"Trading range"

 It's been difficult to follow the cryptocurrency trend in the last couple of weeks. I was curious as to what the Profit and Loss data looked like on the exchange, so copy-pasted some html, ran my auto-formatting script and got this:

import pandas as pd
df = pd.DataFrame(columns=['Type', 'Amount', 'Asset', 'Symbol'],
    data=[['Realized PNL', 16.69441832, 'USDT', 'BTCUSDT'],
          ['Realized PNL', 60.09990598, 'USDT', 'BTCUSDT'], ...],
    index=['2021-03-29T00:25:30', '2021-03-29T00:25:30', ...])

Ok, we need to convert the index to timestamps instead of strings and pull in finplot to do some plotting. finplot requires the time series sorted or it refuses to plot.

import finplot as fplt
df.index = pd.to_datetime(df.index)
df = df.sort_index()

Then we resample to get daily, plot a bar and show it.

df.Amount.resample('D').sum().plot(kind='bar')
fplt.show()

Yep. A sad song.

But innovation requires pressure, and I'm no exception. I'm keeping the long-term strategy the same — it will be effective when we're out of this trading range. But for the short-term/high risk strategy I decided to run a completely different algorithm. I used to run the same algorithm for both, but with more aggressive settings for the short-term one. Now I run "Poem" on BTC tweaked for the last 3.5 years and "Twerk" on ETH tweaked for the last 3.5 months. Poem's starting wallet is 10x that of Twerk, but I count on Twerk catching up in 2 months, as long as this trading range prevails.