[Tutorial] How to configure cryptocurrency trading bots?

[Tutorial] How to configure cryptocurrency trading bots? - Last update: 27/10

<< Click HERE to join our Telegram ( real-time chat program) group - Friendly community, great discounts and more >>

TABLE OF CONTENT

1.	TRADING BOTS
1.1	CRYPTO-PHP
1.1.1	DEFAULT MODE
1.2	ARBI
1.3	HAASBOT
1.3.1	INTELLI ALICE
1.3.2	ORDER BOT
2.	YEARLY RESULTS

1. TRADING BOTS

1.1 🤖 CRYPTO-PHP 🤖

1.1.1 🤖 DEFAULT MODE 🤖

🤖 How to find better settings? 🤖

1. Analyzing trend

The first factor to take into account in CRYPTO-PHP settings when using "Default mode" must be the altcoins trend. This mode must not be run in a downwards trend (except when using CRYPTO-PHP add-ons, see 1.1.2 chapter). In my experience, a comfortable method to determine the current altcoin trend is to visit Coinmarketcap and compare the red Change (24h) values to the green ones. Trades are like a river. This river has two properties: direction and speed. Direction could be "downwards" while speed could be "hard downwards" if this river of trades flows fast down (so-called crash) or "light downwards" if flows slowly down. The more "-" are among Change (24h) values the more likely that trend has downwards direction. The more negative the values are, the harder the trend is.

2. Starting in the right time

Using the Coinmarketcap example at the step 1, the longer the green changes are, the safer to start CRYPTO-PHP in Default mode.

3. Cutting the loss short

The trends are going to change. A CRYPTO-PHP bot started in the right time (see the step 2) is going to meet a downwards trend sooner or later. Besides the values of the Coinmarketcap example (see step 1 and 2), a sympthom of a downwards trend could be the positions turning red. By my reckoning, it is very easy to choose doing nothing. There is a mental trick to break free from this "being stuck with losses" illusion. Red orders are losses already. Cutting losses short is my suggestion. In the CRYPTO-PHP Default mode, cutting losses short is selling all orders (even if loss) and shutting down CRYPTO-BOT then returning to step 1.

The evolution from steps 1 to 3 is iterative. Way I see it, the user may get better at cutting losses short in red market, entering early in green market.

1.1.2 🤖 THIRD PARTY ADD-ONS 🤖

What is a CRYPTO-PHP add-on?

To quote BiLLY CRYPTO-PHP creator:

Quote

It's a tool created by an user

The CRYPTO-PHP add-ons I created are found there:
Here are some Sophisticated Technology Mods for » CRYPTO-PHP

1.2 ▲ ARBI ▲

▲ How to find better settings? ▲

1. Setting all the settings to the most aggressive

It is possible to do it manually in the graphical user interface,

or simply overwrite the content of the corresponding file with the following code

config-Okex.json

Code:

{"ethPairs":true,"usdtPairs":true,"btcPairs":false,"excludedPairs":"HOT_ETH TIO_ETH","tickerCheckInterval":15000,"dynamicStartBalance":true,"balancePerTrade":0.031,"baseCurrency":"BTC","pair1PriceBuy":0,"pair1PriceSell":9,"pair2Price":0,"pair3Price":9,"pair1VolumeMultiplier":1,"pair2VolumeMultiplier":1,"pair3VolumeMultiplier":1}

config-Poloniex.json

Code:

{"xmr_Pairs":true,"eth_Pairs":true,"ticker_Check_Interval":15000,"excludedPairs":null,"dynamicStartBalance":true,"balancePerTrade":0.005,"pair1_Price":0,"pair2_Price":0,"pair3_Price":9,"pair1_Volume_Multiplier":1,"pair2_Volume_Multiplier":1,"pair3_Volume_Multiplier":1,"pair1_Trend_Mode":0,"pair2_Trend_Mode":0,"pair3_Trend_Mode":0,"trend_Monitor_Candle_Size":"5M","trend_Monitor_Periods":12,"profit_Monitor_Refresh_Rate":900000,"profit_Monitor_Stop_ROI":65000,"profit_Monitor_Time_Offset":0}

Now that there are indicators, which are the trades, it is viable to proceed with tests to find better settings.

2. Changing one setting at a time

In order to find better settings, changing only one setting at a time is important.
Before proceeding to the step 3, writing the current value of the setting to be changed and the expected results is critical.

For instance,

Quote

I am going to change the Multiplier of Pair #3 from 1 to 15.
I expect to have more successful Triangular Arbitrages in the following eight hours.

3. Comparing results

If the results meet the expected results, the changed setting xor another one may be changed
else the setting that was changed must be restored to the previous value, see the note in Step 2.

Finding better settings is an iterative process, which is cycling between Step 2 and Step 3.

▲ Main concerns and my solutions ▲

Concern: Ending up having shitcoin bags after trading
Solution: The settings must be brought FURTHER than the most aggressive values, see the code blocks in the "How to find better settings?" section.

Concern: Not a single trade for days
Solution: The settings must be brought CLOSER than the most aggressive values, see the code blocks in the "How to find better settings?" section.

Concern: AV heuristically warns you about using the program
Solution: Add an exclusion for ARBI to the AV program - Detection is false positive.

