Crypto Trading: Catching the Falling Knife
“What is going on with these Cryptos, anyone, just me! No, its everyone !” These are the common thoughts flying around for Cryptos. Everyone from analysts, to novice crypto investors and traders are speculating on “Is the hype over, or does it really over !”.
Folks, does it really matter !
To algo traders, answer is NO! All they want is volatility and liquidity. Put in simple terms, they want the price to move with good quantities being traded. Let me explain! Now with Streak, someone like me (medium account part-time trader, strong opinions) can also get involved by creating strategies based on my simplistic views without code, or worrying about any of the underlying complexities.
Note : All information provided here, is for education purposes only and should not be associated/construed with any form of trading, investing advice. All results are hypothetical and are mathematical analysis. Past returns don’t guarantee future returns.
For the sake of this article, an algo trader is anyone who uses a Strategy to trade, can be a technical traders, quant, HFT, etc (google should help you out with the rest). Also in single line : “Algo trading is using a logic/rule/analysis(other synonyms) to trade”.
Lets looks at an algos which trys to catch the falling knife after it bounces up from hitting a support level, exploit the fact that even when the prices are falling there can simple pull backs opportunities(price upticks) which can be caught and traded, along with opportunities that allow cheaper entries during true rallies.
As illustrated, some pullbacks and entries have been marked.Now what our hypothetical algo should do is, react when prices are going up for some time after a fall, and might or might not continue the same trajectory before we exit with profits.
We can look at various support & resistance charts, draw retrenchments to figure out what would be the right levels or conditions for the pullbacks. We can look back in hind-side and see missed opportunities and various levels, different technical conditions leading to ton of fear, confusion and dilemma.
As an example, here I am using a traditional and simple strategy, basically from chart analysis 101, which just buys when the price starts going up after having been below a support level.
If at 5 minutes interval, the price crosses above first support of Pivot point lines, I will initiate a buy. And I will exit (sell) what is bought if I see some profit say 0.5% or in a loss of 0.15%.
Now as simple as it may sound, we will need a way to verify whether this strategy performs or not ?
Running a backtest on this strategy gives +8.24 % in one month, from 28th Jan 19 to 29th Feb 19 on BTC/USD pair of COINBASE and XBT/USD pair of KRAKEN and +7.3% for BTC/USDC pair of COINBASE.
The algo simply reads
Psychological reasoning behind the trade
When the pair is falling sharply, most of the market participants will be going short at every levels. Since S1(support 1) Pivot point acts as important support level, traders will be initiating short if prices close below it anticipating further downward move. However, when the pair reverts and closes back above S1, fresh buying arises and price moves higher. This causes a panic within the short sellers and they run for cover causing prices to rise further. During this chain, many stoplosses also get triggered which were put up to protect short positions, adding to the upward move.
We can see here that even some of the simplest conditions can help capture profitable trades even when the price has fallen significantly, catching the falling knife without bleeding. The key is in reacting quickly to market changes and exiting systematically with some profit or by accepting smaller losses.
As traders, we need tools that allow us to backtest our strategies so that we can find the ones that have at least some statistical advantage over flip of a coin. This is just one part of the game. The harder part is the execution. If the execution can be automated, it would much faster at placing precise trades and save would save a lot of time which otherwise would be spent staring at the price move. This save time can be invested towards research, analysis and just living life. On the top of this, add an option of paper trading for forward testing or validation of the backtest results, there nothing more to ask for.
Here are my paper trading results for this algo. It was running live for about a week and captured 2 successful opportunities post the 20% drop in BTC prices around 24th Feb 19.
Note : All the analysis, backtesting, paper trading(also my live trading) mentioned here is carried out using streak.world. The platform is currently in in beta, you can signup here.
Platforms and sources :
https://www.streak.world
https://discuss.streak.world
https://medium.com/streakworld/introducing-streak-world-algo-trading-without-coding-a4d74eb706df
