【Couch Investor】这是近年来最便宜的时候：META、谷歌新技术与亚马逊深度估值

1
00:00:00,040 --> 00:00:02,200
大家好，欢迎回到今天的视频

2
00:00:02,200 --> 00:00:02,819
在今天的视频中

3
00:00:02,819 --> 00:00:04,540
我们有很多内容要讨论

4
00:00:04,540 --> 00:00:08,339
我们必须谈谈Meta以及那里发生的一切

5
00:00:08,339 --> 00:00:12,800
我们将讨论谷歌的Turbo Quant及其影响

6
00:00:12,800 --> 00:00:15,990
或者对像K Hidings这样的公司可能造成的影响

7
00:00:15,990 --> 00:00:17,789
包括美光、三星和

8
00:00:17,789 --> 00:00:21,690
当然还有英伟达、AMD、OpenAI及其他公司

9
00:00:21,690 --> 00:00:23,449
然后我还想谈谈

10
00:00:23,449 --> 00:00:25,329
当然还有亚马逊

11
00:00:25,329 --> 00:00:27,079
我将展示DCF分析

12
00:00:27,079 --> 00:00:27,859
基础部分

13
00:00:27,859 --> 00:00:31,440
亚马逊的牛熊案例分析

14
00:00:31,440 --> 00:00:32,299
目前来看

15
00:00:32,299 --> 00:00:32,960
市场

16
00:00:32,960 --> 00:00:34,518
当然已经休市

17
00:00:34,518 --> 00:00:37,999
周四市场表现不佳

18
00:00:37,999 --> 00:00:41,079
正如大家所见

19
00:00:41,079 --> 00:00:43,159
满屏都是红色

20
00:00:43,159 --> 00:00:44,960
英伟达下跌4%

21
00:00:44,960 --> 00:00:46,460
微软下跌7.

