Constant Position Sizing

[systats]

Hi great package! I wondered whether it is it possible to have constant position size of 100$ to prevent compounding metrics and to interpret total profits in percents? Thanks for suggestions.

[polakowo]

Pass size_type=‘value’ (https://vectorbt.dev/docs/portfolio/enums.html#vectorbt.portfolio.enums.SizeType)

[systats]

Hi Polakowo,

I compare different backtesting tools to spot bugs and to increase trust into our open source work. So far I got 4 packages and we can sparsely compare the metrics. First of all I very happy that there is already a lot of constituency. Some questions remain though:

1.) The biggest difference so far is drawdown. Any ideas why this metric is all over the place? 2.) Is there any chance that you can expand the metrics to include for example return_mean, equity_last, equity_max, win_ratio, or win_sd/loss_sd? I will do the same for backer. I tried Pass size_type=‘value’ but it didnt change anything. Although beside equity_cum nothing else seems cumulative.

        var         backer     quantstrat  vectorbt      PA
1  equity_last        307.08     297.04        -       -
2   equity_cum       1231.55          -  1231.55 1231.55
3   equity_max        310.81     311.32        -       -
4    return_sd          8.21       8.27        -    8.21
5  return_mean          2.01       2.07        -    2.01
6     time_exp          0.41          -     0.41       -
7    down_mean         -5.28          -    -1.05   -8.84
8     down_max        -19.76     -43.08    -30.2  -18.52
9    win_ratio          2.20       2.97        -       -
10    win_rate          0.44       0.43     0.44       -
11    win_mean          8.41       8.81     2.01       -
12     win_max         43.56      44.22    43.61       -
13      win_sd          8.68          -        -    8.68
14     win_sum        563.62     572.37        -       -
15   loss_mean         -2.98      -2.96        -       -
16    loss_sum       -256.54    -257.94        -       -
17     loss_sd          2.09          -        -    2.09
18    loss_max         -9.74      -9.87    -9.75       -
19     loss_05         -6.32          -        -   -5.67
20     sortino          0.96          -        -    0.74
21      calmar          0.38        6.9        -    7.62
22    n_trades        153.00        152      153       -

There are a few bugs also on my side so I have also to fix some stuff. The following metrics are already in the dictionary in case you spot an issue.

equity_cum = total_profit
time_exp = gross_exposure
down_max = -max_drawdown_percent
down_mean = -avg_drawdown_percent
win_max = best_trade_percent
win_mean = avg_trade_percent
loss_max = worst_trade_percent
win_rate = win_rate_percent/100
n_trades = trade_count

Thanks for your helpful answers.

[polakowo]

@systats

• what is backer?
• I assume drawdown metrics differ because of difference in starting capital and/or position size (more exposure leads to a greater risk) between packages. The value size type has no effect because you probably use the size of 100$ which uses all starting capital (which is $100 by default in vbt). Setting init_cash to ‘auto’ is going to determine the optimal starting capital for all transactions to go through. Or simply increase init_cash manually. But then you should set up experiments with the same parameters for other backtesters too in order to compare them fairly.
• It’s quite easy to compute those metrics. Last equity is pf.final_value() in vbt, SD of wins is pf.trades.winning.pnl.std(), and so on. Refer to the documentation, and also see vectorbt.portfolio.base.Portfolio._metrics under the development branch, it shows how to query win rate and other interesting metrics.
• What is this dict? I need some context.

Sorry for any typos, I’m not near my computer right now.

[systats]

• Backer is just another backtest tool that we are going to release after testing. Vectorbt is perfect to compare against.
• I will have a look into calculating the additional metrics - okay its great how easy its working!
• Just my internal reference of metrics

I am on the way to normalize all experiments and so far I really like vectorbt.