BuyLowSellHigh

But this is what I mean, your saying of "just stay away" implies that you can adequately judge if the market is in a momentum or sideways mode. And if it would be that easy you could easily make money by switching your signals from momentum-based indicators to oscillation-based indicators.

i get your point, so well just look at the general trend and decide your risk. If it goes sideways between resistance a support X times(3 times?) then wait. I do not have a perfect answer for you there.

This

it's completely random. the only people who tell you it works are the ones who went on a lucky streak with some inane strategy
you're chances of making a profit with random buying and selling are ~50% so it's not like these people are rare.

are you this stupid? if markets are random, why are there clear trends? also being right and the result of being right are 2 different things

Yep this. I only go margin on long for this reason - shorting this is very risky.

by the way: one question I have about the current margin-trading offers in crypto: considering the high volatility and slippage in execution time, do you run the "gap risk" of losing more than your margin so that you have to pay up? Perhaps not as enforceability would be questionable in the more anonymized crypto market, but still I'm curious...

are you this retarded?
it goes up, it goes down. 50% you pick the right thing.

chartbois and pattern soi bois are retarded to think the meme line drawing will help you make a better guess.

it's been backtested to stock data enough times and the conclusion was always the same : you can just as wel flip a coin instead of trying to spot "head and shoulders" or "cup and handle"
fucking retard

human brains are geared to see patterns even if they aren't there.
even if there was was a discernible pattern some trading bot is going to buy/sell it until it isn't there anymore before you can open your browser.

>you're chances of making a profit with random buying and selling are ~50% so it's not like these people are rare
Your chances of making a profit with ONE random trade is 50%. Your chance of making a profit over time with random trades are not

this.
Even then, some coins are manipulated ruthlessly; bouncing between two prices for fucking days like clockwork by an obvious whale bot, aggressive dumping of coins to kill all upward momentum at times etc etc.
Secure your gains.

yes, but it can go up MORE or Less. YEs it can go down from 6k to maybe 5.8k, but the upside that it goes to 8k is higher in the recent market.

This means, it doesnt MATTER if you are "right or wrong"; what matters is that WHEN you are right you get a big win.

Would you rather hit 300 goals in a practice match or the 1 goal in an important championship in football?

well there is a pattern in your chart, whats your point? If it would NOT be the case then the chart would look like random dart board with dots all over.

over a long period your chances just coincide with the larger market trend.
most stock day traders can say they make positive returns because of this, but the bulk of them under perform a single purchase of an index fund.

a bit hard to have profits in a downtrending market when buying and holding, therefore you need to be market trend agnostic and go both short and long

...

that chart was literally made with a random number generator. it's completely unpredictable.
if you're still willing to day trade after understanding that then I don't know. I guess you were going to lose your money one way or another.

>significant degree of subjectivity
my strategy is 100% objective and analytical, no emotions were harmed in its creation...
... and it's working so far (Jan 2016- Feb 2018)

>mfw am a bot author. One among many that outperform the market easily using relatively simple algos.

Academics, whether paid to do so or not (see: disinformation) use the WRONG INPUTS to their models. There's nothing random about the walk down Wall Street.

Price is NOT a function of JUST the previous close price. Nor is the market a closed system. Information enters it from.....what?

Traders. Duh. It's about crowd psychology.

Do you REALLY think people are that hard to predict? Do you REALLY think all those billion dollar trading firms REALLY just flip coins to decide on how to trade, despite hiring quants and ML guys like me by the truckload?

Yes goy. Markets are totally mysterious. Stay poor.

there is a big area between day trade and buy and hold. I usually trade on 3-10 day timeframe...

so what input do you use? do you crawl biz?

Yep. Not only that, but if you use a good position sizing system that lets you further maximize wins and cut your loses short, you can beat buy and hold even if you LOST more guesses at entry points than you won.

And if you have a system to pick entries that is even a slight amount over 50%, you WILL be rich.

There is NO proven technique trading works. Not on crypto, not on stocks. The only possibility is, if you are a such big whale to manipulate the market or work with media together for hypes.

Why? If there such one of the public known techniques would work, everyone would talk about it in this board at least.

So just hodl and dont be a fool.