22
00:00:46,460 --> 00:00:47,200
5%

23
00:00:47,200 --> 00:00:50,490
美光再跌7%

24
00:00:50,490 --> 00:00:54,030
Meta一度下跌超8%

25
00:00:54,030 --> 00:00:57,329
真是惨淡的周四

26
00:00:57,329 --> 00:00:59,670
如果我们查看跟踪的公司列表

27
00:00:59,670 --> 00:01:02,549
仅有少数几只股票上涨

28
00:01:02,549 --> 00:01:03,250
Aggaia

29
00:01:03,250 --> 00:01:04,390
Salesforce、Rout

30
00:01:04,390 --> 00:01:05,030
Adobe

31
00:01:05,030 --> 00:01:05,790
Rubric

32
00:01:05,790 --> 00:01:07,230
PayPal、ServiceNow

33
00:01:07,230 --> 00:01:09,429
诺瓦·诺里斯和Hymns

34
00:01:09,429 --> 00:01:11,680
我们跟踪的其余公司

35
00:01:11,680 --> 00:01:13,760
其余全部下跌

36
00:01:13,760 --> 00:01:15,719
英伟达跌9.6%

37
00:01:15,719 --> 00:01:17,859
Rocket Lab跌9.5%

38
00:01:17,859 --> 00:01:19,799
还有许多其他上涨的股票

39
00:01:19,799 --> 00:01:20,280
七

40
00:01:20,280 --> 00:01:23,498
八至九个百分点

41
00:01:23,498 --> 00:01:24,159
正如我之前所说

42
00:01:24,159 --> 00:01:26,078
我还想重点分析Meta

43
00:01:26,078 --> 00:01:26,659
具体来说

44
00:01:26,659 --> 00:01:28,819
正如今天所见

45
00:01:28,819 --> 00:01:33,030
过去12个月股价下跌15.8%

46
00:01:33,030 --> 00:01:36,039
目前下跌10.4%

47
00:01:36,039 --> 00:01:41,959
该公司当前市盈率TTM为234倍，前瞻市盈率为

48
00:01:41,959 --> 00:01:45,060
17.6倍

49
00:01:45,060 --> 00:01:52,019
当然，股价过去12个月下跌10.4%有诸多原因

50
00:01:52,019 --> 00:01:57,409
或正经历从历史高点30%的回撤

51
00:01:57,409 --> 00:01:57,769
正确

52
00:01:57,769 --> 00:02:00,079
资本支出大幅增加

53
00:02:00,079 --> 00:02:01,540
对油井的担忧

54
00:02:01,540 --> 00:02:06,299
我们可能又要回到2022年看到的那种情况

55
00:02:06,299 --> 00:02:08,150
元宇宙大幅下跌

56
00:02:08,150 --> 00:02:10,469
从顶峰下跌76%

57
00:02:10,469 --> 00:02:12,719
一路跌至底部

58
00:02:12,719 --> 00:02:13,939
我们已经讨论过这一点

59
00:02:13,939 --> 00:02:15,280
我进行了完整的DCF分析

60
00:02:15,280 --> 00:02:16,379
也在这里展示给你看

61
00:02:16,379 --> 00:02:19,180
牛市基础和熊市情景

62
00:02:19,180 --> 00:02:21,819
现在你拥有了所有这些

63
00:02:21,819 --> 00:02:23,419
巨大的周期性投资

64
00:02:23,419 --> 00:02:24,719
事实上是的

65
00:02:24,719 --> 00:02:28,219
当你分析业务的利润率时

66
00:02:28,219 --> 00:02:34,620
顺便说一句，这是一家年营收超2000亿美元、增长率超20%的公司

67
00:02:34,620 --> 00:02:38,110
但当你看业务的利润率时

68
00:02:38,110 --> 00:02:39,509
就能看到这种模式

69
00:02:39,509 --> 00:02:44,310
当然你会开始理解市场担忧的原因

70
00:02:44,310 --> 00:02:46,590
优质业务超级盈利

71
00:02:46,590 --> 00:02:49,889
尤其是核心业务极其盈利

72
00:02:49,889 --> 00:02:52,449
但如果我们要回归

73
00:02:52,449 --> 00:02:59,259
或接近2022年情景，意味着资本支出大幅增加

74
00:02:59,259 --> 00:03:00,439
现金流开始下降

75
00:03:00,439 --> 00:03:01,939
利润率开始下滑

76
00:03:01,939 --> 00:03:09,009
如果几年后这些投资没有显著回报

77
00:03:09,009 --> 00:03:09,650
是的

78
00:03:09,650 --> 00:03:13,250
市场提出所有这些问题显得非常担忧

79
00:03:13,250 --> 00:03:18,759
现在如果在此基础上增加另一个因素

80
00:03:18,759 --> 00:03:22,039
可能损害核心业务

81
00:03:22,319 --> 00:03:28,000
就能完全理解股价今日下跌8%的原因

82
00:03:28,000 --> 00:03:29,280
为何目前处于低位

83
00:03:29,280 --> 00:03:31,199
为何过去12个月持续下跌

84
00:03:31,199 --> 00:03:33,240
如果你还不清楚我在说什么

85
00:03:33,240 --> 00:03:39,620
我在说Meta和YouTube在标志性社交媒体成瘾案中被判定有过失

86
00:03:39,620 --> 00:03:41,419
陪审团认定公司造成损害

87
00:03:41,419 --> 00:03:47,430
因设计成瘾性功能导致年轻用户心理健康受损

88
00:03:47,430 --> 00:03:50,719
我不想详细解释这里发生的事情

89
00:03:50,719 --> 00:03:53,280
以及未来可能引发的后果

90
00:03:53,280 --> 00:03:54,780
如果你喜欢这类视频

91
00:03:54,780 --> 00:03:55,620
他按下按钮

92
00:03:55,620 --> 00:03:56,259
你知道该怎么做

93
00:03:56,259 --> 00:03:57,419
我们非常感谢

94
00:03:57,419 --> 00:03:58,659
如果能进一步支持我

95
00:03:58,659 --> 00:04:00,180
我在描述中是否该删减

96
00:04:00,180 --> 00:04:01,659
并在评论区置顶前十

97
00:04:01,659 --> 00:04:02,870
最佳立即购买推荐

98
00:04:02,870 --> 00:04:04,250
或访问官网

99
00:04:04,250 --> 00:04:05,310
slash investor

100
00:04:05,310 --> 00:04:06,889
非常感谢

101
00:04:06,889 --> 00:04:08,530
因此判决结果是

102
00:04:08,530 --> 00:04:14,030
陪审团认定Meta和YouTube在设计平台时存在过失

103
00:04:14,030 --> 00:04:16,410
这些平台如今正在成瘾儿童

104
00:04:16,410 --> 00:04:18,730
原告现年20岁

105
00:04:18,730 --> 00:04:24,250
六岁时开始使用YouTube，九岁时开始使用Instagram

106
00:04:24,250 --> 00:04:25,589
被提及的功能

107
00:04:25,589 --> 00:04:27,529
无限滚动算法

108
00:04:27,529 --> 00:04:28,310
推荐机制

109
00:04:28,310 --> 00:04:33,670
自动播放视频故意诱导年轻用户沉迷

110
00:04:33,670 --> 00:04:35,079
本案的损害赔偿

111
00:04:35,079 --> 00:04:36,779
总计六百万美元

112
00:04:36,779 --> 00:04:37,480
其中四点

113
00:04:37,480 --> 00:04:42,120
