(binned per thread starter's request) Market bottom rotation trad

EURAUD on track.

GLMD longed, should go straight on from here, or fail.

Good to see your longer term trade working out. I might have missed the post where you explained it, why did you enter the oil short?

GLMD covered break even.

The shorts on Crude were part of what i call "2014" play, when the USA opened asymmetric conflict against emerging markets (few selected countries), the play rendered trough the financial markets from rate hikes on USD with causing capital outflows from EMs, to engaging in trade sanctions to pull up inflation, to putting a heavy bearish media exposure to especially Russia, to tightening the libor markets and narrowing the capital liquidity to those countries, and to push the price of Oil artificially lower in order to deplete the EMs (especially Russias and Brasil) income and also at same time push up the inflation. Thus there was a very hidden but a very real diplomacy of US and Saudis (Kissinger - Bandar deal) to keep the supply of Oil heavily up and surpress the Oil price as much as possible over years to come, but the play itself was structured in two major components, first was deliver a shock which was 2014-15 when Oil price was crushed from 120-28 bottom, and the second part is increase of supply when needed if the demand is somewhat decreased which is the play that came into action when sanctions on Iran were prompted 1 month ago, that is what pushed me into believing that Saudis will pick up the supply since the US cannot quickly up-scale supply but Saudis trough Aramco can do it up to 2 mil barrels a day thats what my thinking was, so even though on initial balance it seems that sanctions on Iran seem very bullish on Oil due to demand shock, the reality is the hidden strategy of US- Saudi will actually reverse the price into favour of supply imbalance.
Thus what got me thinking price will drop even though we saw some initial rally from 67-75. Its just basically understanding the strategy that US and Saudis are playing and trying to think 3 steps ahead of what their moves will be.

Now with marking 68 as the low, its still very likely to see further drop slowly over next few weeks especially as the July Oil output and storage data comes out, which i expect it to be heavily bearish with strong supply increase and judging from the way price is trading atm that data has not yet been priced in to the extent it should be. A 65 - 60 test is likely to be seen for WTI and Brent.

Hey janpec

What are your view son USD MXN , I have been trading this pair since last 2 weeks and have found the same interesting.

It corelated with OIL and moves opposite of Gold.

Still confused if this pair is a good buy at current price.

Is 18.70 levels already a good bottoming scenario.

Would love to hear your perspective.

Interestingly it has given the most profit to me in my 15 days

Cheers

Hey, MXN surely is correlated to Oil but it doesnt have much to do with gold. If there was any correlation was mostly just a random chance in last 2 weeks. Mexican economy obviously gets a large income from crude oil its what makes it exposed to price of crude in itself, but there are many more layers behind to what actually drives true behaviour (as you can see MXN gained in last several days even that Oil dropped, as there are other hidden factors in what makes correlation drive).

USDMXN is quite closely similar scenario to USDTRY which on this thread has been explained many times over and over. Both TRY and MXN are EM currencies, and each of them are part of debt problem and shrinking income especially Mexican oil industry that is under pressure of low Oil prices and surging USD which impacts the USD nominated debt. On each of those pairs since 2014 there is a soft bearish macro from my side, meaning expectations for each of those currencies to further drop but there could be significant pullbacks in between unless some short term macro drive kicks in which was the case for Turkey, but it wasnt for Mexico (was on election of Trump, but since then none).

If you are good at something, keep sticking to that and cut short the rest especially if you are negative on the rest, so sticking to USDMXN might be case for you, but the most important factor is that you 100% understand why you had good performance on it vs the rest. If you dont know, the luck will eventually run out and turn against you. First is to establish were you shorting or longing last 15 days? And if so ....why?

EURAUD likely supply squeeze.

Thanks for thoughts

I am a pure scalper and I see quick pull back in each pair possible after a long long run in the 4 hours or 1 hr or daily charts.I believe in each trade giving me 100 to 200 to 300 USD I exit and I need 10-20 trades a day.

I have been long while 18.8898 levels last week and this dint act much a s a bottom so trying to be long from 18.7498 levels today. shorting the pair as well.

movement relating to gold may be a coincidence .

though the Fibonacci shows further downside I prefer bullish..We are all wrong many times. But I have developed all the patience its needed.

