Backtests

A backtest is an automated statistical analysis of the impact of events on price.

BETA version available to all users (Tier 1, Tier 2, Tier 3)

Why are backtests important?

Strategy Enhancement: Allows for refining and optimizing strategies.
Understanding Risk: Provides a clear picture of potential losses and gains.





A backtest enables checking how many times a given alert or Pro Alert would have been triggered in the past.
The result of a backtest is two lines depicting the statistical price change before and after triggering the alert: the average and the median.
Backtesting can be performed on a single condition or on several simultaneously.

How to do a Backtest?

Select Pro Alert components.
Build your Pro Alert.
Initiate Pro Alert backtesting.
Choose a period for backtesting.
Begin the backtest.
Await the outcome.
Review the trigger frequency of the Pro Alert.
Examine the results chart:
Blue line - average price change
Green line - median price change

How to estimate the potential value of a Pro Alert?

To estimate the potential value of a Pro Alert, conduct its backtest and choose as broad a time range as possible.
Knowing the historical number of triggers in that time range, and the statistical price deviations after triggering, we can estimate what potential profit a given Pro Alert might bring in the future.

For example:

If the backtest result shows that a certain event was triggered 10 times over 8 months, and statistically the price fell by 5% within 4 hours, then started to rise, we can assume that there is a high probability that in the coming year the event will also be triggered about 12 times, and the statistical drops will be similar to the previous ones.

Value of Pro Alert = ( power ( 1 + 1 x √ -5% ² ; 10/8*12 ) * 100 - 100 ) %

Example Result: 107%

The above result may suggest that by playing the given Pro Alert, a trader has a chance to gain an additional annual profit of about 107%.

Actual results may vary from predictions.

Formula to calculate the potential value of a Pro Alert:

Value of Pro Alert = ( power ( 1 + 1 x √ Statistical deviation ² ; Number of repetitions in a year ) * 100 - 100 ) %

Types of Deviations

Drop

Minimal deviations before, maximum downward deviations after the alert (opposite to the final deviation on the right side)

Rise

Minimal deviations before, maximum upward deviations after the alert (opposite to the final deviation on the right side)

Peak

Maximum downward deviations before, maximum downward deviations after the alert (in the same direction as the final deviation on the right side)

Peak in an Uptrend

Maximum downward deviations before, maximum downward deviations after the alert (opposite to the final deviation on the right side)

Trough

Maximum upward deviations before, maximum upward deviations after the alert (in the same direction as the final deviation on the right side)

Trough in a Downtrend

Maximum upward deviations before, maximum upward deviations after the alert (opposite to the final deviation on the right side)

Configuration of Backtests is available from Pro Charts

Updated on: 15/09/2024