Meta赔偿两百万美元，YouTube赔偿一百八十万美元

114
00:04:42,120 --> 00:04:43,319
现在你可能会说

115
00:04:43,319 --> 00:04:44,399
你在开玩笑吗

116
00:04:44,399 --> 00:04:48,579
Meta难道因为这个损失了接近百亿市值

117
00:04:48,579 --> 00:04:50,310
正是因为这个

118
00:04:50,310 --> 00:04:51,449
因为四点

119
00:04:51,449 --> 00:04:53,490
两百万美元罚款

120
00:04:53,490 --> 00:04:57,000
谁在乎呢对吧

121
00:04:57,360 --> 00:05:00,740
并非完全如此因为它确实重要

122
00:05:00,740 --> 00:05:05,269
这是两千余起待审案件中的首例

123
00:05:05,269 --> 00:05:07,569
因此你可以开始理解

124
00:05:07,569 --> 00:05:07,970
好的

125
00:05:07,970 --> 00:05:09,189
如果他们看到

126
00:05:09,189 --> 00:05:09,670
哦

127
00:05:09,670 --> 00:05:12,999
有人获得了这笔赔偿金

128
00:05:12,999 --> 00:05:16,269
这是该案的判决结果

129
00:05:16,269 --> 00:05:21,668
也许我们现在可以尝试碰运气并立即获得赔偿

130
00:05:21,668 --> 00:05:25,028
新墨西哥州陪审团前一天裁定Meta

131
00:05:25,028 --> 00:05:30,959
因未保护儿童免受 predators侵害被罚款三亿七千五百万美元

132
00:05:30,959 --> 00:05:31,500
好的

133
00:05:31,500 --> 00:05:33,240
这已是更大金额

134
00:05:33,240 --> 00:05:37,678
我认为他们应该支付更多

135
00:05:37,678 --> 00:05:42,540
美国还有第230条可能面临风险

136
00:05:42,540 --> 00:05:45,420
新墨西哥州总检察长表示存在现实可能

137
00:05:45,420 --> 00:05:49,910
这些案件可能促使国会重新审视第230条

138
00:05:49,910 --> 00:05:52,529
简言之第230条

139
00:05:52,529 --> 00:05:56,910
是对互联网公司的保护条款

140
00:05:56,910 --> 00:05:58,810
包括社交媒体公司

141
00:05:58,810 --> 00:06:02,540
他们无需对

142
00:06:02,540 --> 00:06:03,139
嗯

143
00:06:03,139 --> 00:06:06,300
平台上发布内容的影响负责

144
00:06:06,300 --> 00:06:08,199
简而言之

145
00:06:08,199 --> 00:06:12,379
这就是为什么Meta今日大幅下跌

146
00:06:12,379 --> 00:06:14,459
谷歌今日也小幅下跌

147
00:06:16,519 --> 00:06:17,660
原因在于

148
00:06:17,660 --> 00:06:22,300
这只是数千潜在案件中的首个

149
00:06:22,300 --> 00:06:26,649
未来判决结果难以预料

150
00:06:26,649 --> 00:06:29,889
或许这能促使政策发生变化

151
00:06:29,889 --> 00:06:35,910
或许这会迫使Meta改变人们在应用内的互动方式

152
00:06:35,910 --> 00:06:36,350
现在

153
00:06:36,350 --> 00:06:38,870
作为Meta的股东

154
00:06:38,870 --> 00:06:40,709
也是谷歌的股东

155
00:06:40,709 --> 00:06:41,910
直接来说

156
00:06:41,910 --> 00:06:43,019
包括YouTube在内

157
00:06:43,019 --> 00:06:46,019
我确实想谈谈这带来的影响

158
00:06:46,019 --> 00:06:47,939
因为作为股东

159
00:06:47,939 --> 00:06:48,399
好的

160
00:06:48,399 --> 00:06:51,160
如果商业模式必须改变的话

161
00:06:51,160 --> 00:06:53,040
当然可能变得更糟

162
00:06:53,040 --> 00:06:57,329
这意味着人们可能减少在YouTube上的使用时间

163
00:06:57,329 --> 00:06:58,550
在Instagram上也是如此

164
00:06:58,550 --> 00:07:02,310
这意味着用户接收到的广告会减少

165
00:07:02,310 --> 00:07:06,930
这意味着公司收入减少，对我不利

166
00:07:06,930 --> 00:07:08,089
作为股东

167
00:07:08,089 --> 00:07:09,350
另一方面

168
00:07:09,350 --> 00:07:14,560
我对为保护孩子所做的改变毫无异议

169
00:07:14,560 --> 00:07:15,259
好的

170
00:07:15,259 --> 00:07:17,300
首先我认为青少年

171
00:07:17,300 --> 00:07:20,439
未成年人不应使用社交媒体

172
00:07:20,439 --> 00:07:24,939
尤其是六岁或九岁这样的年龄

173
00:07:24,939 --> 00:07:28,829
我不认为他们需要在社交媒体上活动

174
00:07:28,829 --> 00:07:31,350
其次对于家长而言

175
00:07:31,350 --> 00:07:33,189
这里不会展开长篇大论

176
00:07:33,189 --> 00:07:33,829
等等

177
00:07:33,829 --> 00:07:34,569
因为我不是家长

178
00:07:34,569 --> 00:07:35,709
所以我不打算谈论

179
00:07:35,709 --> 00:07:37,199
就像自己是家长一样

180
00:07:37,199 --> 00:07:39,310
但家长确实扮演重要角色

181
00:07:39,310 --> 00:07:42,850
家长在孩子的教育中起重要作用

182
00:07:42,850 --> 00:07:45,470
以及决定孩子可以或不可以做什么

183
00:07:45,470 --> 00:07:48,310
当然无法完全控制孩子

184
00:07:48,310 --> 00:07:52,778
但可以管控十八岁以下的Instagram

185
00:07:52,778 --> 00:07:53,579
没错

186
00:07:53,579 --> 00:07:56,699
十六岁以下的Instagram账号

187
00:07:56,699 --> 00:07:57,139
是的

188
00:07:57,139 --> 00:08:00,500
可以对这些Instagram账号进行较多管控

189
00:08:00,500 --> 00:08:08,199
是否需要监护人批准才能创建青少年账号

190
00:08:08,199 --> 00:08:11,759
或使用时需家长同意

191
00:08:11,759 --> 00:08:16,060
由监护人或拥有法律权限的人

192
00:08:16,060 --> 00:08:21,788
可以设置这些应用的使用时长限制

193
00:08:21,788 --> 00:08:24,939
毕竟系统设置本就存在

194
00:08:24,939 --> 00:08:27,500
这些设置旨在保护用户

195
00:08:27,500 --> 00:08:32,460
或帮助用户减少在应用上花费的时间

196
00:08:32,460 --> 00:08:35,179
这些功能早已存在

197
00:08:35,179 --> 00:08:36,480
不仅限于Instagram

198
00:08:36,480 --> 00:08:38,538
其他应用同样具备

199
00:08:38,538 --> 00:08:42,178
只需具备一定自制力

200
00:08:42,178 --> 00:08:45,038
然后我们就可以进入讨论环节了

201
00:08:45,038 --> 00:08:46,109
好的

202
00:08:46,109 --> 00:08:53,828
我们什么时候要讨论人们可以自由做任何事的事实

203
00:08:53,828 --> 00:08:56,729
