provide this information, isnā€™t that enough to predict the future? I would say no (pretty much what I have been argueing over the past few posts).

You cannot reliably use these points to predict the future because you do not know how many speculators are represented, what their intentions or how much the price will deviate from the mean so they give an incomplete picture of past action. More importantly, these charts give no indication of what is going to happen to future speculators who are not already in position, or real money which has yet to enter the market.

The predictive value from charts for new future dealings is unreliable. Charts do not tell you how many 18 year olds are going to go to the USA, or how many X box Ones are going to be shipped to Japan. There is no ā€œ18 year old traveller indicatorā€ or X box future sales pivot figure. They do not tell you what the coin flipper working for the investment bank is going to do. If you try and use charts alone you use incomplete information that does not have sufficient predictive value to predict the future.

Charts cannot be taken in isolation. All information must be built to produce a picture of technical analysis. I do not think technical analysis is one chart - for me it is years of charts building up a total picture so that you can actually build a strategy on ā€˜reversion to mean.ā€™ If you do not have sufficient numbers to create a robust trading platform you cannot trade this reversion reliably.

65

Yes, these information would definitely help a newbie. Maximum newbie enter this market with false assumption. If they read this thread and want to continue with us then they would probably get some guideline.

66

Iā€™m enjoying this, lots of logical/reasonable points. Looking forward to your future posts and finding out just what the hell you do to get the brilliant results you get!

67

Great reading ! Keep it coming Goldenmember

68

Since someone mentioned candlesticks on another forum lets talk about candlesticks: They are one of the tools that is often used - but really how useful are they? The daily timeframe or ā€˜candleā€™ is considered by many to be the most accurate, and the 4 hour candle is also very popular to mark out ā€œthe higher timeframe.ā€

Of course there is some information to be obtained by candlesticks or open close type bars over a line chart, but some people have developed this directly into a candlestick type charting. Others have refined this and trade pinbars, or term certain candles ā€œfakiesā€ - I am not too keen on these ideas (but of course candles do provide some information but just not how they are traditionally taught).
Taking a look at daily candles for instance: Iā€™m sure you have noticed that different metatraders have different timezones. For instance, Oanda uses Eastern Standard Time. Some UK brokers use GMT time. Alpari uses GMT+2. FXDD is an hour later etc etc.
This means that ā€˜candlesticksā€™ look different from broker to broker, and a ā€˜pin barā€™ in one broker could be something very different in another.
This situation also applies to the 4 hour chart - a GMT+2 4 hour candle is different from a GMT candle for instance. Waiting until candle closes, or pinbars to appear or engulfing patterns is highly dependent on your broker - and I do not think that the true interbank market really cares whether an engulfing pattern, dark cloud cover etc has been produced.
There is coverage of patterns and candlesticks from Thomas Bulkowskiā€™s encyclopedia of candlestick charts that covers reversal probabilities, but I donā€™t really see many forum candlestick traders using them - instead they apply what they see through candlesticks into real life. ie: There is a pinbar coming off the resistance indicating that there was rejection of that level pointing the way to price going down.

I think this way of thinking is wrong (of course others think they are correct). Price movements causing those pinbars to appear matters, but not the pinbar itself - instead of trading off the pinbar, examining why the price fell from that level is much more important than a pinbar.

Importantly candlesticks, OHLC bars tell you what happened in the past. No matter what people say, they are lagging indicators and do not reflect future events. A pinbar does not mean price is going to reverse. However the reasons behind the price movement can contribute to your knowledge of what market events are occuring to cause price movements and your experience and general knowledge can tell you whether this price movement will lead to a reversal or whether its an opportunity to buy a pullback.

There are provisos of course - some people trade exclusively technical set ups but what I look for is any clue behind market events and relate that to events that provide good buying or selling opportunities.

69

Thank you for taking the time to share these posts.

70

Hopefully you are Person E and realise that Person A, B, D and [B]author[/B] are all talking total rubbish.

