Statistics
What Does the Ichimoku Cloud Tell You?The technical indicator shows relevant information at a glance by using averages.
The overall trend is up when the price is above the cloud, down when the price is below the cloud, and trendless or transitioning when the price is in the cloud.
When Leading Span A is rising and above Leading Span B, this helps to confirm the uptrend and the space between the lines is typically colored green. When Leading Span A is falling and below Leading Span B, this helps confirm the downtrend. The space between the lines is typically colored red in this case.
Traders will often use the Ichimoku Cloud as an area of support and resistance depending on the relative location of the price. The cloud provides support/resistance levels that can be projected into the future. This sets the Ichimoku Cloud apart from many other technical indicators that only provide support and resistance levels for the current date and time.
Traders should use the Ichimoku Cloud in conjunction with other technical indicators to maximize their risk-adjusted returns. For example, the indicator is often paired with the relative strength index (RSI), which can be used to confirm momentum in a certain direction. It’s also important to look at the bigger trends to see how the smaller trends fit within them. For example, during a very strong downtrend, the price may push into the cloud or slightly above it, temporarily, before falling again. Only focusing on the indicator would mean missing the bigger picture that the price was under strong longer-term selling pressure.
Crossovers are another way that the indicator can be used. Watch for the conversion line to move above the base line, especially when the price is above the cloud. This can be a powerful buy signal. One option is to hold the trade until the conversion line drops back below the base line. Any of the other lines could be used as exit points as well.
Statistical approach to risk management - Python scriptThis script can be used to approximate a strategy, and find optimal leverage.
The output will consist of two columns, one for the median account size at end of trading, and one for the share of accounts liquidated.
The script assumes a 100% position size for the account.
This does not take into account size deviations for earnings and losses, so use with a grain of salt if your positions vary greatly in that aspect.
Code preview
cdn.discordapp.com/attachments/592684708551327764/848701541766529034/carbon.png
TradingView does not allow posting external links until you've reached a specific reputation, so i can't use the url feature
Input explanation
WINRATE : chance of winning trade
AVGWIN : average earning per winning trade
AVGLOSS : average loss per losing trade
MAX_LEVERAGE : maximum leverage available to you
TRADES : how many trades per account you want to simulate
ACCOUNTS : how many accounts you want to simulate
the inputs used in the source code are from one of my older strategies, change them to suit your algorithm
Source code
pastebin.com/69EKdVFC
Good luck, Have fun
-Vin
Support and ResistanceThere is no magical tactic or indicator to "see" the support levels on a chart, but how do astronomers know a black hole is around without seeing it? They use their brain.
There are several types of support/resistance, the obvious ones, and the less obvious ones, and which ones do you think separate the bottom 95% from the top?:
- Horizontal price level (Bitcoin 9000, 6000, 3000)
- Diagonal price level (trendline)
- Moving averages self fulfilling prophecy
- Constant flow of positive news
- Sovereign fund buying or selling (CCP bags) program
- There is more, up to us to find it, good luck
In my weight gain & weight loss example what supports the increase and decrease?
You cannot predict what exactly that person will eat everyday, if someone will drop them home or not,
how much they will walk, if they hear in the media about a "science" paper claiming chocolate is super healthy.
Those are all random factors (to us puny humans), they are part of the normal day to day fluctuations.
Mere mortals including the best know those fluctuations happen but do not trade them, only daytraders do.
The weight has some support during the visible uptrend. This support is the regularly added calories from the weekly pizza.
Exact same as an institution running an algo to buy shares by increments every week at the same hour (that you could scalp).
On the weight loss period, there is a resistance, which is made up of the every other day walk home and the lack of pizza.
Yes mostly the resistance here is not something but the absence of this something!
So while predicting with precision is not possible we can enter on what we consider 2-sigma oscillations for example.
If we bet on the early "reversal" being just noise with a stop at 3 or 4 sigma we get an asymmetric risk to reward.
The average casual investors (and Bill Ackman) are the ones betting on the 4-sigma "never give up! strong hands! we will win! the power of love!".
4 sigma is human nature (for most people), also not understanding probabilities intuitively is. But I'll make a separate idea about that.
Skilled investors not only exit but even if a trend starts with good support (fundamentals + visible with technicals) they will join the new trend.
But careful not mixing everything. "4 sigma" meaning very low p of being noise in the original trend does not mean high odds of opposite trend.
So this is it, to estimate what supports the price the smart market speculator has to work as a detective, find clues, and as a statistician, put these clues together with p weights,
add potential clues with their own p weights.
How can novices possibly imagine they can use "TA" to magically find these supports, they really expect some big red flashy arrows on the chart that anyone can see?
It's so delusional. As detectives they'd expect the killer to scream "I DID IT! I AM HERE!"?
And of course get a big paycheck for their work as a detective. This is not a video game.