如果我想每天使用这个应用五个小时

204
00:08:56,729 --> 00:08:59,159
我应该被允许这样做

205
00:08:59,559 --> 00:09:03,120
这是很多人现在也在问的问题

206
00:09:03,120 --> 00:09:04,659
但我确实同意

207
00:09:04,659 --> 00:09:06,379
尤其是当我们谈到这个案例

208
00:09:06,379 --> 00:09:08,029
这里的青少年

209
00:09:08,029 --> 00:09:09,090
未成年人

210
00:09:09,090 --> 00:09:09,710
是的

211
00:09:09,710 --> 00:09:11,850
现在必须解决这个问题

212
00:09:11,850 --> 00:09:14,360
如果会出现大规模的群体

213
00:09:14,360 --> 00:09:17,899
针对十六岁以下儿童的社交媒体使用

214
00:09:17,899 --> 00:09:18,399
是的

215
00:09:18,399 --> 00:09:19,299
这将产生影响

216
00:09:19,299 --> 00:09:20,940
比如谷歌这样的公司

217
00:09:20,940 --> 00:09:23,009
YouTube和Instagram

218
00:09:23,009 --> 00:09:23,669
TikTok

219
00:09:23,669 --> 00:09:24,710
Snapchat等

220
00:09:24,710 --> 00:09:25,070
等等

221
00:09:25,070 --> 00:09:25,769
因为确实

222
00:09:25,769 --> 00:09:31,730
数十亿用户中有一部分是未成年人

223
00:09:31,730 --> 00:09:33,629
未满十六岁

224
00:09:33,629 --> 00:09:36,730
所以可以理解如果这种情况发生

225
00:09:36,730 --> 00:09:42,409
这对Meta的核心盈利业务有影响

226
00:09:42,409 --> 00:09:47,470
如果Meta的核心盈利业务稍有变化

227
00:09:47,470 --> 00:09:50,750
在此可能面临小幅阻力

228
00:09:50,750 --> 00:09:51,429
嗯

229
00:09:51,429 --> 00:09:54,269
他们已经在资本支出上投入了大量资金

230
00:09:54,269 --> 00:09:56,889
我们还不清楚投资回报率如何

231
00:09:56,889 --> 00:10:00,149
但我们知道核心业务极其盈利

232
00:10:00,149 --> 00:10:03,330
但若这一核心业务

233
00:10:03,330 --> 00:10:06,490
会遭遇此类阻力

234
00:10:06,490 --> 00:10:09,580
市场会相应调整估值

235
00:10:09,580 --> 00:10:11,480
我认为这就是当前的情况

236
00:10:11,480 --> 00:10:16,259
虽然我觉得它从一边直接转向另一边过于激进

237
00:10:16,259 --> 00:10:23,570
这在Meta、Zux和谷歌同样常见

238
00:10:23,570 --> 00:10:27,220
我认为他们会度过这个难关

239
00:10:27,220 --> 00:10:30,019
这确实让Meta面临转变

240
00:10:30,019 --> 00:10:31,860
谷歌2025年的模样

241
00:10:31,860 --> 00:10:34,839
再加上这些风险因素

242
00:10:34,839 --> 00:10:36,158
和去年一样

243
00:10:36,158 --> 00:10:36,619
你知道的

244
00:10:36,619 --> 00:10:38,779
谷歌将被迫出售

245
00:10:38,779 --> 00:10:39,539
安卓系统

246
00:10:39,539 --> 00:10:43,139
谷歌将被迫出售或剥离Chrome

247
00:10:43,259 --> 00:10:43,740
是的

248
00:10:43,740 --> 00:10:44,980
这曾是潜在风险

249
00:10:44,980 --> 00:10:48,740
但实际我认为这会发生吗

250
00:10:48,740 --> 00:10:49,980
正如我去年所说

251
00:10:49,980 --> 00:10:51,899
这里完全不同

252
00:10:51,899 --> 00:10:56,379
我们现在又面临Meta股票的利空压力

253
00:10:56,379 --> 00:10:58,320
这会持续下去吗

254
00:10:58,320 --> 00:11:01,779
在我看来这对未来发展非常不利

255
00:11:01,779 --> 00:11:04,700
不会有太大变化

256
00:11:04,700 --> 00:11:05,320
正如我之前所说

257
00:11:05,320 --> 00:11:07,559
YouTube和Facebook

258
00:11:07,559 --> 00:11:11,230
Meta将对此判决提出上诉

259
00:11:11,230 --> 00:11:11,950
所以目前

260
00:11:11,950 --> 00:11:12,990
如果我们观察Meta股票

261
00:11:12,990 --> 00:11:15,649
首先从日线来看明显超卖

262
00:11:15,649 --> 00:11:16,389
其次

263
00:11:16,389 --> 00:11:19,059
如果我们追溯到大约

264
00:11:19,059 --> 00:11:20,299
2025年五月

265
00:11:20,299 --> 00:11:21,840
我们看到明显的跳空上涨

266
00:11:21,840 --> 00:11:24,659
这个缺口终于被回补

267
00:11:24,659 --> 00:11:25,779
是的

268
00:11:25,779 --> 00:11:26,659
正如你所知

269
00:11:26,659 --> 00:11:28,320
并非所有缺口都会回补

270
00:11:28,320 --> 00:11:29,639
但大多数都会回补

271
00:11:29,639 --> 00:11:31,639
这又是典型案例

272
00:11:31,639 --> 00:11:34,340
请告诉我你对这种情况的看法

273
00:11:34,340 --> 00:11:37,620
你觉得这是明显的深度调整信号吗

274
00:11:37,620 --> 00:11:38,340
或者不是

275
00:11:38,340 --> 00:11:40,460
请在下方评论区留言

276
00:11:40,460 --> 00:11:47,330
接下来分析谷歌宣布的内容及其对Micron等的影响

277
00:11:47,330 --> 00:11:47,690
等等

278
00:11:47,690 --> 00:11:48,330
等等

279
00:11:48,330 --> 00:11:48,929
总之

280
00:11:48,929 --> 00:11:49,409
这就是关键点

281
00:11:49,409 --> 00:11:51,250
这来自谷歌研究

282
00:11:51,250 --> 00:11:53,519
这才是重点所在

283
00:11:53,519 --> 00:11:58,399
这是通过极端压缩重新定义AI效率的技术

284
00:11:58,399 --> 00:12:01,370
这项技术于3月24日发布

285
00:12:01,370 --> 00:12:04,529
这里提到我们推出一系列

286
00:12:04,529 --> 00:12:14,328
基于理论的量化算法，可大幅压缩大型语言模型和向量搜索引擎

287
00:12:14,328 --> 00:12:16,089
很好

288
00:12:16,089 --> 00:12:16,609
我明白

289
00:12:16,609 --> 00:12:19,089
有些人觉得这让我联想到

290
00:12:19,089 --> 00:12:22,129
或者硅谷的Pipe Piper项目

291
00:12:22,129 --> 00:12:23,210
没错

292
00:12:23,210 --> 00:12:24,190
但并非如此

293
00:12:24,190 --> 00:12:28,529
关于Turbo Quant的核心在于

294
00:12:28,529 --> 00:12:29,669
尚未正式发布

295
00:12:29,669 --> 00:12:30,330
目前还未推出

296