Quote

CodeX, [06.07.18 14:10]
[In reply to Mastermind]
This is a heuristic detection. The AV is not saying that the file is not safe, it is saying there exists a malicious file which the AV THINKS is similar to the scanned file. This is merely a warning, nothing horrible.

▲ Tips ▲

VPS time stealing is common, I suggest using dedicated server
I can use better prices in bullish market than in bearish/sideways markets because unfilled orders mean less risk due to market price fall in the bull
I can use better prices at Poloniex than at oKEX because there are less shit coins at Poloniex

▲ About the "-100" result code of ARBI: ▲
I experienced that there is the need make clear: "-100" results represent bank gains. Bank amounts will be turned into base currency gains, sooner or later. Of course, we want to see gains in the base currency as soon as possible. That is why we chose that currency as base currency, after all. However, realistically speaking, each configuration that brings any result will have "-100" results too. If you check @bzcorp or my results at Bitcointalk, it can be seen that besides other numbers, we both have many bank gains. Nevertheless, we are far in profit with ARBI. Why do we prefer to see gains in the base currency then? An inconvenience is the so-called "loss of bank value" event. Loss of bank value is when one or more bank currencies dump while the user has amounts in them. In this case, the "sooner or later" might become "later" or "much later". What can we do? So far, we have discussed multiple options: using less aggressive settings, switching to faster dedicated server, re-configuration using codexBot. Now, I would like to introduce another solution: the bank signal service. Bank signal service is a signalling service that tells each day what bank or banks seem to be safe to use that day. As long as the bank signal service suggests the bank, it may be safe to accumulate gains into that bank. Safe bank amount or amounts are likely to be sold with profit within 24 hours. The service is donation based. The more donation arrives, the more users it can support. You can subscribe by telling it to codexBot#5099 (Discord).

1.3 ⚉ HAASBOT ⚉

1.3.1 ⚉ INTELLI ALICE ⚉

⚉ How to find better settings? ⚉

1. Analyzing prediction models

The more predictable a cryptocurrency price is, the less tinkering Haasbot Intelli Alice will feel necessity for later. In order to choose more predictable cryptocurrencies, the price movements over a time interval could be compared to the prediction models for that time interval. Prediction models are found at WalletInvestor. WalletInvestor is a handy example because for each cryptocurrency, it shows a prediction model and rates how accurately this prediction model has worked so far...

2. Inspecting real price movements

To quote @RusselRick, a valued Haasbot expert in my Telegram groups:

Quote

RusselRick, [14.07.18 12:54]
[In reply to CodeX]
There is a really easy to grasp thing about Alice, you just need to do small mental trick

RusselRick, [14.07.18 12:55]
Think that Alice considering each trade combo (buy+sell) as a separate standalone one

Basically, from what @RusselRick says results that Alice could feel the necessity for some tinkering between a buy and a sell. When there is a large and unexpected (See the prediction model, Step 1.) price movement downwards after a buy, Alice waits until the price comes closer to the buy price.

What happens when the price does not come closer for a very long time?
When this very long time is like eternity?

The history of Alice could be cleared and have her restarted so the trading could go on.

The process that includes Step 1. and Step 2. is an iterative one. No one can predict future prices for sure. The ratings of prediction models should be checked every now and then, and if changes to the more unpredictable, a new pair for Alice could be selected.

⚉ Main concerns and my solutions ⚉

Concern: Not a sell after a buy for days while there was no major price movement downwards
Solution: Waiting more

Concern: Not a sell after a buy for days while there was a major price movement downwards
Solution: Selecting other exchange account and/or currency

⚉ Tips ⚉

Think long-term with Intelli Alice! She could be better than HODLing, but may occupy the allocated revenue for a long time before seeing profit.

1.3.2 ⚉ ORDER BOT ⚉

⚉ How to find better settings? ⚉

1. Paper trading

Before starting an Order Bot, all calculation must be done on paper. Testing the trading on paper first minimizes the probability of misconfiguration.
These calculations should be saved for later usage to compare them to the Logs of the Order Bot.

2. Translating paper trading into Order Bot trading

The input of an Order Bot is a sequence of orders. The trading in Step 1. must be translated into such a sequence where the orders are placed after each other linearly as what time order they should be placed in the exchange.

The Order Bot works well when the Logs match the expectations in the paper trading. Otherway, the bot should be stopped and the difference investigated.

⚉ Main concerns and my solutions ⚉

Concern: If ... else implementation
Solution: Haasbot Order Bot supports If conditions, so-called Trigger Settings, but not their else lines. When the direction of the execution of orders depends on conditions, to all possible direction a condition must be given.

Concern: Not enough funds
Solution: Haasbot Order Bot supposed to do automatically as much as you tell it to do manually - no more, no less. Thus, each calculation must be done on paper first, by the user, so does the fee calculation. The fee amount must be subtracted from the traded amount.

⚉ Tips ⚉

The main advantage of Haasbot Order Bot compared to manual trading is being able to do something else personally, like sleeping, when an order must be placed.

However, there are two more advantages that are worth thinking of:

One of these advantages is that the user's orders are invisible to outsiders until the appropriate Trigger Settings are met.
In this way, hiding the user's strategy from outsider analysis becomes much more easier.

Another advantage is that it is possible to place If control structures into the sequence, therefore it is possible to create pretty complex trading programs.

2. YEARLY RESULTS

[ View yearly results ]

Choose and take advantage of automation!