WEBVTT

1
00:00:00.124 --> 00:00:00.244
OK,

2
00:00:00.404 --> 00:00:01.866
let's let's dig into this.

3
00:00:02.706 --> 00:00:03.206
Thinking back,

4
00:00:03.288 --> 00:00:03.526
you know,

5
00:00:03.747 --> 00:00:04.288
to the early

6
00:00:04.890 --> 00:00:05.147
2000s. Oh,

7
00:00:05.148 --> 00:00:05.308
yeah.

8
00:00:05.729 --> 00:00:06.311
Enron,

9
00:00:06.550 --> 00:00:07.370
WorldCom,

10
00:00:07.952 --> 00:00:08.296
Tyco,

11
00:00:08.413 --> 00:00:08.851
ESOL.

12
00:00:09.694 --> 00:00:10.452
It just felt like,

13
00:00:10.874 --> 00:00:11.171
wow,

14
00:00:11.358 --> 00:00:13.218
another accounting scandal every other week.

15
00:00:13.280 --> 00:00:13.874
It really did.

16
00:00:14.038 --> 00:00:17.624
And suddenly those big numbers everyone focused on earnings balance sheets.

17
00:00:17.843 --> 00:00:17.983
Yeah.

18
00:00:17.999 --> 00:00:19.421
They just didn't feel solid anymore,

19
00:00:19.436 --> 00:00:19.765
did they?

20
00:00:19.843 --> 00:00:21.061
People got really skeptical.

21
00:00:22.046 --> 00:00:23.530
There was even that quote floating around,

22
00:00:23.546 --> 00:00:23.827
I think,

23
00:00:23.874 --> 00:00:26.686
from an analyst saying the statement of cash flows.

24
00:00:27.180 --> 00:00:29.920
was maybe the only thing you could actually trust.

25
00:00:29.981 --> 00:00:30.260
Right.

26
00:00:30.780 --> 00:00:30.960
Which,

27
00:00:31.299 --> 00:00:31.542
you know,

28
00:00:31.663 --> 00:00:35.643
leads us straight to this really fundamental question we're tackling today.

29
00:00:35.745 --> 00:00:40.042
Does the market like actually punish companies where there's this consistent gap,

30
00:00:40.260 --> 00:00:40.479
you know,

31
00:00:40.502 --> 00:00:43.284
between the earnings they report and the cash that's actually moving?

32
00:00:43.424 --> 00:00:43.924
Exactly.

33
00:00:44.331 --> 00:00:45.596
It's one thing to feel suspicious,

34
00:00:45.659 --> 00:00:48.627
but does it does it actually hit the stock price long term?

35
00:00:48.643 --> 00:00:49.502
That's the core of it.

36
00:00:49.940 --> 00:00:52.112
And that is precisely what we are getting into today.

37
00:00:52.487 --> 00:00:55.909
We've got some some really interesting research focusing on.

38
00:00:56.564 --> 00:00:58.444
Something called accrual volatility.

39
00:00:58.645 --> 00:00:58.864
Okay,

40
00:00:58.944 --> 00:00:59.924
sounds a bit technical maybe.

41
00:01:00.325 --> 00:01:00.803
It does,

42
00:01:00.964 --> 00:01:02.464
but the idea behind it is,

43
00:01:02.764 --> 00:01:02.905
well,

44
00:01:03.245 --> 00:01:04.385
pretty straightforward actually.

45
00:01:04.803 --> 00:01:07.210
Accruals are basically the non-cash bits of accounting,

46
00:01:07.264 --> 00:01:07.444
right?

47
00:01:07.608 --> 00:01:07.827
Right,

48
00:01:07.866 --> 00:01:10.428
like sales you booked but haven't collected cash for yet,

49
00:01:10.569 --> 00:01:12.585
or expenses you owe but haven't paid,

50
00:01:13.225 --> 00:01:13.741
stuff like that.

51
00:01:14.007 --> 00:01:14.444
Exactly.