00:12:30,330 --> 00:12:33,009
其次Turbo Quant

297
00:12:33,009 --> 00:12:36,070
这个概念早已被提出

298
00:12:36,070 --> 00:12:38,029
猜测何时会实现

299
00:12:38,029 --> 00:12:39,590
四月二十八日

300
00:12:39,590 --> 00:12:43,389
二十三年发生了什么事去年

301
00:12:43,389 --> 00:12:44,730
当这个出现时

302
00:12:44,730 --> 00:12:45,330
没有

303
00:12:45,330 --> 00:12:46,690
只是现在的情况

304
00:12:46,690 --> 00:12:46,889
是的

305
00:12:46,889 --> 00:12:51,039
这是谷歌研究在社交媒体上发布的

306
00:12:51,039 --> 00:12:53,950
因此获得了更多关注

307
00:12:53,950 --> 00:12:58,669
这让我想起二十三年初的深海时刻

308
00:12:58,669 --> 00:13:01,679
突然所有芯片制造商

309
00:13:01,679 --> 00:13:02,740
每个像英伟达

310
00:13:02,740 --> 00:13:03,659
AMD和新创公司

311
00:13:03,659 --> 00:13:07,039
他们都下跌了二十个百分点左右

312
00:13:07,039 --> 00:13:11,360
但最终没有深海危机

313
00:13:11,360 --> 00:13:13,360
并非这些公司终结

314
00:13:13,360 --> 00:13:13,679
不

315
00:13:13,679 --> 00:13:15,210
这仅仅表明

316
00:13:15,210 --> 00:13:16,210
这些模型

317
00:13:16,210 --> 00:13:17,850
巨大的训练模型

318
00:13:17,850 --> 00:13:22,970
所有这些都能以更高效、更高效的方式完成

319
00:13:22,970 --> 00:13:24,750
更低成本的方式

320
00:13:24,750 --> 00:13:26,990
这并不意味着需要更少芯片

321
00:13:26,990 --> 00:13:27,470
不

322
00:13:27,470 --> 00:13:30,389
这意味着能完成更多任务

323
00:13:30,389 --> 00:13:33,389
这意味着能服务更多用户

324
00:13:33,389 --> 00:13:34,730
使用量随之增加

325
00:13:34,730 --> 00:13:35,110
你知道的

326
00:13:35,110 --> 00:13:36,090
金氏悖论

327
00:13:36,090 --> 00:13:37,230
等等

328
00:13:37,600 --> 00:13:39,220
这里同样适用

329
00:13:39,220 --> 00:13:42,919
如果我们问什么是涡轮量化

330
00:13:42,919 --> 00:13:44,259
这就是答案

331
00:13:44,259 --> 00:13:46,539
像Gemini或ChatGPT这样的大模型

332
00:13:46,539 --> 00:13:48,230
存在记忆墙问题

333
00:13:48,230 --> 00:13:51,029
随着与AI的对话时间延长

334
00:13:51,029 --> 00:13:57,460
模型必须存储当前所有内容的数字备忘录，称为键值缓存

335
00:13:57,460 --> 00:13:59,399
这个缓存不断增大

336
00:13:59,399 --> 00:13:59,899
我意思是说

337
00:13:59,899 --> 00:14:02,480
当你想清理手机内存时就能看到

338
00:14:02,480 --> 00:14:03,379
例如

339
00:14:03,379 --> 00:14:05,240
这个缓存不断增大

340
00:14:05,240 --> 00:14:10,639
最终填满AI芯片的所有可用内存VRAM

341
00:14:10,639 --> 00:14:16,519
涡轮量化是谷歌研究的新数学压缩技术

342
00:14:16,519 --> 00:14:22,740
据称可压缩内存六倍且无精度损失

343
00:14:22,740 --> 00:14:24,539
由于数据更小

344
00:14:24,539 --> 00:14:29,049
芯片处理速度可提升八倍

345
00:14:29,769 --> 00:14:31,509
听起来很棒

346
00:14:31,509 --> 00:14:34,309
为什么呢？几个要点

347
00:14:34,309 --> 00:14:37,610
它解决了推理瓶颈问题

348
00:14:37,610 --> 00:14:39,429
正如我们反复强调的

349
00:14:39,429 --> 00:14:44,139
所有人都说推理远大于训练

350
00:14:44,139 --> 00:14:44,919
但没错

351
00:14:44,919 --> 00:14:46,600
存在一个瓶颈

352
00:14:46,600 --> 00:14:49,100
因为我们已经和Jose讨论过这个问题

353
00:14:49,100 --> 00:14:50,899
很多时候当你想要生成

354
00:14:50,899 --> 00:14:54,419
比如说AI生成的视频之类的东西

355
00:14:54,419 --> 00:14:56,960
这段所谓的视频时长是多少

356
00:14:56,960 --> 00:14:58,669
它是一个八秒的片段

357
00:14:58,669 --> 00:14:59,690
十秒钟

358
00:14:59,690 --> 00:15:01,318
差不多就是这样

359
00:15:01,318 --> 00:15:04,178
最终你希望达到一分钟

360
00:15:04,178 --> 00:15:07,620
当然这也会帮助实现这一点

361
00:15:07,620 --> 00:15:09,620
这将有助于解决

362
00:15:09,620 --> 00:15:16,570
越来越多的实用产品和功能将被全球数十亿用户使用

363
00:15:16,570 --> 00:15:24,399
如果全球数十亿用户能够使用我们AI赋能的实用产品和功能

364
00:15:24,399 --> 00:15:28,460
这些公司的投资回报率可能会大幅提升

365
00:15:28,460 --> 00:15:32,970
因为如果你能比以前运行更多推理任务

366
00:15:32,970 --> 00:15:36,350
而成本却更低

367
00:15:36,350 --> 00:15:37,269
这对你们来说很好

368
00:15:37,269 --> 00:15:41,009
这对目前亏损的公司来说也是好消息

369
00:15:41,009 --> 00:15:44,188
这对今天极度盈利的公司同样有利

370
00:15:44,188 --> 00:15:46,529
因为它们可以更快发展

371
00:15:46,529 --> 00:15:48,149
这对我也会很好

372
00:15:48,149 --> 00:15:48,690
用户

373
00:15:48,690 --> 00:15:49,210
是的

374
00:15:49,210 --> 00:15:49,970
我现在可以拥有

375
00:15:49,970 --> 00:15:52,500
也许未来的对话体验

376
00:15:52,500 --> 00:15:53,559
人形机器人

377
00:15:53,559 --> 00:15:54,620
也许智能眼镜

378
00:15:54,620 --> 00:15:55,240
你随便举个例子

379
00:15:55,240 --> 00:15:57,460
不需要十秒钟

380
00:15:57,460 --> 00:15:58,120
二十秒

381
00:15:58,120 --> 00:16:01,759
三十秒或一分钟就能得到答案

382
00:16:01,759 --> 00:16:04,100
或者当你处理大型项目时

383
00:16:04,100 --> 00:16:05,759
我不确定什么项目

384
00:16:05,759 --> 00:16:08,220
消耗五万多个token

385
00:16:08,220 --> 00:16:08,860
左右

386
00:16:08,860 --> 00:16:12,580
在这种情况下会大幅减少

387
00:16:12,580 --> 00:16:16,990
我们不会减少使用这些服务

388
00:16:16,990 --> 00:16:20,970