No love for USD MXN I trade anything I see

Alluri

I wanted to know if you finding an interesting pattern or observation in USD MXN

Fibonacci doesnt really show anything, no direction whatsoever its just a random walk in the park. Dont hold to any directional view from Fibo numbers.
If you dont believe that....test it. How many times does Fibbo number hold to the test, i can tell you results because i did the test many many times over....its completly unconsistent. Any process that isnt consistent at least to 50% with decent behaviour spike, is un-tradable.

I agree most of the time fibos do look successful in post-mortems

I have been consistent in my own system honestly which is nothing but look at naked charts and trade.

Its like plotting a 1 meter track line with 5 random plots and saying...there is chance that when snail goes from start to finish and pass those 5 lines it will potentially stop at some of those 5 marks. Sure it will in some cases, in some cases it wont at all, and if you actually test what is chance it will consistently stop at 1 specific mark every time you will find out....completly random and unconsistent especially if you judge it by the time it holds there.
It basically is like saying that every time FX pair tests X.355 mark the price will reject it,.....test that statement....it holds no consistency whatsoever. Building strategy around it is straight up suicide.

Thats basically fibbo retracement in nutshell....useless. Put any random plot on the chart and it will be as good as fibbo. Trading is all about efficiency. What is the result? How result behaves and how consistent it is.

I too believe in trading naked is all you need to succeed but there are lots of fibonacci/harmonic stuff on this forum. How do you think of Davit's thread on fibonacci pivots?

only two highs, but the momentum strong, worth watching for HH to bake.

a demand failure continue play

Pretty much the same as any other fibonacci method, there is no statistical relevance of data performance in any Fibonacci tool. If there is no statistical edge in terms of how price behaves around the tool, the key question to ask is : why use it in the first place?
All one is doing by using such indicators is overloading data into charts without adding anything to the edge. Since the trading and markets in general are a massive hole of information one can get very easily lost in all the things to look at, its very important that any info that you add to your analysis has to be efficient.
Personally i follow a long time profitable trader in SMB Capital and a fellow trader who trades debt markets each of them using fibonacci as part of their strategies, do they make gains due to fibo use? Not at all, as soon as you strip both those different strategies into all the variables that make the strategy work you can quickly notice that fibonacci number is just a reason to get them into a trade (additional to few other variables) but if you removed the 38 or 50 fibo number and replaced it with random 10, 20 , or whatever number the strategy would still work with a very similar performance, which means that fibonacci number in itself does not contribute anything meaningful to the whole strategy other than just adding a single additional confirmation variable for entry.
And that is the key reason why fibonacci should not be used, its not my personal subjective opinion, its a fact that you can test yourself that there is no consistency of behaviour when it comes to fibonacci numbers.
Plot a 38% retracement from random high / low distance and over backtest check in how many cases there was significant rejection, what youll find out that there are way too random results, meaning that there is no process that market cares about when it comes to such tool. It also make sense, majority traders do not use fibos, why should the market react to it then?

There are a lot of tools and indicators in trading, but majority of them are non-consistent and non-efficient. Dont be blinded by the fact that there is many threads about fibo tools that doesnt qualify it as useful. There are millions of people who willingly replace science with the fairy tales in Bible or Koran, but that by itself doesnt mean that infos in those two books are efficiency based and tied to consistency of behaviour.

As i pointed out before there are literally only two types of strategies that exists, and everyone is pretty much trading derivates out of those two. One is order flow trading which comes in many different forms, and the other is data / pattern overfitting strategy. Traders and investors that make money with data overfitting strategy dont make it due to use of the tool, but simply due to being there "at right time at right place". You can make money with both of those strategies but you better make sure you know why strategy is working. Fibonacci by itself will not provide any order flow edge thus it cannot be used for order flow strategy, it can be used for data overfitting strategy, but why use it in first place, why not just go for some easier and quicker tool like Stocastic. They both do the same job in terms that they only provide idological reasoning for why to enter trade, on efficiency basis they dont actually contribute anything.

Fibonacci by itself is the same as "support" and "resistance" tool, you can randomly plot the lines from 1 to 100 lines on single chart and call it resistance, but if the behaviour is not consistent (meaning price will pass straight trough in too many of them) why calling it a resistance if there is no resisting process that is consistent.
If it happens once every 10 times its not good enough, especially if that one time it happens is not very similar in how it happens (depth).
So both fibo and plotting sup-res lines fall into same category.

