Django .annotate()

Let's imagine this database. Here we have some records about cryptocurrencies. Notice that we have two records of SLP token.

I need to totalize the invested_amount_usd by sum all the occurences per currency. For example, my two SLP records must sum $250.

Here Django appears, and its pretty simple. I tried to annotate like this, but look that the query has no sense at all...

>>> print(Order.objects.annotate(test=Sum('invested_amount_usd')).query)
SELECT
    "order_order"."id",
    "order_order"."owner_id",
    "order_order"."order_date",
    "order_order"."currency",
    "order_order"."invested_amount_usd",
SUM("order_order"."invested_amount_usd")
AS "test" FROM "order_order"
GROUP BY
    "order_order"."id",
    "order_order"."owner_id",
    "order_order"."order_date",
    "order_order"."currency",
    "order_order"."invested_amount_usd",

Because I tried to GROUP BY every columns of my table, and all records have different data the ORM can't group the records.

>>> print(
    Order.objects.annotate(test=Sum('invested_amount_usd'))
    .values('test')
)
<QuerySet [
    {'currency': 'BTC', 'test': 5000.0}, 
    {'currency': 'ETH', 'test': 700.0}, 
    {'currency': 'SLP', 'test': 50.0}, 
    {'currency': 'DOT', 'test': 5.0}, 
    {'currency': 'SLP', 'test': 200.0}
]>

So, I tried using .values() before annotate.

>>> print(
    Order.objects
    .values('currency')
    .annotate(total=Sum('invested_amount_usd'))
    .query
)
SELECT
    "order_order"."currency",
SUM("order_order"."invested_amount_usd")
AS "total" FROM "order_order"
GROUP BY
    "order_order"."currency"

And this is the result!. I got 250 for SLP token.

>>> print(
    Order.objects
    .values('currency')
    .annotate(total=Sum('invested_amount_usd'))
)
<QuerySet [
    {'currency': 'BTC', 'total': 5000.0}, 
    {'currency': 'DOT', 'total': 5.0}, 
    {'currency': 'ETH', 'total': 700.0}, 
    {'currency': 'SLP', 'total': 250.0}
]>