Really. I am sorry, but I feel like I have to step in againā€¦ There are a couple of good ideas and stuff, but here are so many falsehoods that I donā€™t even know from where to start (I will cover just a few phrases). I hope you arenā€™t writing a book or so, because you have really poor argumentation and very limited knowledge about what you are writing.

the impression I got was that all the future events could be predicted from past events on the charts,

I obviously agree that this is false.

and that you didnā€™t need to know anything else other than what was on the chart. I donā€™t believe this, and I will show you a few examples why.

But not this.

Your examples arenā€™t even related to what you are arguing. It is like saying: ā€œI canā€™t slam a ball through the basket from above, my uncle canā€™t, my aunt canā€™tā€¦ so noone canā€¦ Example why noone can do that.ā€

virtually identical situation

Dude, your knowledge about price and technical analysis is very limited. I donā€™t know a s*** about atomic physics, so I donā€™t go to physic forums to express my false beliefs in teaching mode. Please donā€™t write about something you donā€™t know a s***.

Tools such as support and resistance, fibonacchi lines, pivots, indicators themselves have extremely limited value by themselves and if you use them you should know what value they have
Hey, before you argued about ā€œpriceā€ instead of this mainstream stuff. From your writing it is obvious that you think these things are equivalent, but I will repeat again: your knowledge here is very limited.

If you try and use charts alone you use incomplete information that does not have sufficient predictive value to predict the future.

A single support and resistance line does not tell you the direction - it just tells you that there was sell or buy interest in the past at that point. A single pivot line does not tell you how much the price will retrace - it just tells you the mathematical average.

You are missing the point. Trading is not about predicting future, but to find trades that in long run willl make money.

I believe your intentions are good, so good luck with your thread. But dudeā€¦ think what you write.

71

LOL! He has made almost 300% and $50 000+ and you are saying that his knowledge about price is limitedā€¦ Where is your myfxbook that supports your knowledge ARTjoMiS?

72

Well this is my first post, and I have to be honest the first multi page thread that has captured my attention long enough to read all of it.

I am a newbie and I find this thread quite timely. After a few months on a demo account I have now just started using a live account and find I am asking some very different questions to those I considered while using the demo account. Specifically the ā€˜WHYā€™ questions.

I read about, and tried, many systems on the demo account but taking it all at face value. Sell at X when Y happens etc without knowing ā€˜WHYā€™.

This thread is proving educational for me.

73

ARTjoMS, you canā€™t use words like knowledge without providing evidence to back up claims.

Lets take your example of physics. If you went on a physics forum and expressed an idea that was factually incorrect youā€™d be corrected but with arguments that are based on evidence. Thats the beauty of science evidence is replicateable!

Goldenmembers posts do a good job at highlighting how so many of the so called ā€œfactsā€ of technical analysis cannot be replicated in practice. Now if this was an opinion that was incorrect it could easily be refuted by showing proof much like on your physics forum example. However itā€™s not. You will soon realise that there is no proof offered to support almost all forex gurus/teachers/mentors claims. If a rule or form of analysis is not replicateable (by the same person or others) itā€™s not reliable.

Really all I am saying is if you believe in those types of technical analysis then start a thread and prove they work. Otherwise there is no reason to listen to your arguments. Iā€™m not saying that to be facetious but in my personal experience almost all claims made on this forum are incorrect.

The only reason I am following this thread is because Goldenmember has shown a verifiable record of successful trading. That record is proof that heā€™s doing something right and so I am willing to listen.

74

[QUOTE=ā€œARTjoMS;521305ā€]
You are missing the point. Trading is not about predicting future, but to find trades that in long run willl make money.

