WEBVTT

1
00:00:00.088 --> 00:00:04.911
So we often think being well-informed is just about keeping up with the daily news cycle,

2
00:00:04.973 --> 00:00:05.192
right?

3
00:00:05.493 --> 00:00:05.614
Yeah,

4
00:00:05.673 --> 00:00:06.133
absolutely.

5
00:00:06.575 --> 00:00:09.216
But what if truly understanding what's going on,

6
00:00:09.317 --> 00:00:10.356
especially financially,

7
00:00:11.098 --> 00:00:13.481
meant seeing these underlying patterns?

8
00:00:13.583 --> 00:00:17.778
Patterns that could maybe signal where a company's stock is headed?

9
00:00:17.919 --> 00:00:18.372
Exactly.

10
00:00:18.590 --> 00:00:20.325
That's the kind of deeper level we're getting into today.

11
00:00:20.528 --> 00:00:24.247
We actually got a really interesting piece of academic research sent in by one of you listening.

12
00:00:24.325 --> 00:00:24.450
Oh,

13
00:00:24.512 --> 00:00:24.934
great.

14
00:00:25.075 --> 00:00:25.200
Yeah,

15
00:00:25.309 --> 00:00:27.169
and it looks into just that kind of pattern.

16
00:00:27.622 --> 00:00:28.856
Maybe a surprising clue.

17
00:00:29.524 --> 00:00:30.904
about future stock performance.

18
00:00:31.184 --> 00:00:31.404
Okay,

19
00:00:31.584 --> 00:00:32.365
I like the sound of that.

20
00:00:33.086 --> 00:00:37.326
So today we're doing a deep dive into something called high accruals momentum.

21
00:00:38.388 --> 00:00:39.326
Sounds a bit technical,

22
00:00:39.365 --> 00:00:39.607
maybe.

23
00:00:39.826 --> 00:00:40.248
It does,

24
00:00:40.427 --> 00:00:40.787
it does.

25
00:00:41.068 --> 00:00:42.084
But the core idea,

26
00:00:42.263 --> 00:00:42.529
I think,

27
00:00:42.545 --> 00:00:43.568
is pretty insightful.

28
00:00:44.170 --> 00:00:45.326
The research is behind the paper.

29
00:00:45.404 --> 00:00:47.591
It's titled High Accruals Momentum.

30
00:00:47.873 --> 00:00:48.310
Shouting Hao,

31
00:00:48.795 --> 00:00:49.591
Juwan Zhang,

32
00:00:49.748 --> 00:00:50.482
and Yunzhu Li.

33
00:00:50.904 --> 00:00:51.248
Right.

34
00:00:51.623 --> 00:00:57.873
They basically suggest that if a company consistently reports high levels of what they call discretionary accruals.

35
00:00:58.552 --> 00:00:59.573
four years straight.

36
00:00:59.593 --> 00:00:59.993
Four years,

37
00:01:00.014 --> 00:01:00.174
yeah.

38
00:01:00.334 --> 00:01:03.438
That could actually be a pretty big warning sign for investors.

39
00:01:03.899 --> 00:01:05.700
And this isn't just a small study either.

40
00:01:06.020 --> 00:01:07.160
They looked at a lot of data.

41
00:01:07.200 --> 00:01:10.825
We're talking 1980 right through to 2016.

42
00:01:10.826 --> 00:01:10.887
Wow.

43
00:01:10.888 --> 00:01:11.004
Yeah,

44
00:01:11.106 --> 00:01:12.145
covering NYSE,

45
00:01:12.504 --> 00:01:13.004
Amex,

46
00:01:13.207 --> 00:01:13.684
Nasdaq,

47
00:01:13.910 --> 00:01:14.684
common stocks.

48
00:01:15.075 --> 00:01:17.028
So it's pretty robust.

49
00:01:17.309 --> 00:01:17.512
Okay,

50
00:01:17.575 --> 00:01:19.731
so our mission today then is to unpack this.

51
00:01:20.418 --> 00:01:22.668
What is high accruals momentum?

52
00:01:22.996 --> 00:01:25.215
What are these discretionary accruals anyway?

53
00:01:25.684 --> 00:01:27.559
Why might it matter to you listening?

54
00:01:28.060 --> 00:01:28.260
And,

55
00:01:28.300 --> 00:01:28.541
you know,

56
00:01:28.701 --> 00:01:31.865
what did the research actually find about the impact on stock returns?

57
00:01:31.923 --> 00:01:32.705
We want to make it clear,

58
00:01:32.763 --> 00:01:32.966
right?

59
00:01:33.166 --> 00:01:36.947
Give you those actionable insights without drowning you in accounting jargon.