I think I just wrote a book. PF, RR, WR, etc.Intro, you can skip this part but I think it would be interesting for you to take a quick look:
Statistics estimates and formulas? Trading is mostly about emotions, not statistics estimates and formulas.
Most people do not need all of those formulas, they don't need to make plenty of stats and estimates, but just focus on discipline and emotion control.
I got this quote: "The key to trading success is emotional discipline. If intelligence were the key, there would be a lot more people making money trading."
I agree. Analysis of broker data has shown over and over that over a couple of years 90% or more clients lost money, and often all of their money.
So no point doing stats & formulas for 90% of people that will lose anyway.
What they need most is discipline, to not lose all of their money, but rather just some of their money, and emotional control, to not blow their brains out once they lose everything.
I can make a few quotes too:
Checking a thing quickly...
Stanley Druckenmiller was 46 when he did something stupid with the dot com bubble.
George Soros started at 29 and his biggest (known) mistake was Stanley Druckenmiller.
Oh this one is interesting....
Alot of profitable ones that got really confident after a few years of winning and got wiped out or made huge losses or missed on much returns (Buffet says BRK cost him 200 billion, he'd be way above Jeff Besos). I see alot of late 20s to early 30s. But even older than this after decades, it's never safe, never let your guard down. But most typical is the ~30 yo guy that made lots of money for years and laughs hysterically at noobs (retail traders mostly) and was warned of dangers by people trying to scare them away but proved every one wrong, knew he was at the top, one of the best in the world, so got really arrogant, dropped his guard down, and then boom.
By the way, totally unrelated, should I all in short USO? It's losing money over time and already so many idiots "invested" in it. There can't possibly be more morons that would buy this dead crap right? Lmao USO investors, what a bunch of brainlets. I refuse to lose against idiots just by being outnumbered. All in no SL. 😁
How do I start a show so I can do literal pump and dumps legally like Joseph Granville?
Good. Now that we got this out of the way.
1- Winrate
Pretty simple here. All this shows is what percentage of bets are winners. Doesn't really account for breakeven, doesn't differentiate between small wins big wins. Pretty useless on its own. Implicitly means that every win and loss have the same size, like putting rigid entry target SL, and never touching it.
2- Reward/Risk
How big is the average or expected win, compared to the typical loss. How much are you willing to risk and how much do you expect to make?
Most "educators" repeat how important the risk to reward ratio is, and it kinda is, because it is one of the best predictor of success.
FXCM published some data where they show that over the 3/1/2014 to 3/31/2015 period (1 year), 53% of their clients with a RR of 1 or more were in the green, while only 17% of those without were.
47% of RR >= 1 lose money. 83% of RR < 1 lose more. Their typical win % over a quarter is 25%, and the typical global win % over a year is around 20%.
I would be willing to bet that profitability goes up significantly with reward to risk. Some of it would of course be simply because people that end up with a huge win on their hands balloon the high RR stats.
That said, I doubt just flipping a coin, just randomly buying with a tight stop and a far away target would work. Althought...
The top myfxbook systems are almost all automated garbage systems with an average win 0.20 times the average loss, that were really lucky over a long period (3 std dev of a normal statistical distribution = 0.3% 3/1000, just pick any trash system with high WR and run a binomial probability calculator find the odds of it making profit over 100 rolls). Hey I'll do this later in this idea.
And as I was saying, perfect transition, flipping a coin isn't a viable strategy, the reward to risk alone doesn't say it all, even if traders using a high reward to risk ratio greatly outperform those that don't. If you make 10 times what you lose, but you lose 99% of the time, emm how to say...
And this is why we must look at the profit factor.
3- The profit factor. Oh yes
Pf = (W*R)/(1-W)
I have seen reports with a gross PF of almost 3, and net of barely 1.1.
If you design a strategy you count spreads in it... It's obvious.
Day trading sucks and every analysis of day traders data shows about 1% or less make money, and don't make much.
Probably the only ones making anything are level 2 scalpers, and 'experts' selling day trading robots, or signals, or courses.
First a disclaimer! The argument of day trading having terrible profit factors applies to 95% of the time.
When the average move per unit of time goes way way up (spreads & commissions usually don't especially if volume goes up too),
and you get in 5 hours what you usually get in 2 weeks, then obviously it's different.
I focus my argument on 95% of the time, when volatility is "normal" (within 2 st dev basically, and in particular within 1 - ~70% of the time)
And I have been really nice here.
Getting an idea of what good profit factors are...
If I participated I would take a single bet with huge leverage and hope to get lucky, easy win once every couple of events, but I doubt they allow this.
Lol on the worldcupchampionship site (ran by the CME I think), there are categories, Futures traders at the top have massive returns, way above Forex.
Previous year winners with futures have bigger returns than FX, but this year is just stupid. maybe they blow up soon.
Top 5 FX participants as of May 14 have 40% to 97% returns. Top 5 with futures are already at 200-800%!
