Hi Eric,
Costing in OpenERP is oversimplified IMHO.
It can be good enough for one Company with one Warehouse
without Landed Costs scenario.
For services property field could solve part of the
multi-company problem.
Cost price, stock valuation method, cost price calculation,
accounts for posting, (in some cases Sale price also),
Sequences for Picking (I found 7)
are all related to a Warehouse NOT Product/Company.
As Warehouses are defined for a specific Company,
multi-company is automatically solved.
More :
- Posting should be bound to a picking, not individual move
- same as Invoice (1 Document== 1 account_move)
- Picking is a document (legal), not arbitrary collection of
stock moves from somewhere to somewhere
- Whole Picking must be related to one warehouse. 2 new
fields on Picking: warehouse_id, and contra_location_id.
All Picking stock_moves must go from/to warehouse main
location (or childs) to/from "contra_location" (or child of
contra_loc)
For posting one account is defined on warehose, other on
contra location. I don't use Product stock_input... Doesn't
make sense.
- Between two warehouses there should be picking OUT from 1. Wh
to inter-company virtual "transport" location, and strongly
related picking IN 2. warehouse(with LC).
- Technically we could support only one picking between 2
warehouses inside one company, but I found it too complicated to
implement
- There is a need for a strong (many2...) link between
stock_move(s) and invoice line(s)
- For each "In" picking company could have Landed Costs
(planed, internal, awaiting supplier invoice for a months)
- Production(virtual) -> Prod.Plant(internal in) add
direct LC same way as In picking
- Prod.Plant(internal) -> Finished Goods(internal) add
Costs
- Supplier Invoices for goods and landed costs should create
"corrective" additional posting
adjusting stock value for the goods still in warehouse,
and COGS for sold goods
- Program needs to track all that using FIFO method by
warehouse even if costing method for a Warehouse is Average
I think that FIFO with Landed Costs by warehouse should be
done first. Then Average is trivial.
I wouldn't support LIFO, but there is Specific
Identification method - can be based it on Lots,
and Plan Costing - can be based on special Price lists
(multi-company/currency)
- Margin calculation is now based on current cost price from
product :( ouch
- Physical Inventories should produce:
- Inside picking for quantities diff inside one warehouse
(without accounting posting)
- Surplus picking (new price and comment)
- Loss picking
- Tax allowed Scrap
- Scrap that is not Tax allowed
- Internal Picking between warehouses in same company
- There is a need for special documents for initial state and
price nivelation
- In some(all?/most?) countries selling below cost price is
tricky (VAT!)
- Stock can be valued by cost, sale, retail price
- In some countries "Storno" is mandatory (negative qty and
posting)
- COGS can be taken at Picking Out or Invoice
- Ideally costs should be calculated from account_move_line
(like fixed assets)
to avoid "stock accounting" and "real accounting" discrepancies.
I am not 100% sure.
Right now I am using:
create or replace view account_warehouse_product_sum as (
SELECT min(
aml.id) as id ,
sw.id
as warehouse_id, aml.product_id
...
,CASE WHEN abs(sum(aml.debit_qty -
aml.credit_qty)) > 0
THEN sum(aml.debit - aml.credit)
/ sum(aml.debit_qty - aml.credit_qty)
ELSE 0.000 END as avg_price
...
FROM account_move_line aml
JOIN stock_warehouse sw ON (sw.account_id
= aml.account_id)
WHERE am.state='posted'
--AND
sw.id = %s AND
aml.product_id = %s
GROUP BY
sw.id , aml.product_id
I am from "storno accounting" country and there are more
special issues about retail store accounting, returns,
unnecessary margin and VAT posting ...
All over Launchpad/Git you can find attempts to deal with
parts of a Warehouse costing problem.
Fabian diagnose is absolutely correct:
Some
community members started basic implementations of FIFO/LIFO
but, as they don't rely on strong costing foundations, this
will not work in all situations (lots of side effects).
How important is it?
Top 3 idea:
Top accounting idea:
I invested 2+ months developing "warehouse_base" module for
6.1 that implements described functionality.
Plan is to clean/refactor the code and break it to series of
modules before publishing on LP.
Is there anyone interested in "Costing by Warehouse" aproach?
Goran Kliska
Slobodni
programi d.o.o., Croatia