So to the people that are trading crypto, do you find that TA hasn't been discounted as much as in the stock market? If you try to trade on a single indicator in the stock market you're not gonna make it. Are there less false signals in crypto because the competition between algorithms isn't as high as in the stock market?

so big banks and trading firms are just hiring traders to throw away money?

I actually do kek. I track sentiment and objectivity metrics. Often inversely correlated - basically do the opposite of what Veeky Forums says.

Also reddit, twitter, youtube comments, facebook and news feeds. Those are REALLY helpful - detect those patterns of manipulation. I'm looking at adding a few other sources too.

The Deep Learning models I use learn to identify probable movement based on the market itself (normalized price ohlcv), and the sentiment data. Also utxos, transaction speed, fees and other fundamentals as well.

About 70% accurate on predicting the next candle average price - I have different models for different timeframes, trained seperately. Aggregate the results. Have some rules on when to enter, exit and how to size things. Done!

Even much simpler algos work though - I've personally backtested the good old rsi and macd strategy Veeky Forums loves for example. Works fine, profitable with good position sizing. So anyone claiming that this shit is random is a liar.

You're a fucking idiot, majority of the well known hedge funds firms use bots for trading and guess how these bots function? They follow algorithms given.

See:

Yep. Bot author here. If this was random, bits wouldn't work.

Fucking brainlets. Ah well, least I can take their money.

can you shed some light on your background? I am interested as I am currently actively looking into ML but come from an finance/econometrics/economics side of things.

random walk or in general a stochastic identity model tn = tn-1 + e is just that, the most approximate model. Together with the gbm and similar. No researcher actually claims the true nature of markets is random lol. Random walk itself doesn't even imply that the underlying movement is fully stochastic, it could be partially random and so on.

>Yet, I keep reading about people here daytrading crypto and saying "just buy low sell high" or "use RSI / stoch.RSI".
People on this site lie and a lot are way less experienced than they pretend to be. Most succesful traders I know worry the most about how not to enter trades. Even if they could be profitable, it's about managing risk first. They don't even care if most of their trades are bad, they want their good trades to be really good. But I'm terrible at short term trading myself, maybe I'm just biased and genius really walk among us.

Trading based on technical analysis makes you predictable for the big players. They see how you will react and manipulate the markets accordingly.

For example if they want to invest large amounts of money they cant do that without pushing th price up. So they will make the price crash under the support level to trigger the liquidity in the underlying stop loss cluster.

>Hasty backtracking to obfuscate.

No. You just said it's random. 50/50, straight down the middle, pure guesswork etc. You admit that its crap then? That patterns may exist, that give a better than 50% chance of being correct?

If so, then whether any specific meme line bullshit is itself effective, such a pattern may exist, and tradeable. That right?

>I am not exactly glad not to be the only one beating the market but whatever...
:)

Sorry I cannot say a word without damaging my daily profits.

This is true.

Risk mitigation is huge. There's guys using strategies that only hit one time in 10 (stocks) but when they do, their wins so far exceed their losses they're crazy profitable. All about picking where to cut losers and let winners run.

The thing about trading strategies, if other people know about them, they stop working.

e.g. price is currently 500, you can predict the peak of the pump is 1000.
So you set your sell at 1000.
Another whale who knows this might put their sell at 990 so that their order gets filled first.
Someone sees the wall at 990 and knows it can't be broken by the upward momentum, so puts their sell at 980.
People will keep undercutting each other, weakening the final peak.

Same with the bottom of a dip, which is why when people predict bottoms, it usually stops short of that.
e.g. The BTC crash from 3k to 1600, EVERYONE who did TA predicted 1600 as the bottom.
It stopped short and bottomed at 1675, due to everyone putting their buys on top of everyone else.

So don't expect actual working strategies to be shared anywhere.

CS and Finance. Just a code monkey that likes to gamble, but also win.

I don't wagecuck though. Why build a machine that prints money and give it to a boss? I just trade, tinker with bots and shitpost here. Always looking for other ways to make more too - I've got a lot of other schemes on the boil.

>So don't expect actual working strategies to be shared anywhere.
Pretty much this.

To an extent. It's why anyone selling bots is probably a scammer. Thing is though, there's enough people using different strategies that it offsets the loss of liss of effectiveness.

> Let me tell you exactly why!

Real trading involves real traders, institutions and governments. Real trading involves weekly news that changes the market direction and real trading involving traders who believe in technical trading, most of the time it's a self fulfilling prophecy which is why it works in the long run.