60
00:01:37.025 --> 00:01:37.572
Exactly.

61
00:01:37.728 --> 00:01:38.611
Aiming for clarity.

62
00:01:38.767 --> 00:01:38.892
So,

63
00:01:39.009 --> 00:01:39.134
OK,

64
00:01:39.330 --> 00:01:40.189
let's start at the beginning.

65
00:01:40.314 --> 00:01:41.033
Accruals.

66
00:01:41.509 --> 00:01:42.134
In accounting,

67
00:01:42.173 --> 00:01:43.033
what are we talking about?

68
00:01:43.173 --> 00:01:43.298
OK,

69
00:01:43.478 --> 00:01:43.931
simply put,

70
00:01:44.103 --> 00:01:46.697
accruals are basically about a timing difference.

71
00:01:47.056 --> 00:01:50.197
It's the gap between when a business transaction actually happens.

72
00:01:50.384 --> 00:01:52.681
Like making a sale or providing a service.

73
00:01:52.822 --> 00:01:53.041
Right.

74
00:01:53.338 --> 00:01:55.150
And when the cash actually changes hands.

75
00:01:55.744 --> 00:01:57.369
So you might record revenue when you earn it.

76
00:01:57.560 --> 00:01:58.800
even if the customer pays you later.

77
00:01:59.060 --> 00:02:00.881
That difference is managed through accruals.

78
00:02:01.241 --> 00:02:01.479
Okay,

79
00:02:01.581 --> 00:02:01.881
got it.

80
00:02:01.901 --> 00:02:04.463
So it reflects the business activity,

81
00:02:04.643 --> 00:02:06.565
not just the cash in the bank account rate this second.

82
00:02:07.565 --> 00:02:07.745
Now,

83
00:02:07.784 --> 00:02:10.581
the research specifically mentions discretionary accruals.

84
00:02:10.799 --> 00:02:12.042
What makes them discretionary?

85
00:02:12.760 --> 00:02:12.901
Yeah,

86
00:02:13.002 --> 00:02:13.846
this is the key part.

87
00:02:14.221 --> 00:02:16.565
So some accruals are pretty standard,

88
00:02:16.612 --> 00:02:18.784
very predictable based on how the business operates.

89
00:02:18.877 --> 00:02:21.440
Think like standard accounts receivable.

90
00:02:22.502 --> 00:02:25.721
But there's another component where management has more judgment.

91
00:02:25.992 --> 00:02:28.713
more flexibility in how they estimate and report things.

92
00:02:29.034 --> 00:02:30.332
That's the discretionary piece.

93
00:02:30.493 --> 00:02:30.754
Ah,

94
00:02:30.915 --> 00:02:32.055
so it's less automatic,

95
00:02:32.133 --> 00:02:33.032
more subjective.

96
00:02:33.172 --> 00:02:33.594
Exactly.

97
00:02:34.094 --> 00:02:34.797
And researchers,

98
00:02:34.977 --> 00:02:35.618
like in this paper,

99
00:02:36.235 --> 00:02:37.735
use specific accounting models.

100
00:02:37.954 --> 00:02:42.415
You might hear names like the Jones model or the modified Jones model to try and estimate that part.

101
00:02:42.883 --> 00:02:46.071
They're trying to isolate the atrules that reflect management's choices.

102
00:02:46.321 --> 00:02:46.540
Okay,

103
00:02:46.821 --> 00:02:47.258
choices.

104
00:02:47.321 --> 00:02:49.774
So why might management use this flexibility,

105
00:02:49.852 --> 00:02:50.680
this discretion?

106
00:02:50.774 --> 00:02:51.680
What's the incentive?

107
00:02:52.211 --> 00:02:52.352
Well,

108
00:02:52.353 --> 00:02:53.211
the big implication,

109
00:02:53.696 --> 00:02:55.243
and what the research digs into,

110
00:02:55.868 --> 00:02:58.269
is the potential to manage reported earnings.

111
00:02:58.448 --> 00:02:58.589
Oh,

112
00:02:58.849 --> 00:02:59.028
okay.

113
00:02:59.208 --> 00:03:00.509
So smoothing things out,

114
00:03:00.907 --> 00:03:05.411
maybe making performance look a bit better or more consistent than the raw cash flow might suggest.

115
00:03:05.747 --> 00:03:06.388
That's the idea.

116
00:03:06.849 --> 00:03:09.872
It's about potentially influencing how the company's performance is perceived.

117
00:03:10.028 --> 00:03:10.208
Okay,

118
00:03:10.310 --> 00:03:10.692
interesting.