I believe your intentions are good, so good luck with your thread. But dudeā€¦ think what you write.[/QUOTE]

Uhhhhhhhhhhā€¦

Trading IS about predicting the futureā€¦ Once you move past the novice trader mentality that all you need to know is on the charts and begin realizing how the currency market is mainly affected by central bank monetary policy, capital flows, commodity prices, geo political events, etc you will realize that you are very much trying to predict the futureā€¦

You take pieces of information and try to piece them together and use it to figure out what will happen and how it will affect the currencyā€¦ Itā€™s this type of thinking that is required to progress from being a hobbyist struggling to maintain any type of profitā€¦ To someone doubling and tripling accounts on a regularly basis.

So to youā€¦

ā€œThink what you writeā€ :wink:

76

Very interesting and informative read. Thank you for sharing. I have a few questions Iā€™d like your opinion on.

  1. Since you approach trading from a mathematical standpoint, consider the following. On your mfyxbook, youā€™re currently trading 17 pairs. Letā€™s say you doubled that amount to 34 pairs. Would your win rate double, given double the opportunities? Or would you scale your risk down to half of what it is now? My point is, adding more trades per day with a 87% win rate is no different than trading today, then tomorrow, from a mathematical standpoint. Yes, you have more net risk, but it could be seen as flipping a coin once a day for ten days straight versus all ten today. Because our risk per proposition/trade is already small enough to have a 0% risk of ruin, we can technically take as many trades as we see with more evidence of one direction than another. Will this double your winrate?

  2. Is there such a system that canā€™t be scripted by an EA? Maybe there are too many variables that canā€™t be defined. For instance, in this one system I was reading about, even though the system was giving clear signals of entry, the author would sometimes sit out of trades because he didnā€™t like the setup. There were other factors involved, such as how price was developing, which isnā€™t easily scripted. My current strategy is somewhat complex and definitely does not give black or white signals. What do you make of that?

EDIT: 3. While I do agree with the randomness you depicted through the example of the backpacker, the business man, the airline revenue, wouldnā€™t all that be considered noise in the market? In the COT report, all the speculators are considered a negligible portion of the total positions. From what I understand through my learning (of 1-2 years), the markets exist primarily for the airline company group, basically those who are hedging against possible scenarios. They are a group willing to lose money in the forex market because they are buying insurance for their business. If theyā€™re hedging, someone must take the opposite side of their position. This is where banks come in. These players have so much capital that they are able to move the price 100ā€™s of pips if they wanted to. In this sense, the market is somewhat predictable and NOT completely random. Thatā€™s just how I learned. I could be completely wrong.

Furthermore, if you are trading a strategy that was backtested and proven successful, does that mean you are using automated trading? If not, why arenā€™t you?

77

Yes the market is not completely random, but there are limits to the information that we have at our disposal and that we can process. At a certain level the volume of information is so large and rapid that it is impossible for a human retail trader to predict - I have loosely termed this random, but a better word is chaotic. Prediction of the market comes in where the information that you receive gives you an edge.

edit: Sorry just caught this one about the number of pairs. I am not 100% technical, nor am a 100% fundamental (nor do I like those two words because they are unsatisfying when describing the market in my opinion, but thats more ā€˜philosophyā€™ which some posters seem to very much hate). I do like to keep risk to a minimum -in fact, I have probably been overcautious as a result of hitting a larger drawdown when I first started as a result of starting off purely technical.

I have noticed that a lot of traders ask me about my system, or ask the steps to my system, as if they are expecting that it is just A, B, C, D. Unfortunately its not that way and requires some lateral thinking, not just steps, thus it cannot be automated, nor is it just a single system. I do have systems (in the step A, B, C, D kind of way) that I have used in the past but they tend to be low expectancy and considerably more time consuming and stressful to trade. I am currently trialing as these as automated strategies on a small account, they are performing better than I hoped, and I will be adding this to my trading at the end of the year if everything goes well (unfortunately I will not be sharing these because like many automated strategies the edge is fairly narrow).