公司所需的内存也会减少

389
00:16:20,970 --> 00:16:23,279
更少的HBM内存

390
00:16:23,279 --> 00:16:24,139
并非如此

391
00:16:24,139 --> 00:16:29,019
我认为未来它们需要的内存会大幅增加

392
00:16:29,019 --> 00:16:29,639
现在

393
00:16:29,639 --> 00:16:33,059
具体讨论像美光、铠侠等公司

394
00:16:33,059 --> 00:16:33,600
等等

395
00:16:33,600 --> 00:16:34,159
是的

396
00:16:34,159 --> 00:16:36,500
从利润率来看

397
00:16:36,500 --> 00:16:40,068
可能在2027年前后达到峰值

398
00:16:40,068 --> 00:16:41,188
但从营收来看

399
00:16:41,188 --> 00:16:41,948
我不这么认为

400
00:16:41,948 --> 00:16:42,629
我觉得再想想

401
00:16:42,629 --> 00:16:45,928
未来需要更多内存

402
00:16:45,928 --> 00:16:51,110
因为有了这个你才能真正将AI带到边缘端

403
00:16:51,110 --> 00:16:53,639
真正将AI带到你的电脑

404
00:16:53,639 --> 00:16:54,700
带到你的手机

405
00:16:54,700 --> 00:16:56,080
带到你的智能眼镜

406
00:16:56,080 --> 00:16:56,840
猜猜看

407
00:16:56,840 --> 00:17:01,879
所有这些硬件产品都需要内存

408
00:17:01,879 --> 00:17:05,940
这里特别提到HBM

409
00:17:05,940 --> 00:17:07,680
美光K隐藏式

410
00:17:07,680 --> 00:17:08,960
三星当然也包括在内

411
00:17:08,960 --> 00:17:09,559
没错

412
00:17:09,559 --> 00:17:12,190
这些公司在其历史中

413
00:17:12,190 --> 00:17:14,289
至少以美光为例

414
00:17:14,289 --> 00:17:15,650
让我们拉到极致

415
00:17:15,650 --> 00:17:16,069
是的

416
00:17:16,069 --> 00:17:18,759
或许回顾十年历史

417
00:17:18,759 --> 00:17:20,279
从利润率来看

418
00:17:20,279 --> 00:17:20,720
是的

419
00:17:20,720 --> 00:17:23,240
显然这是一个周期性行业

420
00:17:23,240 --> 00:17:23,759
我们上涨

421
00:17:23,759 --> 00:17:24,240
我们达到峰值

422
00:17:24,240 --> 00:17:24,759
我们下滑

423
00:17:24,759 --> 00:17:25,240
我们再次达到峰值

424
00:17:25,240 --> 00:17:26,279
我们进一步下滑

425
00:17:26,279 --> 00:17:28,298
现在我们正在回升

426
00:17:28,298 --> 00:17:31,039
这是过去十二个月的情况

427
00:17:31,039 --> 00:17:32,039
所以没错

428
00:17:32,039 --> 00:17:32,859
人们说

429
00:17:32,859 --> 00:17:36,789
最终HBM也会成为大宗商品

430
00:17:36,789 --> 00:17:39,710
但这时定制HBM就发挥作用了

431
00:17:39,710 --> 00:17:40,769
尤其是对美光而言

432
00:17:40,769 --> 00:17:42,890
针对HBM的E版本

433
00:17:42,890 --> 00:17:47,059
这时商品属性被搁置

434
00:17:47,059 --> 00:17:49,839
他们成为更具战略性的角色

435
00:17:49,839 --> 00:17:51,240
更战略性的合作伙伴

436
00:17:51,240 --> 00:17:53,138
像英伟达

437
00:17:53,138 --> 00:17:56,519
谷歌等公司

438
00:17:56,519 --> 00:18:00,778
这些公司可能有特定需求

439
00:18:00,778 --> 00:18:02,569
特定的工作负载

440
00:18:02,569 --> 00:18:11,328
他们的工作方式需要特定的软件算法

441
00:18:11,328 --> 00:18:12,909
简而言之

442
00:18:12,909 --> 00:18:14,308
非常技术性

443
00:18:14,308 --> 00:18:17,019
我花了一些时间理解这些

444
00:18:17,019 --> 00:18:18,380
也和何塞讨论过

445
00:18:18,380 --> 00:18:20,799
这就是定制HBM的作用

446
00:18:20,799 --> 00:18:22,960
定制高带宽内存

447
00:18:22,960 --> 00:18:24,559
通过定制高带宽内存

448
00:18:24,559 --> 00:18:26,140
你就不再是普通商品

449
00:18:26,140 --> 00:18:30,099
因为只有少数玩家能实现

450
00:18:30,099 --> 00:18:30,980
就像现在这样

451
00:18:30,980 --> 00:18:31,779
Hbm

452
00:18:31,779 --> 00:18:33,740
只有少数玩家

453
00:18:33,740 --> 00:18:35,920
美光在隐藏三星

454
00:18:35,920 --> 00:18:36,180
是的

455
00:18:36,180 --> 00:18:38,588
你确实有部分中国玩家

456
00:18:38,588 --> 00:18:41,180
但他们占据市场多少份额

457
00:18:41,180 --> 00:18:43,309
他们几乎没什么

458
00:18:43,309 --> 00:18:44,009
最终

459
00:18:44,009 --> 00:18:44,470
是的

460
00:18:44,470 --> 00:18:45,869
也许但再次

461
00:18:45,869 --> 00:18:49,750
这就是这些公司聚焦未来的缘故

462
00:18:49,750 --> 00:18:52,619
专注于定制HBM

463
00:18:52,619 --> 00:18:56,140
以免沦为市场所担忧的

464
00:18:56,140 --> 00:19:00,769
长期成为并保持大宗商品

465
00:19:00,769 --> 00:19:03,950
目前从利润率来看

466
00:19:03,950 --> 00:19:06,329
我们更接近峰值而非

467
00:19:06,329 --> 00:19:07,450
接近底部

468
00:19:08,450 --> 00:19:10,089
但从增长来看

469
00:19:10,089 --> 00:19:16,109
我认为像英伟达这样的公司仍有很长路要走

470
00:19:16,109 --> 00:19:18,409
AMD等等

471
00:19:18,409 --> 00:19:21,128
这一切也利好MD

472
00:19:21,128 --> 00:19:21,689
例如

473
00:19:21,689 --> 00:19:25,400
他们投入大量内存以保持竞争力

474
00:19:25,400 --> 00:19:28,700
但如果现在能通过减少内存竞争

475
00:19:28,700 --> 00:19:30,299
成本将更低

476
00:19:30,299 --> 00:19:31,740
这只会是

477
00:19:31,740 --> 00:19:35,049
可能增加软件压力

478
00:19:35,049 --> 00:19:39,490
这方面我们清楚是英伟达的巨大优势

479
00:19:39,490 --> 00:19:42,990
这也是MD正在努力的方向

480
00:19:42,990 --> 00:19:44,829
而美光目前

481
00:19:44,829 --> 00:19:46,210
这是否过度反应

482
00:19:46,210 --> 00:19:47,069
我认为是的

483
00:19:47,069 --> 00:19:49,900
基于当前增长速度和预期

484
00:19:49,900 --> 00:19:50,480