One can go on youtube and watch 1000 videos in single day about many different traders predicting ahead at exactly which high will Bitcoin have a rejection, none of that data has any use and the reason why not is because if onerepeats the test or "prediction" 100 times over, the results are all over the place. Any piece of data that has been over 5000 year of history written in science books is 99% based on very consistent behaviour, not everything, but a lot especially in the most robust sciences.
2 plus 2 is 4 always is and will be, thus is extremely cosnistent, and if something is very consistent you can strap a name on it. If its not,...then the name strapping is meaningless.

And the worst of the fibo tool is usually the way its used. Usually youll see a trader plotting a tool on historical chart and then saying...aha see the price got a bump exactly at 50% fibo retracement. That same person will next time say the same for 38 and 100% bump, because there is no exact pre-set condition or number that he is looking in the first place. Its extreme data overfitting. In order to have edge one has to know ahead in advance exactly what he/she is looking for because that kind of view comes from the test, if there was no test done ahead, then trading is just all a guess work.

Well said. Knowing what underlying element is making a strategy work is truly important, and it is also why I don't use fibonacci. The only possible reason why it may work that I can think of is mass psychology, eg. traders tend to pull the price to around 68% after a rally, but no one can prove this. Adam Grimes has done some statistical analysis and found out that trends tend to pull back to 68%. His research seems legit, but personally I won't take it.

Sure, thing is that Grimes statistical data is not useful, making backtests for checking behaviour is a very delicate matter in trading, the statistical analysis has to be carefully structured if you want to have good data. I generally do counts every week at least one and it often takes 3 hours just to carefully set up exactly what and how you will test and collect the data, because if there are too conceptual things you are collecting the data will not be useful.
The problem with most of statistical counts that i have seen reported usually in books by other traders, is that the way traders define a start and end of trend is completly different from trader to trader, which means that one cannot cross use the data gathered from first test and paste it to second if the initial variables are very different.
Basically what you get is what i call "EURUSD problem", which was a very popular thread on this forum, but the major problem was that there it was just an idea exchange not the execution exchange thread, it doesnt have any use if there are 10 traders in same room each telling other what to do meanwhile they all using very different strategies. Its a CEO of McDonallds shouting at CEO of Boeing how to construct next Dreamliner.....pointless.
Initial variables have to be the same in order to cross use the data.

And to let me state not that i think Grimes in particular is a bad influence, in my view he is probably one of the best traders to follow to when it comes to things on how to approach to trading, but the thing is for you to be data useful you do need to identify the trend and trading approach exactly the same as he does.

And then to make things worse, then you add another layer of problem on top of it, which is that majority of traders are too late to identifying the trend changes and too inconsistent to defining when the trend shifted in case A vs case B, which makes the whole statistical data even more questionable.
Thats why for statistical behaviour data to be useful one really has to carefully construct the count and also make sure that the variables that were used to define a pattern has to be transparent.
Like if i gave you a data that 34% of trend shifts come after a strong volatility shift (yesterdays GBPJPY example), that result of data is by itseful useless to you IF i dont explain how to define exactly what price info was used to identify trend and potential rotation.

Data over-fitting is the biggest problem of statistical counts in trading. The best way to avoid them is by having very robust and specific thing that you are counting in the first place. Trend by itself is too conceptual and not something that you really want to do in statistical count, because its too broad on how to even identify the trend in first place. The best you can do is to count something that is very consistent, and very robust. For example if you count a pattern that had 10 times done same behaviour from same price area, chances are that count will be more reliable and the data from it than counting what trend A vs trend B does on average.

Thing is that in trading one should always prioritize to focus on changing or improving the things that you are fully in control of instead of worrying about the things you cant do nothing about. So one can do hundreds of statistical counts on how trend is likely to behave in future, the actual better statistical count is to focus on the past trades that trader has executed and check the under-performance in there and then change something in trading approach to get better results, because for most traders there will be under-performance problems on trade execution that will cause the most problems, not actually being 100% right and predicting where the price will go and how it will go there. The main reason being is that by default trade patterns are not really consistent enough to really fully trust the data. I can call supply takeout pattern A and the next one B being similar, the reality is the order flow is very different in each of them and also very hard to fully determine how it looks like. One can do statistical counts in biology all you want because animal A will have 99% consistency towards animal B of same species, in trading there is no such thing, which makes data results more questionable.