119
00:03:10.989 --> 00:03:14.145
So we have accruals and this discretionary part management can influence.

120
00:03:14.270 --> 00:03:16.536
How does this connect to the main idea,

121
00:03:17.552 --> 00:03:18.630
high accruals momentum?

122
00:03:18.849 --> 00:03:19.099
Right.

123
00:03:19.114 --> 00:03:23.036
So the study defines high accruals momentum very clearly.

124
00:03:23.770 --> 00:03:24.427
It's when a company

125
00:03:24.816 --> 00:03:27.597
doesn't just have high discretionary accruals in one year,

126
00:03:28.338 --> 00:03:32.357
but consistently reports high levels for four years in a row.

127
00:03:32.775 --> 00:03:34.002
Four consecutive years,

128
00:03:34.041 --> 00:03:35.916
that feels like a deliberate pattern,

129
00:03:35.978 --> 00:03:37.595
not just a one-off adjustment.

130
00:03:37.861 --> 00:03:38.338
Exactly.

131
00:03:38.541 --> 00:03:41.002
The researchers see it as a sustained pattern,

132
00:03:41.283 --> 00:03:45.455
potentially indicating that management is consistently using these accounting judgments,

133
00:03:45.798 --> 00:03:48.048
perhaps to inflate reported earnings over time.

134
00:03:48.486 --> 00:03:49.595
Is this something you see a lot?

135
00:03:49.705 --> 00:03:52.127
Is it common for companies to have this four-year streak?

136
00:03:52.544 --> 00:03:52.684
Well,

137
00:03:52.685 --> 00:03:54.266
that's actually one of the interesting findings.

138
00:03:54.286 --> 00:03:54.406
No,

139
00:03:54.465 --> 00:03:55.587
it's relatively rare.

140
00:03:55.868 --> 00:03:56.028
Oh.

141
00:03:56.108 --> 00:03:56.227
Yeah.

142
00:03:56.268 --> 00:03:57.589
Across that whole period,

143
00:03:57.590 --> 00:03:59.049
1980 to 2016,

144
00:03:59.549 --> 00:04:02.092
they found high accruals momentum in only about 1%

145
00:04:02.733 --> 00:04:04.960
of all the company year observations they looked at.

146
00:04:05.014 --> 00:04:05.460
Only 1%.

147
00:04:06.077 --> 00:04:06.421
Wow.

148
00:04:07.421 --> 00:04:08.217
And on the flip side,

149
00:04:08.975 --> 00:04:10.116
low accruals momentum,

150
00:04:10.522 --> 00:04:13.944
consistently low discretionary accruals for four years.

151
00:04:14.163 --> 00:04:15.053
That was also pretty rare,

152
00:04:15.069 --> 00:04:16.928
maybe around 1.4%.

153
00:04:16.929 --> 00:04:20.288
So the fact that it's uncommon might make it a stronger signal when you do see it?

154
00:04:20.491 --> 00:04:21.225
That's the thinking.

155
00:04:21.408 --> 00:04:22.269
It's not just noise.

156
00:04:22.329 --> 00:04:23.951
It's a distinct pattern that stands out.

157
00:04:24.230 --> 00:04:25.012
When it does happen,

158
00:04:25.351 --> 00:04:27.894
the research suggests maybe we should pay attention.

159
00:04:27.976 --> 00:04:28.094
OK,

160
00:04:28.273 --> 00:04:28.773
that makes sense.

161
00:04:29.336 --> 00:04:33.984
So what kinds of companies tended to show this rare pattern?

162
00:04:34.023 --> 00:04:36.523
Did the research identify any common characteristics?

163
00:04:37.023 --> 00:04:37.265
Yes,

164
00:04:37.828 --> 00:04:41.765
they did find some common traits among these high accruals momentum firms.

165
00:04:42.250 --> 00:04:44.594
They tended to be smaller companies,

166
00:04:44.687 --> 00:04:45.859
lower market capitalization.

167
00:04:45.906 --> 00:04:46.031
OK,

168
00:04:46.156 --> 00:04:46.750
smaller firms.

169
00:04:46.953 --> 00:04:47.375
Anything else?

170
00:04:47.547 --> 00:04:49.422
They also typically had lower leverage ratios,

171
00:04:49.781 --> 00:04:50.859
so less reliant on

172
00:04:51.052 --> 00:04:51.673
debt financing.

173
00:04:51.913 --> 00:04:52.353
Interesting.

174
00:04:52.554 --> 00:04:52.933
Smaller,

175
00:04:52.994 --> 00:04:53.394
less debt.

176
00:04:53.836 --> 00:04:54.215
What else?

