Validators

Validators in Dejavu can be seen as a harness. While they could replicate the logic in your strategy, we strongly suggest using Validators as a safety harness.

Validators don't stop the signal generation of a trade as logic in your trading Strategy would, but they do stop the placement of orders if conditions aren't met.

For example, if you had a cash-only long-only strategy, you can also validate this by going:

from dejavu.execution.orders import SimulatedExecutionHandler
from dejavu.execution.validators import CashValidator, ShortValidator

slippage = ...
commission = ...
executor = SimulatedExecutionHandler(slippage, commission, validators=[CashValidator(), ShortValidator(allow_short=False)])

This way, if your strategy logic happens to generate a short signal, the signal will still be generated, but the order will not be executed.

Dejavu provides a few validators out the of the box, but we suggest creating your own to suit your business or investment needs.

Extending Validators¶

Validators being composable in nature make them really helpful for maintaining a level of predictability and safety.

They also have access to your Portfolio, so you can get quite advanced with the validation you want. Perhaps ensuring you're not overweight in a stock or asset class for example, or ensuring you're not making too many trades in a day.