This post is not about the crypto data site: https://cryptoquant.com, here we’re using ‘Crypto Quant’ descriptively. It’s a simple way to describe the programmatic work of understanding crypto currency movements.
‘Quant’ on Crypto is more interesting, in certain ways, than regular Quant:
- the crypto market is ON 24×7, no overnight holds, no holidays
- crypto trading fees can be minimal (if you know how)
- the volatility and volume on BTC and ETH are significant (turbulence)
https://www.binance.us/ (for US residents)
The site and app are a great way to get into crypto currencies. You will need to verify your identity in a multi-step process. The signup process is similar to Coinbase and other similar ‘wallet’ apps.
It makes intuitive sense that you can buy US$ for crypto coins on these ‘wallet’ apps. What isn’t necessarily obvious is that they encourage you to ‘spot’ trade one crypto currency for another, this is how you achieve low trade commissions!
Spot-trades
Trade commissions of as low as 0.0075% of trade with no minimums.
Here’s a USD purchase of ADA coin, notice the 0.5% commission fee.
And here’s a spot-trade of BTC for ADA with BNB transaction fee, this works out to 0.075% (25% discount on 0.1% spot-trade fees) explained here.
This is a game-changer for programmatic trading: no minimum commissions. To achieve this you need to use another coin (rather than US$ in your account and accept that the commission will be paid in Binance coin (BNB). To put this into perspective: a $1,000 trade would carry $0.75 commission when done as a spot trade.
So now we have a way to trade with nominal fees. Just make sure you have sufficient BNB (Binance coin) for the fees.
But what if we want to exit positions to ‘cash’ in between trading sessions? This takes us outside the volatility as we tune our models and re-calibrate. If we leave our funds in BTC the volatility could be a problem during this time.
The answer is crypto coins ‘tied to’ US$, on Binance at the time of this piece there are 3 such coins: BUSD, USDT, USDC
You can see this in the upper right corner of the ‘basic’ trading desktop view: https://www.binance.us/en/trade/BTC_USD
So we have a way to go from US$ to a ‘stable’ coin tied to US$ from which we can trade to/from various crypto coins with nominal fees. Psyched.
What else does a Crypto Quant hacker need? Data!
Data
Quant is nothing without lots of data! Fortunately Binance makes it easy for account holders (remember all the verification steps you were forced to go through) to get historical trade data without charge.
Get all the Crypto historical data you need free of charge.
In the pull-down menu below your name see ‘API Management’ screen. Creating a new API is easy.
And now you are ready to Quant. Let’s switch our attention to Python and Jupyter notebook where all our Quant dreams can come true! If you are not familiar with this environment, take some time to get into it, it’s an essential tool for Quant coding and Python data-science in general.
Quant is essentially data-science applied to market data.
Let’s get minute data for BTC from Jan 1’st 2021 in CSV format, for convenience I’ve uploaded it here.
Several other intervals are available, see Binance API docs.
We are now ready to analyze crypto trading data, we can backtest strategies, analyze model profitability and trade on the platform.
Part 2 of 3 is here.
- 000
- 7
- Account
- ADA
- All
- api
- app
- apps
- binance
- Binance Coin
- bnb
- BTC
- BUSD
- buy
- charge
- Coding
- Coin
- coinbase
- Coins
- commission
- Creating
- crypto
- crypto data
- Crypto Market
- crypto trading
- currencies
- Currency
- CZ
- data
- Discount
- dreams
- Environment
- ETH
- EU
- EV
- Exit
- Fees
- format
- Free
- funds
- General
- great
- hacker
- here
- How
- hr
- HTTPS
- ia
- IP
- IT
- LG
- Market
- medium
- model
- Other
- perspective
- platform
- profitability
- purchase
- Screen
- sense
- Simple
- Spot
- Switch
- time
- trade
- Trading
- transaction
- us
- USD
- USDT
- Verification
- View
- Volatility
- volume
- Work
- works