52
00:01:14.866 --> 00:01:18.803
So what these researchers did was measure how much those accruals bounce around,

53
00:01:19.069 --> 00:01:19.288
you know,

54
00:01:19.303 --> 00:01:19.991
quarter to quarter.

55
00:01:20.194 --> 00:01:25.194
As a way to see how consistently the reported profit lines up with the actual cash flow.

56
00:01:25.868 --> 00:01:26.329
Precisely.

57
00:01:26.829 --> 00:01:28.731
Because the thinking is over the long run,

58
00:01:29.052 --> 00:01:32.175
these accruals should kind of wash out.

59
00:01:32.214 --> 00:01:34.054
They should revert towards zero.

60
00:01:34.475 --> 00:01:34.600
OK.

61
00:01:34.975 --> 00:01:39.061
So if you see these really big persistent swings in accruals,

62
00:01:39.460 --> 00:01:45.304
it kind of flags a consistent difference between the profit a company claims and the cash it's really pulling in.

63
00:01:45.585 --> 00:01:47.616
So the key isn't just having accruals.

64
00:01:47.694 --> 00:01:49.819
It's how much they jump around the volatility.

65
00:01:50.069 --> 00:01:50.319
Yes.

66
00:01:50.788 --> 00:01:52.897
Why the focus on volatility specifically?

67
00:01:52.980 --> 00:01:53.120
Well,

68
00:01:53.140 --> 00:01:55.923
it's that consistency of the difference that seems to matter.

69
00:01:56.362 --> 00:02:00.647
A company might have high accruals one quarter for perfectly normal timing reasons,

70
00:02:00.667 --> 00:02:00.890
you know?

71
00:02:00.909 --> 00:02:01.187
Sure.

72
00:02:01.507 --> 00:02:03.007
A big sale at quarter end or something.

73
00:02:03.327 --> 00:02:03.874
Exactly.

74
00:02:04.069 --> 00:02:08.132
But if there's repeatedly a big gap between earnings and cash flow quarter after quarter,

75
00:02:08.593 --> 00:02:09.976
that signals something different.

76
00:02:10.069 --> 00:02:12.772
Accrual volatility captures that ongoing pattern.

77
00:02:12.960 --> 00:02:13.179
Okay.

78
00:02:13.366 --> 00:02:13.944
That makes sense.

79
00:02:15.085 --> 00:02:16.022
And here's where it gets,

80
00:02:16.460 --> 00:02:16.616
well,

81
00:02:17.288 --> 00:02:18.241
really fascinating.

82
00:02:18.929 --> 00:02:19.819
The main finding,

83
00:02:19.882 --> 00:02:21.554
the big takeaway from this research.

84
00:02:22.024 --> 00:02:30.032
is a really strong and long-lasting negative link between this accrual volatility and future stock returns.

85
00:02:30.189 --> 00:02:30.611
Negative.

86
00:02:30.696 --> 00:02:34.259
So higher volatility means lower returns later on.

87
00:02:34.415 --> 00:02:34.775
That's it.

88
00:02:34.876 --> 00:02:37.376
They found that if you hypothetically built a strategy,

89
00:02:37.978 --> 00:02:42.259
buying stocks with low accrual volatility and shorting stocks with high volatility,

90
00:02:43.009 --> 00:02:48.353
you could have potentially seen an average annual risk-adjusted return around 10%.

91
00:02:48.509 --> 00:02:48.837
Wow.

92
00:02:48.978 --> 00:02:49.384
10%

93
00:02:49.446 --> 00:02:49.993
risk-adjusted.

94
00:02:49.994 --> 00:02:50.196
That's...

95
00:02:50.540 --> 00:02:51.461
That's significant.

96
00:02:51.481 --> 00:02:51.981
It really is.

97
00:02:52.182 --> 00:02:52.302
Now,

98
00:02:52.382 --> 00:02:52.642
obviously,

99
00:02:52.663 --> 00:02:54.003
we're not giving investment advice here.

100
00:02:54.105 --> 00:02:54.964
Don't run out and do this.

101
00:02:55.023 --> 00:02:55.265
Right.

102
00:02:55.304 --> 00:02:56.085
Standard disclaimer.

