I would like to know which of the following implementations of a trading strategy would result in the fastest response time (from a processor loading standpoint) of a trading rule. The Trading Rule Indicator will be applied to a chart with 1 second periodicity with the assumption that the rule will be evaluated once per bar (= once per second).

The strategy buys 1 contract when price equals the value of a moving average. The moving average can either be implemented on the 1 second chart using the MPD token, or a reference line referenced back to a V# variable containing the moving average value.

The first case (using the MPD token) seems like it might be faster since the calculation is the only step necessary to get the moving average value.

In the second case, the moving average value is first calculated on a second chart (happens to be a PNF chart), then stored in the V# variable, and then retrieved from the variable to be used in the trading rule.

Thanks.