177
00:04:54.355 --> 00:04:55.316
And interestingly,

178
00:04:55.375 --> 00:04:57.476
they often had higher book to market ratios,

179
00:04:57.477 --> 00:04:57.680
you know,

180
00:04:57.681 --> 00:05:00.023
the kind of ratios sometimes associated with value stocks.

181
00:05:00.383 --> 00:05:02.820
And they also showed a higher ROA return on assets.

182
00:05:04.411 --> 00:05:05.511
That's a bit of a mixed bag.

183
00:05:05.931 --> 00:05:06.371
Smaller,

184
00:05:06.552 --> 00:05:07.031
less debt,

185
00:05:07.132 --> 00:05:10.453
but high book to market and seemingly high profitability based on ROA.

186
00:05:11.992 --> 00:05:15.953
Any theories on why these traits cluster together with high accruals momentum?

187
00:05:16.593 --> 00:05:16.710
Well,

188
00:05:16.711 --> 00:05:18.117
the researchers speculate a bit.

189
00:05:18.210 --> 00:05:19.210
For smaller firms,

190
00:05:19.312 --> 00:05:20.835
maybe there's just less scrutiny,

191
00:05:21.242 --> 00:05:22.742
fewer analysts following them,

192
00:05:22.835 --> 00:05:24.570
perhaps giving management more leeway.

193
00:05:24.679 --> 00:05:24.898
Right.

194
00:05:24.960 --> 00:05:25.710
Less spotlight.

195
00:05:25.882 --> 00:05:31.960
The lower leverage maybe means less pressure from creditors or maybe it's linked to the way they're managing earnings,

196
00:05:31.992 --> 00:05:33.992
perhaps making them look like they don't need debt.

197
00:05:34.812 --> 00:05:35.712
And the high ROA,

198
00:05:37.054 --> 00:05:37.193
well,

199
00:05:37.334 --> 00:05:38.154
that could create pressure,

200
00:05:38.195 --> 00:05:38.537
couldn't it?

201
00:05:39.095 --> 00:05:40.919
Pressure to keep reporting strong numbers,

202
00:05:41.240 --> 00:05:45.482
maybe leading them to use discretionary accruals if the underlying business isn't quite keeping up.

203
00:05:45.865 --> 00:05:45.982
OK,

204
00:05:46.099 --> 00:05:46.865
that paints a picture.

205
00:05:47.201 --> 00:05:48.162
So we know what it is.

206
00:05:48.185 --> 00:05:49.029
Who tends to show it?

207
00:05:49.084 --> 00:05:49.209
Now,

208
00:05:49.210 --> 00:05:49.966
the big question,

209
00:05:50.490 --> 00:05:52.927
what did the research find about future stock performance?

210
00:05:53.052 --> 00:05:55.068
Is this pattern actually a red flag?

211
00:05:55.271 --> 00:05:56.896
This is really the core of the study.

212
00:05:57.193 --> 00:05:57.912
And yes,

213
00:05:57.959 --> 00:05:59.052
the findings are pretty stark.

214
00:05:59.459 --> 00:06:03.802
Firms identified with high accruals momentum tended to have significantly lower stock returns.

215
00:06:03.818 --> 00:06:03.943
and

216
00:06:04.283 --> 00:06:05.224
the following periods.

217
00:06:05.584 --> 00:06:06.244
Lower returns.

218
00:06:06.565 --> 00:06:06.766
Okay.

219
00:06:06.767 --> 00:06:07.868
And what about the opposite,

220
00:06:08.207 --> 00:06:09.969
the low accruals momentum firms?

221
00:06:10.028 --> 00:06:11.008
They found the reverse.

222
00:06:11.411 --> 00:06:17.700
Firms with consistently low discretionary accruals tended to experience higher subsequent stock returns.

223
00:06:18.020 --> 00:06:23.864
So it really seems like the market eventually reacts negatively to that sustained high accruals pattern.

224
00:06:24.098 --> 00:06:24.801
It suggests that,

225
00:06:25.145 --> 00:06:25.426
yes,

226
00:06:26.504 --> 00:06:30.832
the market might initially be fooled or maybe just slow to recognize the pattern,

227
00:06:30.911 --> 00:06:32.223
but eventually it seems to catch up.

228
00:06:32.883 --> 00:06:34.045
Can we put some numbers on that?

229
00:06:34.324 --> 00:06:36.306
How significant was the difference in returns?

230
00:06:36.467 --> 00:06:43.511
They tested a strategy basically going long the low accruals momentum stocks and shorting the high accruals momentum stocks.

231
00:06:43.512 --> 00:06:44.316
A hedge portfolio,

232
00:06:44.371 --> 00:06:44.511
yeah.

