SQL Server 2017

By using db<>fiddle, you agree to license everything you submit by Creative Commons CC0.

CREATE TABLE dbo.historial
    ([fecha] date, [codigo_empleado] varchar(4), [sueldo_mensual] int)
;
INSERT INTO dbo.historial
    ([fecha], [codigo_empleado], [sueldo_mensual])
VALUES
    ('2019-01-01', 'E001', 250000),
    ('2019-02-01', 'E001', 550000),
    ('2019-01-01', 'E002', 350000),
    ('2019-02-01', 'E002', 450000),
    ('2019-01-01', 'E003', 50000),
    ('2019-02-15', 'E003', 150000)
;
6 rows affected

WITH CTE AS
(
	SELECT	*,
			ROW_NUMBER() OVER(PARTITION BY codigo_empleado ORDER BY fecha DESC) RN
	FROM dbo.historial
)
SELECT	fecha,
		codigo_empleado,
		sueldo_mensual
FROM CTE
WHERE RN = 1
;


fecha
codigo_empleado
sueldo_mensual


01/02/2019 00:00:00
E001
550000

01/02/2019 00:00:00
E002
450000

15/02/2019 00:00:00
E003
150000

fecha	codigo_empleado	sueldo_mensual
01/02/2019 00:00:00	E001	550000
01/02/2019 00:00:00	E002	450000
15/02/2019 00:00:00	E003	150000