Thread by @santimentfeed

Are some days better than others for trading @bitcoin and other #crypto?

If so, how does trading $BTC only on (historically) the best days compare to simply #HODLing?

Check out our full analysis here! tinyurl.com/y6z55ulw

[explanation continued in comments]

2/ To answer the above, we started by calculating the average $BTC price returns for each day of the week, including weekends.

Here are the average log returns for Bitcoin from 2017 until present day:

3/ We can see major differences between days:

Monday and Saturday emerge as clear winners, with average returns of +0.69% and +0.5%, respectively.
Wednesday, Thursday and Sunday all returned neg. values for the past 2 years. To date, Weds proved to be the worst performing day.

4/ With this info, we can now try and backtest a day-sensitive $BTC strategy.

Here’s how our $BTC portfolio would have performed if - starting in 2018 - we only bought @Bitcoin on Mondays and Saturdays and sold at the end of the day:

5/ In this time frame, our strategy would pocket us a cool 68% profit, while HODLing #Bitcoin results in a 38% loss.

Unfortunately for our model, the infamous drop at the end of 2018 happened to be on a Monday, making the comparison a bit tighter than it would’ve been otherwise.

6/ Looking further in the past provides interesting results as well.

Backtesting the same day strategy, only this time from 2017 onwards, gives the following outcome:

7/ This time, our strategy does not outperform #HODLing, which is not too surprising considering that there were relatively few bad days for #Bitcoin in 2017.

In this time frame, our day strategy would’ve returned 469% profit - or about 71% of our benchmark’s (HODLing) returns.

8/ While that sounds like a failed experiment, it is worth noting that we get 71% of the returns by only being in #BTC for two days of the week.

Also we can see there’s now a *massive* difference in volatility between our strategy and #HODLing.

9/ The annualized standard deviation of our day strategy for the observed time frame was 0.43; for holding it was 0.84.

Annualized Sharpe ratio of our strategy was 1.65; for holding, it was 1.35

The strategy reduced volatility substantially and improved risk-adjusted returns.

10/ So far, however, these backtests are riddled with hindsight bias. (investopedia.com/terms/h/hindsi…)

To eliminate it, instead of testing our day strategy back, let’s try and test it ‘forward’ - starting in 2018, and using data from the previous 2 years (2016-2018)

11/ Below are the 2016-17 average log returns for Bitcoin.

We actually get similar outputs to before, if you consider $BTC’s bullish tendencies throughout 2016.

The best performers are again Monday, Thursday and Saturday. Wednesday & Sunday are now also returning + values.

12/ With this info back in 2018, we might have decided to trade #Bitcoin only on Mondays, Thursdays and Saturdays - and sell at the end of each.

Here are the results of that day strategy from January 2018 onwards:

13/ With only past data to work with, the strategy again outperforms #HODLing $BTC while making only a small loss of -4% compared to the -38% of holding.

Again, the analysis shows how some days of the week can be more auspicious for #crypto traders than others.

14/ However, while #BTC is a decent proxy for market, the best days might be different for other coins

That’s why we created a ‘daily performance’ template, where you can find the best trading days for any coin in the Santiment database - docs.google.com/spreadsheets/d…

15/ Download the #Sansheets plugin (sheets.santiment.net) so you can import #Santiment data into Google Spreadsheets, then use this pre-made template to understand which days have had the best historical returns for any coin.

16/ One big caveat - this analysis doesn’t take into account one obvious downside of any active trading strategy - transaction costs.

Those costs would chip away at the projected returns, narrowing the gap between the outlined strategy and our benchmark in reality.

17/ Trading on specific days is obviously not the holy grail of strategies. Rather, it's complementary to other models.

If you’re already an active trader, it could help offset some of #crypto’s volatility, and warn you of days when odds have been stacked against in the past.

18/ Read the full analysis here, and please note that this is NOT financial advice. tinyurl.com/y6z55ulw