233
00:06:44.816 --> 00:06:45.379
Exactly.

234
00:06:45.472 --> 00:06:50.816
And they found the strategy generated statistically significant positive abnormal returns,

235
00:06:50.879 --> 00:06:54.629
meaning returns above what you'd expect given the risk over the next three,

236
00:06:54.707 --> 00:06:55.629
six and 12 months.

237
00:06:55.769 --> 00:06:56.972
And the size of those returns?

238
00:06:57.269 --> 00:06:57.425
Well,

239
00:06:57.504 --> 00:07:02.566
one calculation they ran suggested an annualized abnormal return of around 13.2%.

240
00:07:02.663 --> 00:07:03.504
13%.

241
00:07:04.184 --> 00:07:04.985
That's not trivial.

242
00:07:05.065 --> 00:07:07.688
That's a substantial potential edge if this holds.

243
00:07:07.809 --> 00:07:10.071
It's definitely a number that makes you sit up and take notice.

244
00:07:10.493 --> 00:07:10.868
Absolutely.

245
00:07:11.172 --> 00:07:11.290
So,

246
00:07:11.852 --> 00:07:12.032
okay,

247
00:07:12.094 --> 00:07:12.290
why?

248
00:07:12.735 --> 00:07:13.555
Why does this happen?

249
00:07:14.415 --> 00:07:16.938
The researchers dug into two main possibilities,

250
00:07:17.000 --> 00:07:17.219
right?

251
00:07:17.680 --> 00:07:20.313
Was it just earnings management or something else?

252
00:07:20.813 --> 00:07:21.297
That's right.

253
00:07:21.407 --> 00:07:23.110
They considered two main hypotheses.

254
00:07:23.704 --> 00:07:24.188
The first,

255
00:07:24.282 --> 00:07:25.329
as we've kind of hinted at,

256
00:07:25.438 --> 00:07:27.250
is the earnings management explanation.

257
00:07:27.791 --> 00:07:33.598
The idea that managers are actively manipulating the numbers and eventually the market figures it out or there's a correction.

258
00:07:34.117 --> 00:07:34.699
Precisely.

259
00:07:35.320 --> 00:07:40.484
Maybe future earnings disappoint or there are restatements and the stock price adjusts downwards.

260
00:07:41.024 --> 00:07:43.164
The second idea was the growth anomaly.

261
00:07:43.547 --> 00:07:43.688
Ah,

262
00:07:43.742 --> 00:07:48.492
the puzzle where high growth stocks sometimes underperform lower growth or value stocks.

263
00:07:48.570 --> 00:07:49.055
Exactly.

264
00:07:49.305 --> 00:07:50.133
So the question was,

265
00:07:50.836 --> 00:07:56.695
maybe these higher cruels are just a sign of a high growth company and the subsequent low returns are just part of that phenomenon.

266
00:07:56.983 --> 00:07:58.884
Not necessarily related to shady accounting.

267
00:07:59.185 --> 00:07:59.304
OK,

268
00:07:59.646 --> 00:08:00.746
two plausible stories.

269
00:08:00.906 --> 00:08:02.750
How did they try to disentangle those?

270
00:08:02.769 --> 00:08:05.949
How did they test whether it was really about earnings management or just growth?

271
00:08:06.254 --> 00:08:07.675
They did some clever analysis.

272
00:08:07.996 --> 00:08:10.855
They looked at the relationship between accruals,

273
00:08:10.918 --> 00:08:12.675
momentum and future returns.

274
00:08:13.144 --> 00:08:17.957
But then they specifically controlled for various measures of actual company growth.

275
00:08:18.269 --> 00:08:18.957
What kind of measures?

276
00:08:19.160 --> 00:08:20.785
Things like the book to market ratio again.

277
00:08:20.847 --> 00:08:21.050
Right.

278
00:08:21.410 --> 00:08:22.910
But also asset growth,

279
00:08:23.425 --> 00:08:25.957
growth in the number of employees trying to get a handle.

280
00:08:26.207 --> 00:08:28.088
on how much these firms were really expanding.

281
00:08:28.909 --> 00:08:29.030
OK,

282
00:08:29.229 --> 00:08:31.491
so they separated the firms based on growth.

283
00:08:31.792 --> 00:08:32.573
What happened then?

284
00:08:33.073 --> 00:08:38.459
Did the low return effect disappear for the high growth firms that also had high accruals momentum?

285
00:08:39.100 --> 00:08:39.397
Actually,

286
00:08:39.522 --> 00:08:39.639
no,

287
00:08:40.022 --> 00:08:41.178
it was kind of the opposite.

