-
|
Hello! |
Beta Was this translation helpful? Give feedback.
Replies: 3 comments 4 replies
-
|
@Sanjan611 that's correct. Generally, each column in vectorbt is a separate backtesting instance, so having multiple columns shouldn't be considered a single portfolio, but a set of separate portfolios, each with its own initial capital. To specify which columns belong together, vectorbt introduces the idea of grouping using |
Beta Was this translation helpful? Give feedback.
-
|
Great, thank you! A different question on
10000
that same notebook. If I run An extension to that question - for |
Beta Was this translation helpful? Give feedback.
-
|
cash of zero indicates that all cash has been used. This makes sense since you allocate all your capital at the very first timestamp. You can read on |
Beta Was this translation helpful? Give feedback.
-
|
Thanks a lot! I really like this library and expect to be using it quite a bit in the upcoming weeks, so I'll try and support it where I can. |
Beta Was this translation helpful? Give feedback.
-
|
I place entries and exits on every month first day. 10 stock of total 5000 . my settings: init_cash = 1000
fees = 0.002 # in %
vbt.settings.portfolio['freq'] = 'd'
vbt.settings.portfolio['init_cash'] = init_cash
vbt.settings.portfolio['fees'] = fees
vbt.settings.portfolio['slippage'] = 0
vbt.settings.returns['year_freq'] = '252 days'
pf = vbt.Portfolio.from_signals(
close, vbt_entries, vbt_exits, sl_stop=None,tp_stop=None,
init_cash=10_000,
group_by=True,
cash_sharing=True,
)close and entries preview: but vbt didn't share the cash: I have set |
Beta Was this translation helpful? Give feedback.
-
|
work out, what I was missing is size_type='targetppercent' . |
Beta Was this translation helpful? Give feedback.
-
|
Realized that |
Beta Was this translation helpful? Give feedback.
@Sanjan611 that's correct.
Generally, each column in vectorbt is a separate backtesting instance, so having multiple columns shouldn't be considered a single portfolio, but a set of separate portfolios, each with its own initial capital. To specify which columns belong together, vectorbt introduces the idea of grouping using
group_by. For example,group_by=[0, 0, 1]for three columns will build one group with two columns and one with one column (a separate number ingroup_byindicates a separate group). When you docash_sharing=True, cash is shared among the columns of each group (in the above example, the first two columns would share $100 and the last column also $100). For cash sharing…