Charting (more)

I thought I would do a piece on candle charts since they are quite popular. Lots of people trade candlesticks by themselves, and this particularly applies to price action type traders. For instance, a long wick on the 4 hour chart meaning that price is rejected is a typical price action set up. Although it can give you a setup, and some traders can make money on it, its quite often a false signal.
Let me explain why by giving a bit of background:
Last time I wrote a bit about timeframes being different for different brokers, which is very important on the make up of a candle or a OHLC bar. So lets have a few examples of brokers (they may have changed since I last looked at them):

DF Markets 0 GMT
FXCM +2 GMT
Alpari +3 GMT
Oanda US -5 GMT

Adapting that to the chart, you can shift the 4 hour candle 1 hour either way depending on the broker you use to form a variety of different candle shapes. You can make lots of rejections and wicks appear on one broker and not appear on another broker just by selecting the server time.
Now looking at a ā€˜typical dayā€™ as described by a ā€˜trading guru/expertā€™ (not that there is a typical day, this description that follows is an oversimplification that itself is a marketing ploy to make you believe trading is simple and predictable).

London opens and the eurusd is sold. There is some ā€˜consolidationā€™ at New York and then eurusd is sold again until the London Close whereā€™profit takingā€™ takes place around some area of support so price comes back to the 38% fib line for the day. A ā€œtypical trending dayā€ (not that this actually happens that often unlike some people would have you believe). And in a downtrend, the same pattern repeats itself again and again. However, that profit taking can be made into a ā€˜rejectionā€™ candle depending on which timezone your server is in. Obviously this makes ā€˜rejectionā€™ candles very unsatisfactory ā€“ the example of the downtrend I described is a ā€˜rejection off support but in the longer model, does not mean rejection at all, and is a ā€˜false signal.ā€™
You can formulate the same hypothesis for any candle. Candles are a representation of price (obviously). There are a few gurus who say: I just trade trade price action(rejection, candles, S/R) and ignore everything else because ā€œits all in the chartā€ are missing so much of the picture. Its not to say that ā€˜candlesā€™ are entirely useless, but they only provide part of the puzzle. Candles, OHLC bars all represent buying and selling in the market and to be able to predict whether price is going to go up or down you need reasons behind the buying or selling, or reasons behind the candles or price (or you could ignore candles and just use reasons of course) rather than just ā€˜priceā€™ itself.

78

I am a newbie trader. Newbie traders ask newbie questions. Maybe it is the wrong question, but:

When there is no specific system or any specific steps, could you maybe then explain your trading decision making process by some of your recent trades?

What about these two:
a) 09.03.2013 GBPUSD Buy @1.55564
b) 08.28.2013 USDJPY Buy @96.88000

79

I have to disagree with this as a short term trader who holds trades from a few hours up to 24 hours, absolutely everything I need to know about my trading analysis such as entry, exit and management is indeed in the charts as a visual representation. I agree with what you are saying for traders who want to hold a trade for extended periods of time as I believe fundamentals do determine the overall direction of the market, however within this ā€œoverall movementā€ there are always opportunities to take an opposite bias to that of the bigger picture, retraces being an evident example which can be seen in the most dominant of trends.

80

I outlined these trades in my journal well before I took them. There were some overwhelming indications (technical and fundamental) that GBP was being bought over the last month, and prior to that there were reasons that USD was rangebound and it was a technical level.

81

Looking at the charts one of the most popular ā€˜indicatorsā€™ is support and resistance. As Iā€™ve emphasised I only trade what makes logical sense. I donā€™t trade random 38% fib lines or GMT pivot lines because there I donā€™t think I can attribute any real reason why they should be areas of buying and selling. This goes against what a lot of traders say - a few traders Iā€™ve spoken to swear to their fibonacchi lines, and other traders swear by pivots saying thats what pit traders used to use to scalp with. Well, I donā€™t use them because I cannot really say why someone would be buying/selling there, or how much is being bought/sold there. Of course, if someone has found statistically that they can take X pips with only risking Y pips and getting a Z% of wins over a long time period, I am not against it, but I donā€™t expect these lines and levels to guide me as to where the market is going to go.