In 2018 futures winner made 250% FX winner made 200%, sometimes futures traders make huge gains. The gap is already so big lol. Anything to do with NatGas & Oil? 😆
www.worldcupchampionships.com
Looking at a "war of traders" results. 27 days... Not sure what their leaderboard is. Looks like a great way to get suckers to deposit money and pay fees asap.
First place has a PF of 44%, I assume this means 1.44, second place 160% I assume it means 2.6. Followed by 1.3, 1.05, 1.13, 1.26, 6.85, 1.11....
Prob easier to get a higher PF with commodity futures where they are so much hedgers, much fewer care about hedging FX risk, plus central banks use it to manipulate everything, more people trying to make money.
Sometimes the sharpe ratio is mentionned. Quick definition:
The Sharpe ratio measures the performance of an investment compared to a risk-free asset, after adjusting for its risk. It is defined as the difference between the returns of the investment and the risk-free return, divided by the standard deviation of the investment.
I looked at some hedge funds reports a while ago, since they diversify and hedge alot PF isn't as high.
Warren Buffet has a PF of what? 100? He makes one trade every 10 years.
It's basically impossible to find those numbers, unless you work at a brokerage, and apart from your own, with the exception of the few times a broker releases some data.
You have to take into account how many opportunities you get also, and more but alot if implicit.
I would say that a PF too low is bad, because when conditions change you will take long to notice with certainty and you will also lose way faster! If you had a PF of 1.1 you spent 5 years to grow, and that can be lost very fast AND it takes you longer to realize it is not working anyway.
A high PF has a high margin for error, profits grow fast enough so drawdowns don't eliminate years of progress, and going from 2.5 to 0.75 over a period kinda is extreme.
I think typically for operations that target 1 to 5 daily ATR, (days to weeks holding period), and you get more than something like 1 single bet a year, good profit factors are in the 1.5-2.5 range. Lower than this gets a little dangerous, more than this is the holy grail.
A 25% winrate 5R system has a PF of 1.67.
4- Max Drawdown & risk per operation & max risk
Here you use a binomial probability calculator.
Plenty on the internet.
Winrate 25%, Reward/Risk 5, PF 1.67
==> After 60 bets, on average you should get 15 wins 45 losses.
The odds of getting more than 15 wins (P: 16 or more out of 60) are 43%.
The odds of getting less than 5 wins (55 or more losses) are 0,0956%. 1/1000.
10k account. Flat $100 risk per bet.
55 loss 5 wins = $5500 in loss, $2500 in wins, down $3000.
60 loss 0 wins = $6000 in loss, $ZERO in wins, down $6000 (rekt.)
So every 1000 trades you should expect something like this right?
Even with a very decently profitable strategy it will happen.
You have to decide at what point you consider the odds of it just being bad luck to be too high, and you just want to drop it.
Smaller drawdowns are going to happen absolutely all the time.
If you are risking 1% every time and adjusting, 55L 5W would be a 27% rekt, and 60L would be a 46% rekt.
The odds of losing 18 or more out of 20 are greater than 9% (9/100). Will happen ALL THE TIME.
With 1% risk, drawdown of 8 to 18%. Expect it very often.
Some clients use funds to diversify, to get returns with low risk.
Some expect less risk and volatility than the stock market, but expect better returns. Cute.
5- Expected returns after 100 bets
Say you got a system like the one I used in my example (that you backtested + used over a great number, or just used over a greater number of operations).
Winrate 25%, Reward/Risk 5, PF 1.67
If you do not care about eating 20% punches in the face,
and risk 1% per trade, on AVERAGE, after 100 gambles,
then your results will be as such:
75 Losses, 25 wins
(0.99^75)*(1.05^25) = 1.6. Up 60%.
If you risk 1% of your 20 years life saving, you would get 20% drawdowns on a regular basis, meaning you worked for free 4 years.
You can play around with calculators and notepad to estimate how big drawdowns you'll get, how often etc.
With a 2% risk:
(0.98^75)*(1.1^25) = 2.38. Up 138%.
And regular drawdowns not of 8-18% but 23.3%-33.3%.
And once in a while drawdowns of 60% to 70%.
And a few times in a lifetime of 80% to ....
What is the max drawdown before divorce + jump off a cliff?
6- Expected returns after 1 year
And here we are...
Traders should have a vague idea to start with but mostly look at all of this after running a strategy correctly and with some profits, over a "significant" amount of time, kek can't give a number.
First of all what is the amplitude of moves you manage to catch?
So the first limit is obviously the number of waves / moves.
No matter what sytem you have you will not be able to join more waves than they are waves in the first place!
And then... how many you can catch, is much, MUCH, lower than how many there are. Duh!
Anyone with half a brain should be able to understand all of this at some point...
Someone that manages to be profitable and doesn't blow up should make 5 to 40% I guess.
That's that. It's exponentially harder, but also exponentially more profitable.
I think I should build a new income stream writting books...