288
00:08:41.202 --> 00:08:42.288
This was a key finding.

289
00:08:42.725 --> 00:08:47.413
The negative link between high accruals momentum and future returns was stronger,

290
00:08:47.772 --> 00:08:48.913
more pronounced for firms.

291
00:08:49.219 --> 00:08:52.002
that actually had low growth according to those indicators.

292
00:08:52.061 --> 00:08:52.262
Wow.

293
00:08:52.342 --> 00:08:52.563
Okay.

294
00:08:52.623 --> 00:08:55.905
So if you have high accruals momentum and you're not actually a high growth company,

295
00:08:56.307 --> 00:08:58.248
the market reacts even more negatively later.

296
00:08:58.327 --> 00:08:59.670
That's what the evidence suggests.

297
00:08:59.811 --> 00:09:01.952
So what does that tell us about the two explanations?

298
00:09:02.233 --> 00:09:05.873
It strongly points toward the earnings management explanation being more likely.

299
00:09:06.717 --> 00:09:08.952
If it were just a proxy for growth,

300
00:09:09.436 --> 00:09:12.436
you wouldn't expect the effect to be stronger in low growth firms.

301
00:09:12.623 --> 00:09:12.920
Right.

302
00:09:12.998 --> 00:09:18.155
It suggests investors become particularly skeptical when they see high accruals that aren't backed up.

303
00:09:18.271 --> 00:09:20.373
by strong underlying business growth.

304
00:09:20.895 --> 00:09:23.438
It raises questions about the quality of those earnings.

305
00:09:23.496 --> 00:09:23.938
Exactly.

306
00:09:24.016 --> 00:09:29.203
It makes the high accruals look less like investment for the future and more like,

307
00:09:30.242 --> 00:09:30.406
well,

308
00:09:30.727 --> 00:09:31.281
something else.

309
00:09:31.344 --> 00:09:31.461
OK,

310
00:09:31.602 --> 00:09:33.102
that's a really important distinction.

311
00:09:33.164 --> 00:09:33.328
Now,

312
00:09:33.484 --> 00:09:35.305
you mentioned earlier this general idea,

313
00:09:35.492 --> 00:09:36.664
the accruals anomaly,

314
00:09:37.383 --> 00:09:40.352
that high accruals often mean lower future returns.

315
00:09:40.353 --> 00:09:41.352
That's been around a while.

316
00:09:41.836 --> 00:09:44.242
But didn't the study say something about that maybe fading?

317
00:09:44.570 --> 00:09:44.899
Yes,

318
00:09:44.961 --> 00:09:46.680
that's another fascinating layer to this.

319
00:09:47.131 --> 00:09:50.175
There's other research suggesting that the general accruals anomaly,

320
00:09:50.354 --> 00:09:52.274
just looking at one year's worth of accruals,

321
00:09:52.657 --> 00:09:54.700
might have weakened or even disappeared in the U.S.

322
00:09:54.759 --> 00:09:55.020
market,

323
00:09:55.040 --> 00:09:57.180
particularly after maybe 2004 or so.

324
00:09:57.344 --> 00:09:59.165
Maybe because investors got wise to it.

325
00:09:59.962 --> 00:10:02.063
Or Sarbanes-Oxley changed things.

326
00:10:02.478 --> 00:10:03.799
Could be a combination of factors.

327
00:10:04.140 --> 00:10:04.921
Markets adapt,

328
00:10:05.121 --> 00:10:06.121
regulations change.

329
00:10:06.621 --> 00:10:12.906
But the basic idea is that the simple one-year accrual signal might not be as potent as it once was.

330
00:10:13.187 --> 00:10:13.312
OK,

331
00:10:13.507 --> 00:10:15.328
so if the general anomaly faded,

332
00:10:15.554 --> 00:10:17.070
what about this momentum version?

333
00:10:17.593 --> 00:10:20.679
Did the four-year pattern also lose its predictive power?

334
00:10:21.023 --> 00:10:22.023
And this is the kicker?

335
00:10:22.757 --> 00:10:22.882
No,

336
00:10:23.132 --> 00:10:23.585
it didn't.

337
00:10:24.304 --> 00:10:27.882
The researchers found that the predictive power of high accruals momentum,

338
00:10:27.945 --> 00:10:28.882
that four-year streak,

339
00:10:29.429 --> 00:10:31.226
remains strong and significant.

340
00:10:31.438 --> 00:10:34.861
even in the periods after the general anomaly seemed to dissipate.

341
00:10:35.122 --> 00:10:35.443
Whoa.

342
00:10:36.142 --> 00:10:36.321
Okay,

343
00:10:36.345 --> 00:10:38.505
so this isn't just the old anomaly in disguise.