是的

485
00:19:50,480 --> 00:19:53,240
这里的回调相当剧烈

486
00:19:53,240 --> 00:19:55,179
根据数据来看

487
00:19:55,179 --> 00:19:57,699
我们接近100日移动平均线

488
00:19:57,699 --> 00:20:00,338
位于约331美元

489
00:20:00,338 --> 00:20:01,538
RSI偏低

490
00:20:01,538 --> 00:20:02,798
但未超卖

491
00:20:02,798 --> 00:20:03,219
嗯

492
00:20:03,219 --> 00:20:03,909
暂时还没

493
00:20:03,909 --> 00:20:04,269
哦

494
00:20:04,269 --> 00:20:04,648
顺便说

495
00:20:04,648 --> 00:20:06,148
当你看市盈率P

496
00:20:06,148 --> 00:20:06,669
500倍

497
00:20:06,669 --> 00:20:08,229
周线图上

498
00:20:08,229 --> 00:20:09,028
发生了什么

499
00:20:09,028 --> 00:20:09,648
嗯

500
00:20:09,648 --> 00:20:10,548
它确实有影响

501
00:20:10,548 --> 00:20:14,359
现在它正位于五十日移动平均线上

502
00:20:14,359 --> 00:20:15,259
最后一点

503
00:20:15,259 --> 00:20:17,420
让我们看看亚马逊

504
00:20:17,420 --> 00:20:18,819
我看好亚马逊

505
00:20:18,819 --> 00:20:23,339
它始终是那种无需多想的买入持有型股票

506
00:20:23,339 --> 00:20:24,019
因为没错

507
00:20:24,019 --> 00:20:27,569
它通过广告订阅业务实现增长

508
00:20:27,569 --> 00:20:28,829
人工智能、AWS

509
00:20:28,829 --> 00:20:30,970
机器人技术之类领域

510
00:20:33,750 --> 00:20:36,210
它同样能提供这种保障

511
00:20:36,210 --> 00:20:38,009
或者说确定性

512
00:20:38,009 --> 00:20:41,049
我不确定这里是否该用确定性

513
00:20:41,049 --> 00:20:46,569
但即使在衰退期人们仍会使用亚马逊购物

514
00:20:46,569 --> 00:20:48,269
这是事实

515
00:20:48,269 --> 00:20:51,380
在我看来这家公司表现会很好

516
00:20:51,380 --> 00:20:52,339
无论怎样

517
00:20:52,339 --> 00:20:54,440
当然表现会更佳

518
00:20:56,138 --> 00:20:59,839
但在衰退期仍希望持有那些

519
00:20:59,839 --> 00:21:04,490
部分业务仍能稳健运营的公司

520
00:21:04,490 --> 00:21:06,730
让我们再看亚马逊

521
00:21:06,730 --> 00:21:08,549
当前股价是多少

522
00:21:08,549 --> 00:21:11,888
我们支付了什么价格

523
00:21:11,888 --> 00:21:14,588
这基本是当前总结

524
00:21:14,588 --> 00:21:16,348
牛市、基准和熊市情景

525
00:21:16,348 --> 00:21:19,838
我给牛市35%概率

526
00:21:19,838 --> 00:21:23,479
基准50%、熊市15%

527
00:21:23,479 --> 00:21:25,298
合计36%

528
00:21:25,298 --> 00:21:28,430
当前价格有6.6%上行空间

529
00:21:28,430 --> 00:21:30,480
概率加权价格

530
00:21:30,480 --> 00:21:32,119
我是如何得出这个结论的

531
00:21:32,119 --> 00:21:33,339
我就是这样计算的

532
00:21:33,339 --> 00:21:35,759
牛市情景如下

533
00:21:35,759 --> 00:21:36,359
是的

534
00:21:36,359 --> 00:21:38,359
这可能有些激进

535
00:21:38,359 --> 00:21:43,119
但我认为AWS增长将持续加速

536
00:21:43,119 --> 00:21:47,000
广告业务仍将稳健增长

537
00:21:47,000 --> 00:21:48,480
这就是亚马逊的特点

538
00:21:48,480 --> 00:21:52,640
快速发展的业务也是高利润业务

539
00:21:52,640 --> 00:21:55,720
这支持了我的观点

540
00:21:55,720 --> 00:21:57,900
这里存在看涨理由

541
00:21:57,900 --> 00:21:59,420
可能略显乐观

542
00:21:59,420 --> 00:21:59,680
好吧

543
00:21:59,680 --> 00:22:00,420
没关系

544
00:22:00,420 --> 00:22:01,819
你可以调整这些数字

545
00:22:01,819 --> 00:22:02,900
所以是的

546
00:22:02,900 --> 00:22:07,200
营收将保持低双位数增长

547
00:22:07,200 --> 00:22:09,789
直到2032年

548
00:22:09,789 --> 00:22:12,069
然后我们下探到九七点

549
00:22:12,069 --> 00:22:15,930
五到六点五个百分点，有一定缓冲空间

550
00:22:15,930 --> 00:22:16,309
是的

551
00:22:16,309 --> 00:22:20,269
我认为这就是我们即将看到亚马逊盈利能力的时刻

552
00:22:20,269 --> 00:22:20,970
真正核心

553
00:22:20,970 --> 00:22:23,339
盈利能力开始显现

554
00:22:23,339 --> 00:22:25,319
预计将达到两到三点二

555
00:22:25,319 --> 00:22:27,409
三点五个百分点

556
00:22:27,409 --> 00:22:30,969
十年后自由现金流利润率

557
00:22:30,969 --> 00:22:33,900
再次达到十七点五百分比

558
00:22:33,900 --> 00:22:34,779
这就是牛市情景

559
00:22:34,779 --> 00:22:35,380
这是其中之一

560
00:22:35,380 --> 00:22:35,759
是的

561
00:22:35,759 --> 00:22:37,420
一切顺利

562
00:22:37,420 --> 00:22:38,619
AWS业务顺利

563
00:22:38,619 --> 00:22:40,319
机器人效率

564
00:22:40,319 --> 00:22:41,900
广告订阅

565
00:22:41,900 --> 00:22:42,759
你随便说

566
00:22:42,759 --> 00:22:44,119
所有这些都顺利进行

567
00:22:44,119 --> 00:22:47,160
亚马逊成为非常盈利的业务

568
00:22:47,160 --> 00:22:48,059
由此可见

569
00:22:48,059 --> 00:22:51,339
这就是你现在得到的基本情况

570
00:22:51,339 --> 00:22:52,400
基于基准情景

571
00:22:52,400 --> 00:22:54,079
其实就是牛市情景

572
00:22:54,079 --> 00:22:58,269
但不包含额外假设

573
00:22:58,269 --> 00:22:59,549
所以在这里同样

574
00:22:59,549 --> 00:23:01,890
我仍预期稳健增长

575
00:23:01,890 --> 00:23:06,839
但到2031年已进入高个位数增长

576
00:23:06,839 --> 00:23:11,109
到2035财年降至五个百分点

577
00:23:11,109 --> 00:23:12,250
有一定缓冲空间

578
00:23:12,250 --> 00:23:15,289
在二一点百分比自由现金流利润率

579
00:23:15,289 --> 00:23:17,219
达到十五个百分点

580