So onto support and resistance. This has a reason behind it - it is effectively where traders who are caught in the wrong direction liquidate their trades. If they are long, and the market drops, then there will be some traders who will get out at break even when the market goes back in their favour. This forms sell orders at a certain price level, and people with access to the order book will naturally place sell orders at the same place.

There are different ways to work out the support and resistance - (I donā€™t just use the chart, but others do: its the most simple way) and just line up highs and lows). However, there are some limitations. When does the trader with a trade in the red decide to set his trade to break even? Is it when the trade goes 10 pips against him? 50 pips? 100 pips? How much money has he traded in the wrong direction? How many traders place orders with him - and do they place it 5 pips away to be filled, 10 pips away? And is the trade willing to take a small loss of 5-10 pips, or is he aiming for +1 pip to make his trade worthwhile? And of course, what is the prevailing reason why the price has returned? None of these questions are simply answered by the chart. And of course, support and resistance lines by themselves do not predict overall direction.

If you bother to do any statistical study of support and resistance, the actual edge is quite slim. Of course, the edge depends on whether you use a 15 minute, 1 hr, 4 hr, or daily timeframe (some people use 5 minute but I have not really tested it for any time period >3 months so my results are worthless here), and whether you take a 1:1 take profit: stop loss ratio, or a 2:1, or a 3:1 etc etc. It may surprise you to learn (well, it certainly surprised me) that the win rate of a 1:1 ratio is pretty close to 50%. Depending on the ā€œtimeframeā€ (Iā€™m not a fan of tmeframes as I have mentioned, but this was for testing purposes) and the take profit target it may vary by a 2-4%, but thats what I found testing (over 10 years on the EURUSD, AUDUSD, GBPUSD pairs). With a 2:1 the win rate is less (30-40%), and 3:1 less again (about 30%), and it is a narrow edge. But what surprised me was that on a 1:1 ratio at a line of support and resistance, the edge given by this ā€˜support and resistanceā€™ was slightly more than flipping a coin. Sure, you can make money this way, but I pretty much came to see that it was a laborious and generally unrewarding task (unless automated).

So next time you run into a vendor or thread spouting how great support and resistance is, I hope this post is of interest. I am not saying it is useless - I have mentioned it has a real basis, but there are limitations - however, it can be improved if those limitations can be supported by other methods.

82

[B]Charting 5 (indicators)[/B]

One of the first things I looked at when I looked at a chart, and I expect a lot of people look at was indicators. A lot of people think that they are the worst thing ever, and that they are meaningless, but I think they are undervalued. Of course, they have their weaknesses, and just like price action or support and resistance I consider them to be poor if used to predict but they do have their uses.
First of all - very simple - what are indicators? Of course they differ from indicator to indicator, but many of them (stochastic, RSI, Moving average, CCI, bollinger band) are based by setting a range over the last periods that you determine. For example, RSI by default on metatrader 4 is set to the last 14 periods (or candlesticks, OHLC, or whatever you call them). So for the RSI based on the last 14 periods, the formula of the indicator sets a value on that range for the current price. Simplifying it further, the indicator is nothing more than setting a probability against a backdrop of the last few periods.
Commonly I read forums or blog posts where people claim the RSI/stochastic is ā€˜more oversold than ever.ā€™ Lets say that the RSI is under 10 - an uncommon occurence, and lets say for example the formula for RSI dictates that there is 5% chance (this is not the exact figure for all you pedants out there) that it goes under 10. Instead of thinking - its impossible that it will breach 9, 8, 7, instead I think along the lines of it has 5% chance to do so, if its breached 10 then it is one of those times. I donā€™t go along the line of thinking that a price must reverse because it is ā€˜extreme RSIā€™ or 'extreme stochastic.'
Its also worth noting that the range is based on the number of periods that is set as well. If you have the previous 14 candlesticks consisting 1 pip range, any movement will push the RSI into extremes.
To use indicators successfully in my opinion, a purely statistical method is required because they do not have a predictive element or a ā€˜common senseā€™ reason behind them like support and resistance. However I have found that the performance of indicators can have a very similar performance to price action. Of course you need to find the right set of indicators, but indicator based methods such as the 3 ducks method (I will call moving averages indicators although some might disagree) I have found are just as effective as price action from support and resistance (more to say about the ineffectiveness of price action than anything else) when backtested consistently.