103
00:02:56.366 --> 00:03:03.890
But it strongly suggests that maybe paying closer attention to how earnings and cash flow relate could be pretty valuable looking at companies.

104
00:03:04.257 --> 00:03:05.976
And connecting back,

105
00:03:06.374 --> 00:03:10.694
it really implies the market does seem to notice this disconnect over time.

106
00:03:11.335 --> 00:03:14.960
It penalizes firms where earnings and cash are consistently out of whack.

107
00:03:15.382 --> 00:03:16.101
And what's striking,

108
00:03:16.132 --> 00:03:16.460
like you said,

109
00:03:16.461 --> 00:03:18.710
this isn't just about the Enrons and Worldcoms,

110
00:03:18.711 --> 00:03:19.773
the big fraud cases.

111
00:03:19.972 --> 00:03:20.092
No,

112
00:03:20.432 --> 00:03:20.852
not at all.

113
00:03:20.973 --> 00:03:22.952
They looked at a huge range of companies,

114
00:03:23.153 --> 00:03:23.653
NYSE,

115
00:03:23.974 --> 00:03:24.653
NASDAQ,

116
00:03:24.794 --> 00:03:26.052
AMMX listed firms.

117
00:03:26.134 --> 00:03:26.294
Yeah,

118
00:03:26.352 --> 00:03:27.032
over 20 years,

119
00:03:27.095 --> 00:03:27.255
right?

120
00:03:27.256 --> 00:03:28.552
1988 to 2008.

121
00:03:28.575 --> 00:03:28.997
That's right.

122
00:03:29.091 --> 00:03:31.052
And they found this deviation was widespread.

123
00:03:31.411 --> 00:03:32.958
It wasn't just a few bad apples.

124
00:03:33.013 --> 00:03:34.075
And the numbers really show it.

125
00:03:34.372 --> 00:03:35.497
Across all those companies,

126
00:03:35.536 --> 00:03:36.528
the average connection,

127
00:03:36.700 --> 00:03:40.731
the correlation between quarterly earnings changes and quarterly cash flow changes was only,

128
00:03:40.841 --> 00:03:40.997
what,

129
00:03:40.998 --> 00:03:42.200
0.55?

130
00:03:42.231 --> 00:03:42.403
Yeah,

131
00:03:42.435 --> 00:03:42.903
about that.

132
00:03:43.075 --> 00:03:43.731
Kind of moderate.

133
00:03:43.778 --> 00:03:48.216
But then the average link between changes in quarterly accruals and cash flow changes was...

134
00:03:49.320 --> 00:03:50.100
strongly negative,

135
00:03:50.421 --> 00:03:51.540
minus 0.78.

136
00:03:51.581 --> 00:03:52.901
Minus 0.78,

137
00:03:53.100 --> 00:03:53.602
whoa.

138
00:03:53.603 --> 00:03:54.962
It really drives the point home,

139
00:03:54.981 --> 00:03:55.380
doesn't it?

140
00:03:55.622 --> 00:03:56.762
When cash flow changes,

141
00:03:57.239 --> 00:04:02.739
it's actually more likely to be offset by an opposite move in accruals than it is to just flow through to the earnings number.

142
00:04:02.786 --> 00:04:04.645
The accruals are acting like a buffer almost,

143
00:04:05.184 --> 00:04:06.091
smoothing things out.

144
00:04:06.419 --> 00:04:08.544
Maybe sometimes smoothing things out a bit too much.

145
00:04:09.262 --> 00:04:12.059
It's almost like that post-Enron skepticism,

146
00:04:12.091 --> 00:04:12.309
you know,

147
00:04:12.325 --> 00:04:13.700
questioning the earnings numbers.

148
00:04:14.169 --> 00:04:17.184
Maybe it was tapping into something real and broader.

149
00:04:17.640 --> 00:04:19.121
even outside of outright fraud.

150
00:04:19.262 --> 00:04:20.863
That's a really interesting way to put it.