Crypto trading isn't run by news all the time, recently it has been, but not like normal trading. Most importantly most traders in the crypto are not traders, they don't know much about trading anyways, they are holders. They don't when to buy or sell. Technical analysis could work as there are now traders as well as holders

The only thing that works consistently is drawing channels and areas of previous resistance and support

Same with new companies and their Investors. They invest to throw their money away? No, they try to make 20% per year (more then the world market). But the majority of new companies fail.

Investors will every try to play out the market, some suceed, some not.

What do you think where the money is coming from? From the people which are loosing.

ok so never invest in any company, never try anything beause you can fail?

This.

Therefor there is not the final technique which works. 99% what you will hear and try will fail. So be warned.

Of course you will hear how many people make 20% a day. But you will never hear from the other half who loose the 20% per day.
Money just change the owner. In crypto of course you can profit from new FOMO traders, but also this is a very very high risk.

>Of course you will hear how many people make 20% a day.
And if they give away too much information they are probably lying.

the only problem is this

you rebuy or resell accordingly

>rebuy/resell
... and I was probably giving away too much information.

fuck, I just watched that scene. You guys weren't joking when they said this is the anime of Veeky Forums

take loss immediately or you wait till you are in + again ?

immediately

if I forecast that it will dip and it doesn't I correct my assumption immediately

and it's a matter of how often you update your forecasts too

See the problem with that is that whales and bots shake out weak hands all the time on most every asset available. But I guess it is a safer play. If I used stop losses, I wouldn't be waiting for my shitcoin to rally 17% just so I could break even on that order.
On the other hand, if you're that trigger happy, you never make any gains. I don't know.

>trigger happy
please refer to the post above the post above this post

Well you said you take a loss immediately. Which is fine, I'm not arguing that it's bad. I'm simply poiting out a possible problem with that. But slow and steady wins the race anyway, so I'm probably wrong.

...

crypto jumps between +/- 1% value in hours sometimes minutes the trick is not to tell when the market is going up or down, the trick is to take all your indicators and grab those 0,5% profit over and over and over and over again.

Sure. I understand you.

I respect your approach although it is only anecdotically similar to mine.

Buy red, sell green.

I wish I could redpill someone about it, but alas that would defeat the very purpose of it all. :(

the problem is a lot of exchanges will fuck you up with fees one way or another

this was a nice thread, back to lurker mode

This. Although that really depends on the specific exchange---and approach.

It's actually less than 50% because of fees.

thinking of TA that it increases your chances of guessing it write is fucking retarded. you obviously don't even know what TA is used for.

how do i become you? seriously what you’ve responded with here (if no larp) sounds like the most interesting and fascinating shit to do. any tips at all? what should i learn?

i’m seriously so intrigued. part of it is of course the money aspect but the part that’s most pulling is the process of finding out those patterns that are successful and putting your mind to it and shit.

i don’t want to bother you further but is there any other way to contact you? i’m so passionate about this, i fucking love social sentiment and patterns

Its an expanding market, statistically BTC is more likely to go up than down at any given moment. Thats why there are huge squuezes at bottoms to buy that dip and the price recovers almost instatanious after days or weeks of falling.

Day trading is just about knowing the market circumstances and buying low selling high like you normally would with any market.

Eternal FPBP

Man this sentence is so fucked up, how in the world can you call someone retarded you fucking mong.

>BTC is more likely to go up than down at any given moment
underrated piece of information

I dream of a smart contract which would reveal/share my strategy to another person but only if that other's person strategy is the same as mine

whats your setup ubuntu, cuda, jupyter? python, matlab, octave?

what models you using
Do you collect your own data, or do you pay for it.

I am a student and need to pay for my studies, I am doing ok trading and have made a image classifcation app and retrained a model on my local machine after learning to use ubuntu, bash and everything in about 1 year. I am a non comper science student, i started doing this after an interest in computer vision, sift, point cloud decimation, and triangulation interested me in my background readings and from using photogrammetry for my practical aspect of my phd.

I need to make more money fast.

My biggest question is how do you secure your trades?

What setup you use, this is my biggest concern for making a bot.

Also this means that most pumps on b are pnds then

obviously

I am to retarded to learn english properly sry