00:23:17,219 --> 00:23:23,419
在此采用三点五百分比终端增长率作为熊市情景

581
00:23:23,419 --> 00:23:25,558
我再次重复强调

582
00:23:25,558 --> 00:23:26,750
在熊市情景下

583
00:23:26,750 --> 00:23:30,769
不应过度压低股价

584
00:23:30,769 --> 00:23:32,750
仅仅出于防御目的

585
00:23:32,750 --> 00:23:34,219
需保持合理性

586
00:23:34,219 --> 00:23:36,318
在此熊市情景中

587
00:23:36,318 --> 00:23:39,419
仍采用基准情景

588
00:23:39,419 --> 00:23:42,219
但假设这些投资

589
00:23:42,219 --> 00:23:47,190
这些增长加速可能不会发生

590
00:23:47,190 --> 00:23:49,210
如我们预期的那样

591
00:23:49,210 --> 00:23:53,349
业务可能不如预期盈利

592
00:23:53,349 --> 00:23:56,400
这就是我们得出这些数据的原因

593
00:23:56,400 --> 00:23:57,680
这些营收增长

594
00:23:57,680 --> 00:23:59,099
这些利润率

595
00:23:59,099 --> 00:24:03,329
以及2035年的自由现金流利润率

596
00:24:03,329 --> 00:24:06,480
尤其是这里的终端增长率

597
00:24:06,480 --> 00:24:07,880
我设定为二点八个百分点

598
00:24:07,880 --> 00:24:10,279
终端增长率为2.9%

599
00:24:10,279 --> 00:24:10,920
有些人说

600
00:24:10,920 --> 00:24:13,179
哦你得定2%的利率

601
00:24:13,179 --> 00:24:13,499
我不确定

602
00:24:13,499 --> 00:24:15,849
我认为亚马逊应得2.8%

603
00:24:15,849 --> 00:24:17,410
现在是2.9%的增长率

604
00:24:17,410 --> 00:24:19,970
如果你想快速了解当前状况

605
00:24:19,970 --> 00:24:21,950
可以查看此处的隐含股价

606
00:24:21,950 --> 00:24:24,138
波动率与终端增长率

607
00:24:24,138 --> 00:24:26,759
今日股价意味着我们处于约2%

608
00:24:26,759 --> 00:24:27,419
5%的水平

609
00:24:27,419 --> 00:24:30,179
终端增长率5%和10%波动率

610
00:24:30,179 --> 00:24:31,898
如果我们稍微调高

611
00:24:31,898 --> 00:24:33,179
则为3%的增长率

612
00:24:33,179 --> 00:24:34,190
10%的波动率

613
00:24:34,190 --> 00:24:35,950
如果你想达到9%

614
00:24:35,950 --> 00:24:36,910
3%的波动率

615
00:24:36,910 --> 00:24:39,680
我们应该在205.8左右

616
00:24:39,680 --> 00:24:40,599
等等等等

617
00:24:40,599 --> 00:24:41,240
但再次强调

618
00:24:41,240 --> 00:24:42,680
如果你想尝试调整

619
00:24:42,680 --> 00:24:43,440
完全免费

620
00:24:43,440 --> 00:24:44,299
无需注册

621
00:24:44,299 --> 00:24:44,819
不收邮件

622
00:24:44,819 --> 00:24:45,519
无需任何操作

623
00:24:45,519 --> 00:24:46,319
可以自由尝试

624
00:24:46,319 --> 00:24:48,640
此处所有蓝色数字均可编辑

625
00:24:48,640 --> 00:24:49,799
这些将影响

626
00:24:49,799 --> 00:24:50,839
当然你的乐观

627
00:24:50,839 --> 00:24:52,819
悲观和基准情景

628
00:24:52,819 --> 00:24:53,900
可以自由调整

629
00:24:53,900 --> 00:24:55,140
故事需要合理

630
00:24:55,140 --> 00:24:56,859
不能随意填入数字就说

631
00:24:56,859 --> 00:25:00,710
哦看这里的隐含增长率现在20%

632
00:25:00,710 --> 00:25:05,039
仅仅因为需要股价更高而不符合逻辑

633
00:25:05,039 --> 00:25:07,779
基于当前故事

634
00:25:07,779 --> 00:25:09,079
基于我的假设

635
00:25:09,079 --> 00:25:09,700
是的

636
00:25:09,700 --> 00:25:12,299
乐观情景预估87.3%

637
00:25:12,299 --> 00:25:14,930
当前价格有3%的涨幅空间

638
00:25:14,930 --> 00:25:19,029
基准情景当前价格有2.6%的涨幅

639
00:25:19,029 --> 00:25:22,799
而悲观情景有3.5%的跌幅

640
00:25:22,799 --> 00:25:25,869
在下方评论区告诉我你的看法

641
00:25:25,869 --> 00:25:28,109
总的来说这就是我的全部内容

642
00:25:28,109 --> 00:25:29,009
在今天的视频中

643
00:25:29,009 --> 00:25:29,369
是的

644
00:25:29,369 --> 00:25:30,049
就我个人而言

645
00:25:30,049 --> 00:25:33,619
我认为Meta值得买入

646
00:25:33,619 --> 00:25:33,940
是的

647
00:25:33,940 --> 00:25:35,339
我认为亚马逊

648
00:25:35,339 --> 00:25:37,720
这是我的退休投资组合的一部分

649
00:25:37,720 --> 00:25:38,599
位列前三

650
00:25:38,599 --> 00:25:40,799
在Meta和Mercado Libre之后

651
00:25:40,799 --> 00:25:41,759
还有亚马逊

652
00:25:41,759 --> 00:25:41,960
在那里

653
00:25:41,960 --> 00:25:44,240
我也认为这几乎是毫无疑问的

654
00:25:44,240 --> 00:25:44,640
是的

655
00:25:44,640 --> 00:25:47,819
我仍然认为这些大科技公司是

656
00:25:47,819 --> 00:25:50,950
值得持有的优质资产，安心入睡

657
00:25:50,950 --> 00:25:54,109
你可能不会获得十倍回报

658
00:25:54,109 --> 00:25:57,650
但我确实认为你仍能获得市场超越回报

659
00:25:57,650 --> 00:26:00,859
而且无需每日焦虑

660
00:26:00,859 --> 00:26:01,559
没错

661
00:26:01,559 --> 00:26:05,380
这些公司目前估值极具吸引力

662
00:26:05,380 --> 00:26:06,059
现在

663
00:26:06,059 --> 00:26:07,140
它们极其强大

664
00:26:07,140 --> 00:26:08,779
它们极其盈利

665
00:26:08,779 --> 00:26:09,920
这其中存在风险吗

666
00:26:09,920 --> 00:26:10,200
是的

667
00:26:10,200 --> 00:26:11,940
每家公司都有其风险

668
00:26:11,940 --> 00:26:13,779
只是你支付的价格

669
00:26:13,779 --> 00:26:16,430
现在看来这很合理

670
00:26:16,430 --> 00:26:17,630
告诉我你的看法吧

671
00:26:17,630 --> 00:26:18,789
在下方评论区

672
00:26:18,789 --> 00:26:20,349
我们下次再见

673
00:26:20,349 --> 00:26:21,240
再见

共 4 条，按点赞数排序。

采集完成，该视频无弹幕