344
00:10:38.607 --> 00:10:39.767
It's something distinct.

345
00:10:40.087 --> 00:10:40.946
That's what it looks like.

346
00:10:41.368 --> 00:10:43.532
It suggests that looking at the consistency,

347
00:10:43.892 --> 00:10:44.814
the multi-year pattern,

348
00:10:45.173 --> 00:10:46.415
provides different information,

349
00:10:46.470 --> 00:10:48.579
information that the market still seems to react to,

350
00:10:49.157 --> 00:10:52.392
even if it's gotten better at pricing in single-year accrual levels.

351
00:10:52.564 --> 00:10:53.220
That's really interesting.

352
00:10:53.221 --> 00:10:55.267
So the persistence of the behavior matters.

353
00:10:55.298 --> 00:10:55.939
It seems so.

354
00:10:56.423 --> 00:10:58.189
A sustained pattern sends a different,

355
00:10:58.298 --> 00:10:59.173
perhaps stronger,

356
00:10:59.564 --> 00:10:59.845
signal.

357
00:11:00.018 --> 00:11:02.200
Did they look at anything else like how this interacts with,

358
00:11:02.260 --> 00:11:02.481
say,

359
00:11:02.760 --> 00:11:03.961
earnings surprises?

360
00:11:04.180 --> 00:11:04.563
They did.

361
00:11:04.844 --> 00:11:08.508
They wanted to see if the market's reaction changed depending on whether a company hit,

362
00:11:08.868 --> 00:11:09.086
beat,

363
00:11:09.227 --> 00:11:10.508
or missed its earnings targets.

364
00:11:10.868 --> 00:11:16.110
And does a negative surprise make things worse for a high accruals momentum firm?

365
00:11:16.188 --> 00:11:16.391
Yes.

366
00:11:17.016 --> 00:11:26.829
They found the negative relationship between high accruals momentum and subsequent returns was significantly more pronounced for companies that also reported negative earnings surprises.

367
00:11:26.860 --> 00:11:28.188
That makes intuitive sense.

368
00:11:28.344 --> 00:11:29.500
If you've got this history of...

369
00:11:29.538 --> 00:11:31.040
of potentially aggressive accounting,

370
00:11:31.340 --> 00:11:32.581
and then you miss your numbers.

371
00:11:32.681 --> 00:11:32.962
Right.

372
00:11:33.382 --> 00:11:34.905
Kind of validates the skepticism,

373
00:11:34.964 --> 00:11:35.405
doesn't it?

374
00:11:35.765 --> 00:11:36.585
Investors might think,

375
00:11:36.706 --> 00:11:36.827
ah,

376
00:11:37.046 --> 00:11:37.226
see,

377
00:11:37.284 --> 00:11:39.030
those high accruals weren't sustainable,

378
00:11:39.569 --> 00:11:40.686
and react more strongly.

379
00:11:41.069 --> 00:11:46.913
It suggests maybe investors scrutinize that accrual history more closely when things start to go wrong operationally.

380
00:11:47.116 --> 00:11:47.241
OK.

381
00:11:47.976 --> 00:11:48.772
And just to be thorough,

382
00:11:48.804 --> 00:11:49.944
they did all the usual checks,

383
00:11:50.022 --> 00:11:50.257
right?

384
00:11:50.413 --> 00:11:51.538
Robustness tests,

385
00:11:51.632 --> 00:11:55.069
making sure this wasn't just some fluke in the data or how they measured things.

386
00:11:55.116 --> 00:11:55.241
Oh,

387
00:11:55.257 --> 00:11:55.726
absolutely.

388
00:11:55.788 --> 00:11:57.741
Standard practice and good academic research.

389
00:11:58.190 --> 00:12:00.432
They tried measuring momentum over three years,

390
00:12:00.954 --> 00:12:01.512
five years,

391
00:12:01.553 --> 00:12:02.235
not just four.

392
00:12:02.614 --> 00:12:03.735
Looked at total accruals,

393
00:12:03.797 --> 00:12:04.837
not just discretionary.

394
00:12:05.196 --> 00:12:07.602
Even looked at non-discretionary accruals momentum.

395
00:12:07.962 --> 00:12:08.157
Okay.

396
00:12:08.422 --> 00:12:10.399
They checked pre and post Sarbanes-Oxley.

397
00:12:11.024 --> 00:12:11.782
Across the board,

398
00:12:11.962 --> 00:12:13.587
the main finding generally held up.

399
00:12:14.251 --> 00:12:19.297
High discretionary accruals momentum consistently pointed towards lower future returns.