151
00:04:20.902 --> 00:04:21.023
So,

152
00:04:21.104 --> 00:04:21.223
OK,

153
00:04:21.684 --> 00:04:23.848
what does this all mean for you listening in?

154
00:04:24.367 --> 00:04:24.488
Well,

155
00:04:24.489 --> 00:04:25.945
it suggests that earnings,

156
00:04:26.008 --> 00:04:27.609
the number that grabs all the headlines,

157
00:04:28.211 --> 00:04:29.992
might not always give you the full,

158
00:04:30.492 --> 00:04:34.094
unbiased story of a company's real financial pulse.

159
00:04:34.438 --> 00:04:36.484
Which brings us to the mission for this deep dive.

160
00:04:36.562 --> 00:04:37.703
We want to understand this,

161
00:04:38.016 --> 00:04:39.969
this accrual volatility anomaly.

162
00:04:40.188 --> 00:04:40.328
Yeah.

163
00:04:40.484 --> 00:04:45.281
Why does this consistent gap between earnings and cash seem to predict?

164
00:04:45.884 --> 00:04:47.387
lower stock returns in the future.

165
00:04:47.667 --> 00:04:49.370
And maybe more practically for you,

166
00:04:50.033 --> 00:04:51.833
how can you use this knowledge to,

167
00:04:52.255 --> 00:04:52.497
you know,

168
00:04:52.615 --> 00:04:56.521
be a smarter investor without getting lost in super complex financial weeds?

169
00:04:56.701 --> 00:04:56.904
Okay,

170
00:04:56.966 --> 00:04:58.724
so we have this really intriguing pattern.

171
00:04:59.431 --> 00:05:00.752
High accrual volatility,

172
00:05:01.173 --> 00:05:02.215
lower future returns.

173
00:05:02.695 --> 00:05:03.894
The big question is why?

174
00:05:04.676 --> 00:05:05.738
What's going on underneath?

175
00:05:05.855 --> 00:05:06.039
Right,

176
00:05:06.040 --> 00:05:06.758
what's the mechanism?

177
00:05:07.277 --> 00:05:09.824
The researchers explored a couple of main possibilities.

178
00:05:09.863 --> 00:05:13.348
The first one is something they call the earnings fixation hypothesis.

179
00:05:13.543 --> 00:05:14.262
Earnings fixation.

180
00:05:14.504 --> 00:05:14.683
Yeah,

181
00:05:14.769 --> 00:05:16.793
it connects right into investor psychology,

182
00:05:16.824 --> 00:05:17.933
which is always fascinating.

183
00:05:18.090 --> 00:05:18.371
How so?

184
00:05:18.824 --> 00:05:24.230
The idea is basically that investors tend to get really fixated on that reported earnings number.

185
00:05:24.871 --> 00:05:25.121
You know,

186
00:05:25.230 --> 00:05:25.840
the bottom line.

187
00:05:25.902 --> 00:05:26.058
Sure,

188
00:05:26.059 --> 00:05:26.871
it's the headline figure.

189
00:05:27.159 --> 00:05:27.600
Exactly.

190
00:05:27.800 --> 00:05:34.726
And maybe they don't fully grasp or perhaps appreciate the difference between the cash part of that number and the non-cash part,

191
00:05:34.746 --> 00:05:35.324
the accruals.

192
00:05:35.425 --> 00:05:35.550
OK.

193
00:05:36.167 --> 00:05:43.191
So for companies with high accrual volatility or maybe earnings are temporarily pumped up by aggressive accrual accounting,

194
00:05:43.722 --> 00:05:45.394
investors might get too optimistic.

195
00:05:45.472 --> 00:05:45.722
Right.

196
00:05:46.050 --> 00:05:51.082
They overvalue the stock because they're focused on that possibly inflated earnings number.

197
00:05:51.410 --> 00:05:52.191
And the flip side.

198
00:05:53.063 --> 00:05:55.466
For low volatility companies.

199
00:05:55.626 --> 00:05:56.187
Conversely,

200
00:05:56.245 --> 00:05:56.446
yeah.

