Check out my latest app — TradingGYM. It’s a trading simulator that helps you practice trading with a much faster feedback/learning loop and minimal lookahead bias. Once you open the app, you will see a chart with a random asset at a random point in time. Make trades and fast-forward time to see how it played out.
This is Part 1 in multi part series:
- Part 1: Basic strategies, introduction, setup and testing vs June-July market.
- Part 2: Advanced strategies and where to find them, testing vs June-July market.
- Part 3: Basic and Advanced strategies testing vs August market.
- Part 4: Neural Network strategies description and backtests against September market.
- Part 5: Neural Network strategies backtests against October market.
- Part 6: Did Neural Network strategies predict November 14th price drop?
- Part 7: Crypto Trading 2018 in Review: 17 Advanced + 15 Neural Net strategies tested
- Part 8: Intro to Statistical Arbitrage in Crypto — Pairs Trading
- [NEW] Part 9: Crypto Trading 2019 Half Year Review: 17 Advanced + 15 Neural Net strategies tested
Last 2 months have been rocky for Crypto. In mid June we were hit with some of the lowest BTC prices we have seen in a long while, going down around 5500 USD, and since most of the coins are following BTC chart, they saw a similar drop, some more some less. After spending some time in a dark valley under 6000, mid July brought some promise and BTC climbed up well over 8000, hitting 8400 in it’s highest point before, only to drop to 7500s a week later. Now we are even lower in 7000s again.
I thought it would be a good time to take a step back and look at how the markets have done from trading standpoint and what are the chances of making a profit with basic trading strategy. In trading, I only believe in strategies that you can backtest, so I will be using a few basic strategies, gather data for some coins and backtest them for the time period from June 1 to July 31.
With basic I mean a strategy that you run into when you first start to learn about trading, like RSI, MACD, etc. Since I will be using Gekko trading bot, I will take the basic strategies that comes packaged with source, which is basically this list. They will be used with default parameters, which you can see here.
I decided I will use TOP20 from coinmarketcap based on market cap and import COIN-USD or COIN-USDT price data to keep it simple. When I tried to gather the candle data necessary for backtesting, I discovered that not all coins are so easily accessible for importing (getting data through exchange API), for example, Tezos (Nr. 18) is only found on some relatively exotic exchanges (see here) so I decided I will just skip these coins, which are:
- Tether (USDT) itself
- NEM (XEM)
- Tezos (XTZ)
Just to be clear, I have a bit old Gekko version and I also have a lot of local modifications made, so my results will differ a bit (5% based on some testing) from what you would see if you tried the same, but I think the difference is negligible in this context.
For candle sizes, I will use 5, 10, 30 and 60 minutes. Since 10m seems like the most reasonable, I will start with that. So each strategy gets backtested with each coin for date range of June 1 — July 31. Which makes it 9 strategies X 20 coins = 180 samples for each chart below.
10 minute candles
Looks bad. Of course, the market is going down, but I still hoped to see something better than getting my account wiped (mostly) clean in just 2 months by following a basic strategy like this.
Two things to notice here.
- We have an interesting outlier — VEN with DEMA strategy. While pretty much everything we have is 50–90% down, VEN with DEMA is up 31%. I find a VEN chart for Jun-July and to my surprise it’s mostly down for the time period. I load the strategy into Gekko to make sure backtest was correct and yes, this is what I see.
Unfortunately, seeing other results for DEMA, is have a feeling this is just pure luck, no magic here. Let’s move on.
2) Leading Reversal strategies (CCI, RSI, TSI) seem to be doing relatively better here, which makes total sense, since we have a BEAR market and these kinds of strategies are theoretically better at that, at least on paper.
Ok so 10m candles are pretty much a no go, let’s go lower then.
5 minute candles
There is not much to comment here, this is even worse, let’s try the other way.
30 minute candles
A bit better, but still — a recipe for disaster. A few outliers here and there but overall pretty hopeless.
60 minute candles
For the longest candle size in this experiment, there simply were very few signals generated, which results in half the strategies not making a single trade. RSI is by far the best from the worst in this one, and with some parameter tweaking we could even get it profitable on average on all coins.
Quite obviously, this is not the way to trade in the current market. Keep in mind these are just the most basic strategies with default parameters.
I’m not saying that it’s not possible to be profitable with these strategies. Even though default parameter results are bad, we could find some combination that would return positive profit in backtests, but would that mean strategy itself is profitable and has predictive nature, or is it just pure luck? Most probably the latter.
The thing is, you need to understand what you are doing before getting into this. You got to start somewhere and these are good strategies to start with just to get the feel and learn the basics, but once you have done that you need to look forward and learn something more advanced.