There are some advantages of indicators over price action methods I have noted.

  1. Support and resistance is commonly hit during volatile periods where the spread is considerably higher. The number of orders at support and resistance can also cause substantial slippage. This means that if you backtest a support/resistance system with a 1.5 pip spread and you get a 55% win rate, in reality the spread widens when support/resistance is hit so the forward tested win rate is lower. This doesnā€™t occur so much on purely indicator based systems.

  2. Indicator systems are easier to program into EAs than price action.

Overall, this means that if you trade indicators in an automated fashion, you can trade over more pairs, 24 hours a day, 5 days a week with less slippage which eclipses any advantage a price action strategy has.

Of course, I am not saying that indicators are the best, or that everyone should trade with indicators, but they certainly have their uses if you can recognise their significant limitations.

83

[B]Charting Part 6 (overbought and oversold)[/B]

Another common use of the chart is to see whether a currency/stock etc is overbought or oversold. Some people base this on indicators, some people just think that going up or down X amount of pips is too much, or some peopleā€™s positions might just be deep in the red so they claim a move is ā€˜overdoneā€™. There is also a statistical basis for this, but like all statistical methods you need a sufficient population of results to make it work.

Looking at a chart is there such a thing as a overbought or oversold currency? If we move away from currency charts for a second and look at other charts then you can see that there is some basis for it.

For example if you look at an unemployment chart for a country over the years. You will notice that unemployment rates seldom fall under 1% - is this because of the chart or is it something else? 1% unemployment is unlikely and is going to be certainly a low because there will always certainly be a proportion of the adult population who are unsuitable for work. You can assume If you look at the high - at the moment for the USA and the UK the employment is 7-8%. For Japan it is 4-5%. In Greece it is 25%. If you plotted Greeceā€™s unemployment it would have been over the top for a long time. Is there a top? Is the ā€˜topā€™ dictated by the chart of the unemployment, or is it something ā€œfundamentalā€ which differs from country to country?

Lets talk about the GBPUSD pair as another example - is this oversold at 1.00 (and is that why it rebounds?) - or is there a fundamental reason? Is it overbought at 2.00 - is that why it dropped in 2009 or is it fundamental?

The answer in my mind is that there is a fundamental reason for ā€œoversold or overboughtā€ - a discrepency between the current price being paid and the current value. You can see the same thing in shares more clearly - stocks are bought when the company is overvalued - and that is overbought. Stocks are sold when the company still holds value, and this is oversold. In my mind you cannot work out what is overbought or oversold solely by the price. It is also highly debatable whether a 200 pip move is ā€˜oversellingā€™ because in the grand scheme it is not really that much of a difference (unless you are already in a situation where you have judged the fundamental value to already differ from the price, and the 200 pip move exacerbates it further).

Of course, the chart is necessary to see whether something is oversold or overbought because it tells you the current price. Price action believers would have you believe the nature of the movement of the market or the pattern of the market ALONE will tell you consistently whether something is oversold or overbought - I have illustrated chart situations previously where the same chart patterns are not consistent and do not consistently give information in previous posts. Charts must be used in conjunction with fundamental analysis to bring up an understanding of ā€˜oversoldā€™ or ā€˜overboughtā€™ if they are to be useful and is an important part of trading.

84

I just have one word as a comment: Entropy.