201
00:05:56.526 --> 00:06:00.331
For companies where earnings trap cash flow more closely,

202
00:06:00.972 --> 00:06:05.534
low accrual volatility investors might actually underappreciate that stability and quality.

203
00:06:05.558 --> 00:06:07.058
They might undervalue those stocks.

204
00:06:07.214 --> 00:06:07.339
OK,

205
00:06:07.534 --> 00:06:13.690
so it's like a potential mispricing based on focusing too much on the reported earnings and not enough on the,

206
00:06:14.128 --> 00:06:14.425
let's say,

207
00:06:14.550 --> 00:06:17.050
the quality or cash backing of those earnings.

208
00:06:17.362 --> 00:06:17.925
Pretty much.

209
00:06:18.315 --> 00:06:22.659
The other main explanation they looked into revolves around information uncertainty.

210
00:06:23.147 --> 00:06:24.549
Information uncertainty.

211
00:06:25.249 --> 00:06:28.952
How does accrual volatility create uncertainty?

212
00:06:29.393 --> 00:06:29.534
Well,

213
00:06:29.635 --> 00:06:30.214
think about it.

214
00:06:30.432 --> 00:06:35.557
If a company's reported profits consistently look quite different from its actual cash coming in,

215
00:06:36.120 --> 00:06:39.331
it kind of suggests that management has more wiggle room,

216
00:06:39.643 --> 00:06:44.581
more discretion in how they paint the financial picture through accounting choices.

217
00:06:45.049 --> 00:06:45.253
Ah,

218
00:06:45.596 --> 00:06:45.815
okay.

219
00:06:46.049 --> 00:06:48.768
More choices mean maybe less clarity for outsiders.

220
00:06:48.878 --> 00:06:49.440
Exactly.

221
00:06:49.659 --> 00:06:50.549
That flexibility.

222
00:06:51.087 --> 00:06:54.991
While sometimes perfectly legitimate and necessary for accounting rules,

223
00:06:55.292 --> 00:06:57.932
can also make it tougher for investors to get a really clear,

224
00:06:58.014 --> 00:07:00.557
objective view of the underlying business reality.

225
00:07:00.714 --> 00:07:01.635
Leading to uncertainty.

226
00:07:01.799 --> 00:07:02.034
Right.

227
00:07:02.339 --> 00:07:03.057
And interestingly,

228
00:07:03.299 --> 00:07:07.628
the research did find that accrual volatility tends to be higher in companies where,

229
00:07:07.878 --> 00:07:08.378
for example,

230
00:07:08.440 --> 00:07:11.924
financial analysts disagree more about future earnings forecasts.

231
00:07:12.034 --> 00:07:12.581
That makes sense.

232
00:07:12.909 --> 00:07:14.737
More disagreement means more uncertainty.

233
00:07:14.924 --> 00:07:19.049
And also where the stock price itself is just generally more volatile day to day.

234
00:07:19.112 --> 00:07:20.049
It all kind of fits together.

235
00:07:20.378 --> 00:07:20.737
It does.

236
00:07:21.679 --> 00:07:23.841
If it's harder to pin down the true performance,

237
00:07:24.282 --> 00:07:26.782
investors likely feel less sure-footed.

238
00:07:27.446 --> 00:07:33.251
They might demand a higher return or essentially pay a lower price to compensate for that extra uncertainty or risk.

239
00:07:33.407 --> 00:07:33.970
Precisely.

240
00:07:34.095 --> 00:07:34.251
Now,

241
00:07:34.454 --> 00:07:36.798
we've talked about this pattern looking at groups of stocks,

242
00:07:36.852 --> 00:07:37.532
portfolios,

243
00:07:37.954 --> 00:07:40.876
but they also drilled down to see if it holds for individual companies.

244
00:07:41.032 --> 00:07:41.345
They did,

245
00:07:41.376 --> 00:07:41.517
yeah,

246
00:07:41.813 --> 00:07:42.095
using,

247
00:07:42.454 --> 00:07:42.860
what was it,

248
00:07:42.923 --> 00:07:44.938
Fama-Macbeth regressions.

