WITH roa_table AS (
SELECT
symbol,
report_date,
MAX(CASE WHEN item_name = 'net_income_common_stockholders' THEN item_value END) AS net_income_common_stockholders,
MAX(CASE WHEN item_name = 'total_assets' THEN item_value END) AS total_assets
FROM
'{stock_statement}'
WHERE
symbol in ({symbols})
AND item_name IN ('net_income_common_stockholders', 'total_assets')
AND report_date != 'TTM'
AND period_type = 'quarterly'
AND finance_type in ('income_statement', 'balance_sheet')
GROUP BY symbol, report_date
),
base_data AS (
SELECT
symbol,
report_date,
net_income_common_stockholders,
total_assets,
YEAR(report_date::DATE) AS report_year,
QUARTER(report_date::DATE) AS report_quarter,
YEAR(report_date::DATE) * 4 + QUARTER(report_date::DATE) AS continuous_id
FROM
roa_table
WHERE
net_income_common_stockholders IS NOT NULL AND total_assets IS NOT NULL
),
base_data_rn AS (
SELECT
symbol,
report_date,
net_income_common_stockholders,
total_assets,
report_year,
report_quarter,
continuous_id,
ROW_NUMBER() OVER (PARTITION BY symbol ORDER BY continuous_id ASC) AS rn_asc
FROM
base_data
),
grouped_data AS (
SELECT
*,
continuous_id - rn_asc AS group_id
FROM
base_data_rn
),
base_data_window AS (
SELECT *
FROM (
SELECT
*,
MAX(group_id) OVER (PARTITION BY symbol) AS max_group_id
FROM grouped_data
) t
WHERE group_id = max_group_id
),
equity_with_lag AS (
SELECT
symbol,
report_date,
net_income_common_stockholders,
total_assets as ending_total_assets,
LAG(total_assets, 1) OVER (PARTITION BY symbol ORDER BY report_date) AS beginning_total_assets
FROM base_data_window
),
net_incomet_and_avg_asserts AS (
SELECT
symbol,
report_date,
net_income_common_stockholders,
(beginning_total_assets + ending_total_assets) / 2.0 AS avg_asserts
FROM equity_with_lag
WHERE beginning_total_assets IS NOT NULL
),
pivoted AS (
SELECT *
FROM net_incomet_and_avg_asserts
PIVOT (
ANY_VALUE(net_income_common_stockholders) AS net_income_common_stockholders,
ANY_VALUE(avg_asserts) AS avg_asserts
FOR symbol IN ({symbols})
)
)
SELECT
*
FROM pivoted
ORDER BY report_date;