400
00:12:19.532 --> 00:12:22.251
And did they find anything interesting about the non-discretionary part,

401
00:12:22.252 --> 00:12:24.141
the more operational accruals?

402
00:12:24.704 --> 00:12:24.954
Yes,

403
00:12:25.297 --> 00:12:25.672
actually.

404
00:12:26.026 --> 00:12:32.150
they found that consistently low non-discretionary accruals momentum could also be seen negatively,

405
00:12:32.912 --> 00:12:37.080
which kind of makes sense if the basic operational accruals are consistently low.

406
00:12:37.478 --> 00:12:40.377
Maybe that signals weakness in the core business itself.

407
00:12:40.783 --> 00:12:40.939
Huh.

408
00:12:41.681 --> 00:12:43.642
So it's not just about the discretionary choices.

409
00:12:43.767 --> 00:12:48.283
Extremely low operational accruals might be a different kind of warning sign about fundamental health.

410
00:12:48.361 --> 00:12:48.814
Potentially,

411
00:12:48.845 --> 00:12:49.095
yes.

412
00:12:49.314 --> 00:12:50.642
It adds another layer of nuance.

413
00:12:50.830 --> 00:12:50.986
Okay.

414
00:12:50.987 --> 00:12:51.908
So wrapping this all up,

415
00:12:51.955 --> 00:12:53.611
what's the main takeaway here for you?

416
00:12:53.946 --> 00:12:55.168
listening to this deep dive.

417
00:12:55.207 --> 00:12:57.689
I think the big message is that consistency matters.

418
00:12:57.990 --> 00:12:59.713
A sustained pattern in this case,

419
00:13:00.353 --> 00:13:06.662
four years of high discretionary accruals seems to be a pretty meaningful negative signal for future stock performance.

420
00:13:06.998 --> 00:13:09.803
And it's not just capturing the old accruals anomaly.

421
00:13:10.342 --> 00:13:12.779
It seems to be providing distinct information,

422
00:13:13.232 --> 00:13:19.529
possibly linked to sustained earnings management that persists even when the market might have adjusted to simpler signal.

423
00:13:19.732 --> 00:13:20.217
Exactly.

424
00:13:20.389 --> 00:13:22.795
It's that multi-year aspect that seems key.

425
00:13:23.182 --> 00:13:24.422
So practically speaking,

426
00:13:24.623 --> 00:13:26.422
while obviously not investment advice,

427
00:13:26.563 --> 00:13:32.180
this suggests that looking beyond just one quarter or one year's financials could be really valuable.

428
00:13:32.727 --> 00:13:34.782
Watching for these kinds of persistent patterns,

429
00:13:34.821 --> 00:13:36.266
especially maybe in those smaller,

430
00:13:36.368 --> 00:13:38.508
lower leverage firms the study highlighted.

431
00:13:38.688 --> 00:13:38.946
Right.

432
00:13:39.461 --> 00:13:41.071
It encourages digging a bit deeper,

433
00:13:41.352 --> 00:13:45.180
looking at trends over time and how a company reports its performance.

434
00:13:45.743 --> 00:13:49.227
It's another piece of the puzzle to consider when you're evaluating a company.

435
00:13:49.508 --> 00:13:50.664
Definitely food for thought.

436
00:13:50.830 --> 00:13:51.731
It really makes you wonder,

437
00:13:51.751 --> 00:13:52.112
doesn't it,

438
00:13:52.131 --> 00:13:53.954
what other multi-year patterns,

439
00:13:53.973 --> 00:13:55.254
maybe beyond just accruals,

440
00:13:55.633 --> 00:13:58.700
might be hiding clues about a company's future that we should be looking for.

441
00:13:59.098 --> 00:14:00.379
That's a great question to ponder.

442
00:14:00.840 --> 00:14:03.520
Are there other consistent signals in financial statements,

443
00:14:03.801 --> 00:14:06.387
maybe in cash flows or investment patterns,

444
00:14:06.707 --> 00:14:08.473
that can tell a similar story over time?

445
00:14:08.832 --> 00:14:10.301
Something for all of us to think about.

446
00:14:10.395 --> 00:14:10.723
And hey,

447
00:14:10.770 --> 00:14:15.161
if you come across research or have thoughts on that or anything else you'd like us to explore,

448
00:14:15.254 --> 00:14:16.192
please send it our way.

449
00:14:16.582 --> 00:14:19.457
We love getting suggestions for future deep dives.

450
00:14:19.551 --> 00:14:20.036
Absolutely.

451
00:14:20.314 --> 00:14:25.399
keep them coming thanks everyone for joining us for this depth dive into hyacruel's momentum