249
00:07:44.954 --> 00:07:45.642
That's the one.

250
00:07:46.115 --> 00:07:47.056
It sounds complex,

251
00:07:47.177 --> 00:07:55.284
but think of it as a statistical way to test the impact of accrual volatility while controlling for other things we already know affect stock prices.

252
00:07:55.362 --> 00:07:56.323
Like company size,

253
00:07:56.487 --> 00:07:57.565
value characteristics,

254
00:07:57.847 --> 00:07:58.487
book to market,

255
00:07:59.026 --> 00:07:59.526
momentum,

256
00:08:00.690 --> 00:08:01.589
the usual suspects.

257
00:08:01.690 --> 00:08:02.292
Exactly.

258
00:08:02.714 --> 00:08:08.433
It's like trying to isolate the effect of just the accrual volatility stripping out those other influences.

259
00:08:08.511 --> 00:08:09.261
And what did they find?

260
00:08:09.886 --> 00:08:11.839
Even after accounting for all those other factors,

261
00:08:11.840 --> 00:08:12.948
they still found that yes,

262
00:08:13.319 --> 00:08:17.900
Higher accrual volatility was statistically linked to lower future returns for individual stocks.

263
00:08:18.322 --> 00:08:21.322
It wasn't just explained away by size or value effects.

264
00:08:21.720 --> 00:08:21.837
OK,

265
00:08:22.001 --> 00:08:22.743
that's pretty compelling.

266
00:08:22.923 --> 00:08:28.118
And they were also careful to check if this was just like a different version of the standard accrual anomaly,

267
00:08:28.165 --> 00:08:28.525
weren't they?

268
00:08:28.603 --> 00:08:28.720
Oh,

269
00:08:28.822 --> 00:08:29.259
absolutely.

270
00:08:29.322 --> 00:08:29.915
Critical point.

271
00:08:29.947 --> 00:08:35.462
The standard accrual anomaly is about the level of accruals companies with high total accruals tend to underperform.

272
00:08:35.603 --> 00:08:35.868
Right.

273
00:08:35.884 --> 00:08:36.400
So you might think,

274
00:08:36.431 --> 00:08:36.572
well,

275
00:08:36.587 --> 00:08:40.165
maybe companies with high accruals just naturally have more volatile accruals,

276
00:08:40.228 --> 00:08:40.400
too.

277
00:08:40.712 --> 00:08:42.400
That's the potential overlap you need to check.

278
00:08:42.959 --> 00:08:44.560
But they showed it's distinct.

279
00:08:44.800 --> 00:08:50.086
They did things like sorting stocks based on both the level of accruals and the volatility.

280
00:08:50.507 --> 00:08:50.765
Yeah.

281
00:08:50.929 --> 00:08:58.336
And they found the negative link between volatility and returns held up even within groups of stocks that had similar levels of accruals.

282
00:08:58.937 --> 00:09:02.500
High volatility hurt returns whether overall accruals were high or low.

283
00:09:02.859 --> 00:09:03.140
I see.

284
00:09:03.203 --> 00:09:05.156
So it's not just the same thing measured differently.

285
00:09:05.297 --> 00:09:05.422
No.

286
00:09:06.078 --> 00:09:08.406
They also checked the correlation between the two measures,

287
00:09:08.625 --> 00:09:10.312
high level versus high volatility,

288
00:09:10.750 --> 00:09:12.406
and found it wasn't super high.

289
00:09:12.891 --> 00:09:15.672
They seem to be capturing different aspects of the financial reporting.

290
00:09:15.913 --> 00:09:17.651
Did they look at different types of accruals,

291
00:09:17.713 --> 00:09:17.893
too,

292
00:09:18.194 --> 00:09:21.530
like operating accruals versus maybe more discretionary ones?

293
00:09:21.553 --> 00:09:21.670
Yeah,

294
00:09:22.014 --> 00:09:22.413
they did.

295
00:09:22.491 --> 00:09:25.694
They broke it down to see if the effect was stronger for certain types,

296
00:09:25.874 --> 00:09:28.967
maybe ones where management has more judgment involved.

