Rosalie:
I am going to say there is a misunderstanding of the basics of forecast and the difference between direct and indirect demand.
The usage of forecast is to prevent "out-of-stock" situations by creating "indirect demand" so the machinery is put to work to satisfy upcoming demand. The longer your manufacturing lead times the more important because you want to plan ahead.
Forecast is also used to "smooth" requirements given to manufacturing so that manufacturing does not have to be changing production planning. Basically you decide to use a "statistical" control instead of using direct demand.
Reorder levels and forecasting are examples of statistical control of inventory. The usage of MP WO orders is heuristic management with human intervention (thus master planners) where planners "risk" and "commit" infraestructure to certain schedule to be ahead of future direct demand (SO, DRP, etc).
Let's say you have a lead time ("to promise") of 45 days to build some equipment. If we were an organization that solely relies on MRP (pure MRP), in theory a customer would never get what they are requesting before those 45 days (ignore shipping). The only way this could be solved is if our customers send us their requirements with 45 or more days in advance. In this utopia, forecasting would not be necessary since you have real demand by means of future SO.
Now, what would happen if we are shop that reacts solely on real demand (that is, sales orders)... it is possible that our workload goes to the roof and that our capacity is not able to support that demand on a given period (e.g. week). We would end up with weeks with little or no production and then we would have weeks where we would not meet the demand (and thus, cost of opportunity, extra cost on extra hours and decrease in customer satisfaction and so on).
So, MRP has to run in combination with statistical control and/or human planning.
If you are familiar with project planning think of the "critical path". First we have to control the bottlenecks and this includes both items (purch or manuf) and processes/operations that have a direct connection to the critical path of a product. We need to have stock that can absorb demand. Planners need to determine what is coming. They decide on a combination of factors such as: what customers say will happen in the near future (but still not a hard direct demand), the market (e.g. known shortages of certain components like chips), maintenance schedules, consolidate production quantity through "economic order quantity" lots to avoid line changes and so on.
A planner can use direct WO to create stock and start the planning and moving the wheels through MRP or can use FCST to get the same effect. In this case these inputs should/can be seen as "indirect demand".
Now, here comes the point... (assuming a B/F of 0/0) if in a series of periods (weeks) the forecast is for 100, then we are saying that every week we want to manufacture that much regardless of how much our customers order. In theory FCST has been smoothed so that combining our current stock levels and the forecasted production we will meet/satisfy customer demand, or at least we expect.
If for a given week we get SO qty that exceeds the forecast it will use the highest of the two (again, 0/0). Bwd/Fwd consumption of forecast will work here to absorb that variability and smoothen direct demand. If we used B0/F0 we would end up manufacturing the higher of the two (FCST or SO). When have B/F different from zero, our production will be smoothed so we do not have huge "excess requirements" and that they are distributed against back and fwd periods. Basically it states: mmmmh there is mode direct demand for this week than what we had forecasted... how did we do against our forecast on the last 2 weeks... we did not reach the forecast... then instead of requesting this "extra" from the production department let's say it could be covered with last week's unconsumed forecast.
OK, so what is the point here? Is there one?
As you say, on a 0/0 you WOULD end up manufacturing the excess qty every week.
But since you are starting with forecast what will be the effect of putting forecast for October while you already are in november? Or are you saying you want to put that forecast today? What would be the effect of entering forecast for october once we are in november?
Let's say you add forecast to October just to accomplish backward consumption (e.g. 100). Let's say your orders for the first week of november are 120. QAD will first go back and then to the future. What will be effect on planning and inventory levels if in the end those 20 can no longer be produced in october?
If the forecast for the october is higher than the actual orders for those weeks, MRP will produce MRP actions as overdue. If the forecast is same or lower than the actual SO for those weeks, MRP won't say a thing (unless they have not been already met through inventory availability, WO or PO).
So basically, the effect BW will have in NOVEMBER for forecast entered against OCTOBER is:
If the forecast entered for those weeks in october is higher than the actual demand, then it will end-up requesting extra.
If the forecast entered for those weeks is same or smaller than actual demand for those periods, then direct demand will be used. There should be no difference if the range is (0 to actual demand).
If there is excess demand on november and the FCST is zero for october then the excess will go against future weeks.
In case I created confusion with any references to "zero" in my last email, I am not saying the B/F should be zero; it should be what your company thinks it should.
Your original question was whether to setup forecast for october once in november. My response is that:
If you setup a forecast in october and this forecast becomes larger than your direct demand (SO) it will trigger an overdue action message. Do you really want to manufacture that? (Your planners would know whether they should) if so, it means that october's plan was short and that in fact you need to catch-up. On the other hand, if this is not the case, then there is no purpose for entering forecast for october.
Bwd/Fwd is used to smoothen direct demand so we do not end up with high peaks on a given period and the rationale here is that if we get a "large" order it may be possible it is because the customer did not request it last week and/or that maybe it is requesting more than it normally does but next week is going to require less. When B/F are used, then it will not be the max(real demand, FCST) for the period unless both prior and future "buckets" own forecast has run out. It will be the max(FCST, total demand less what could be allocated against other B/F week buckets).
I hope this helps you.
.