SQLクエリのパフォーマンス要因（異なる列の指定による変化）

SQLに似たような性質（データ型:decimal）の列に条件を置き換えてSQLクエリを回したところ、大幅にパフォーマンスが異なり、困惑しております。

前提・実現したいこと

クエリのパフォーマンスの悪化原因を明らかにし、パフォーマンスの大幅な改善を図りたい。

発生している問題・エラーメッセージ

SQLに似たような性質（データ型:decimal）の列に条件を置き換えてSQLクエリを回したところ、大幅にパフォーマンスが異なり、困惑しております。

DTP_POINTS
Decimal(16,０) null) - NOT ALLLOW NULL
処理時間５分

DTP_POINTS_BASIC
Decimal(18,０）not null) - ALLLOW NULL
処理時間２時間以上

該当のソースコード

２つあります

SQL
1違いは指定した列の違いになります。指定列が[DTP_POINTS]と [DTP_POINTS_BASIC] だけが以下のコードで異なります。
2
3DTP_POINTS
4Decimal(16,０) null) - NOT ALLLOW NULL
5処理時間５分
6
7DTP_POINTS_BASIC
8Decimal(18,０）not null) - ALLLOW NULL
9処理時間２時間以上
10
11**__処理時間５分__**
12
13WITH BASE AS (SELECT [DTP_MEMBER_ID] 
14FROM  
15[DWH_PP].[dbo].[DAILY_TRX_PLMS_2018] 
16WHERE  
17[DTP_COMPANY] = '001' AND 
18MONTH([DTP_TRX_DATE]) = '5' AND  
19YEAR([DTP_TRX_DATE]) = '2018' AND 
20[DTP_TRANSACTION_CODE] = 'POI') 
21 
22SELECT 
23(SELECT COUNT(DISTINCT [DTP_MEMBER_ID]) FROM BASE) AS 'UNIQUE_MEMBER_PREVIOUS_MONTH', 
24COUNT(DISTINCT BASE2.[DTP_MEMBER_ID]) AS 'UNIQUE_MEMBER_NEXT_MONTH', 
25SUM(BASE2.[DTP_AMOUNT]) AS 'AMOUNT_AT_FIRST_TRX', 
26SUM(BASE2.[DTP_POINTS]) AS 'POINT_ISSUE_AT_FIRST_TRX', 
27SUM(CASE WHEN BASE2.[DTP_AMOUNT] >= 100000 THEN 500 ELSE BASE2.[DTP_POINTS] END) AS 'BONUS_POINT_2X', 
28SUM(CASE WHEN BASE2.[DTP_AMOUNT] >= 100000 THEN 750 ELSE FLOOR(BASE2.[DTP_POINTS] * 1.5) END)  AS 'BONUS_POINT_3X', 
29SUM(CASE WHEN BASE2.[DTP_AMOUNT] >= 100000 THEN 1000 ELSE FLOOR(BASE2.[DTP_POINTS] * 2) END)  AS 'BONUS_POINT_4X', 
30SUM(CASE WHEN BASE2.[DTP_AMOUNT] >= 100000 THEN 1250 ELSE FLOOR(BASE2.[DTP_POINTS] * 2.5) END)  AS 'BONUS_POINT_5X' 
31 
32FROM ( 
33 SELECT
34 [DTP_MEMBER_ID],
35 [DTP_AMOUNT],
36 [DTP_POINTS],
37 ROW_NUMBER() OVER (PARTITION BY [DTP_MEMBER_ID] ORDER BY [DTP_TRX_DATE] ASC) AS 'TRX_COUNT'
38 FROM [DWH_PP].[dbo].[DAILY_TRX_PLMS_2018]
39 WHERE [DTP_COMPANY] = '001' AND
40 MONTH([DTP_TRX_DATE]) = '6' AND 
41 YEAR([DTP_TRX_DATE]) = '2018' AND
42 [DTP_TRANSACTION_CODE] = 'POI' AND [DTP_MEMBER_ID] IN 
43 (SELECT [DTP_MEMBER_ID] FROM BASE)) AS BASE2
44WHERE TRX_COUNT = 2 
45
46
47**__処理時間２時間以上__**
48
49WITH BASE AS (SELECT [DTP_MEMBER_ID] 
50FROM  
51[DWH_PP].[dbo].[DAILY_TRX_PLMS_2018] 
52WHERE  
53[DTP_COMPANY] = '001' AND 
54MONTH([DTP_TRX_DATE]) = '5' AND  
55YEAR([DTP_TRX_DATE]) = '2018' AND 
56[DTP_TRANSACTION_CODE] = 'POI') 
57 
58SELECT 
59(SELECT COUNT(DISTINCT [DTP_MEMBER_ID]) FROM BASE) AS 'UNIQUE_MEMBER_PREVIOUS_MONTH', 
60COUNT(DISTINCT BASE2.[DTP_MEMBER_ID]) AS 'UNIQUE_MEMBER_NEXT_MONTH', 
61SUM(BASE2.[DTP_AMOUNT]) AS 'AMOUNT_AT_FIRST_TRX', 
62SUM(BASE2.[DTP_POINTS_BASIC]) AS 'POINT_ISSUE_AT_FIRST_TRX', 
63SUM(CASE WHEN BASE2.[DTP_AMOUNT] >= 100000 THEN 500 ELSE BASE2.[DTP_POINTS_BASIC] END) AS 'BONUS_POINT_2X', 
64SUM(CASE WHEN BASE2.[DTP_AMOUNT] >= 100000 THEN 750 ELSE FLOOR(BASE2.[DTP_POINTS_BASIC] * 1.5) END)  AS 'BONUS_POINT_3X', 
65SUM(CASE WHEN BASE2.[DTP_AMOUNT] >= 100000 THEN 1000 ELSE FLOOR(BASE2.[DTP_POINTS_BASIC] * 2) END)  AS 'BONUS_POINT_4X', 
66SUM(CASE WHEN BASE2.[DTP_AMOUNT] >= 100000 THEN 1250 ELSE FLOOR(BASE2.[DTP_POINTS_BASIC] * 2.5) END)  AS 'BONUS_POINT_5X' 
67 
68FROM ( 
69 SELECT
70 [DTP_MEMBER_ID],
71 [DTP_AMOUNT],
72 [DTP_POINTS_BASIC],
73 ROW_NUMBER() OVER (PARTITION BY [DTP_MEMBER_ID] ORDER BY [DTP_TRX_DATE] ASC) AS 'TRX_COUNT'
74 FROM [DWH_PP].[dbo].[DAILY_TRX_PLMS_2018]
75 WHERE [DTP_COMPANY] = '001' AND
76 MONTH([DTP_TRX_DATE]) = '6' AND 
77 YEAR([DTP_TRX_DATE]) = '2018' AND
78 [DTP_TRANSACTION_CODE] = 'POI' AND [DTP_MEMBER_ID] IN 
79 (SELECT [DTP_MEMBER_ID] FROM BASE)) AS BASE2
80WHERE TRX_COUNT = 2 
81