297
00:09:29.030 --> 00:09:30.295
And the results were consistent.

298
00:09:30.389 --> 00:09:30.999
Pretty much.

299
00:09:31.108 --> 00:09:31.295
Yeah.

300
00:09:31.702 --> 00:09:33.811
Across different ways of measuring accruals,

301
00:09:33.842 --> 00:09:35.624
the volatility effect kept showing up,

302
00:09:35.999 --> 00:09:40.999
which really reinforces the idea that it's acting as this broader signal about the potential disconnect.

303
00:09:41.079 --> 00:09:43.240
between earnings and cash reality.

304
00:09:43.699 --> 00:09:46.582
Regardless of the specific accounting lever being pulled,

305
00:09:46.699 --> 00:09:47.000
maybe.

306
00:09:47.262 --> 00:09:48.801
That seems to be the implication.

307
00:09:48.902 --> 00:09:49.101
Okay,

308
00:09:49.141 --> 00:09:49.281
now,

309
00:09:49.617 --> 00:09:50.461
thinking practically,

310
00:09:50.820 --> 00:09:53.180
any strategy that involves short selling,

311
00:09:54.266 --> 00:09:54.445
well,

312
00:09:54.805 --> 00:09:56.984
it raises questions about real-world feasibility,

313
00:09:57.039 --> 00:09:57.398
doesn't it?

314
00:09:57.461 --> 00:09:57.930
Definitely.

315
00:09:58.195 --> 00:10:00.398
These limits to arbitrage ideas...

316
00:10:01.073 --> 00:10:02.174
Can you actually do this?

317
00:10:02.494 --> 00:10:06.037
Is it too costly or too difficult to short the high volatility stocks?

318
00:10:06.279 --> 00:10:06.498
Right.

319
00:10:06.599 --> 00:10:07.760
Some stocks are hard to borrow,

320
00:10:07.959 --> 00:10:08.842
expensive to short,

321
00:10:09.322 --> 00:10:13.025
or maybe just too risky for many investors to touch on the short side.

322
00:10:13.885 --> 00:10:15.564
So the researchers looked at this directly.

323
00:10:15.744 --> 00:10:19.205
They considered things like how volatile a stock's individual return is,

324
00:10:19.799 --> 00:10:22.111
how easy it is to trade its liquidity.

325
00:10:22.502 --> 00:10:25.017
Did the effect disappear for harder to trade stocks?

326
00:10:25.471 --> 00:10:25.596
No.

327
00:10:25.905 --> 00:10:31.471
they found the accrual volatility effect was still significant even after controlling for those factors.

328
00:10:32.112 --> 00:10:33.213
What about trading costs?

329
00:10:33.791 --> 00:10:35.471
Bid-ask spreads and stuff.

330
00:10:35.533 --> 00:10:36.393
They looked at that too.

331
00:10:37.135 --> 00:10:39.057
Using measures of transaction costs,

332
00:10:39.096 --> 00:10:40.580
they concluded that yes,

333
00:10:40.760 --> 00:10:43.041
while costs eat into profits,

334
00:10:43.447 --> 00:10:46.041
the anomaly still seemed potentially profitable,

335
00:10:46.526 --> 00:10:49.026
especially if you held the positions for longer periods,

336
00:10:49.072 --> 00:10:50.807
not just trading in and out constantly.

337
00:10:51.166 --> 00:10:51.322
Okay.

338
00:10:51.445 --> 00:10:56.229
And they even checked if the effect held up specifically for stocks that have options traded on them,

339
00:10:56.288 --> 00:10:58.651
since those are generally easier and cheaper to short.

340
00:10:58.913 --> 00:10:59.475
And it did.

341
00:10:59.655 --> 00:11:00.014
It did.

342
00:11:00.053 --> 00:11:01.897
The negative relationship was still there.

343
00:11:02.014 --> 00:11:07.749
So it doesn't look like it's just some theoretical finding that vanishes once you factor in real world fictions.

344
00:11:08.046 --> 00:11:09.139
So the evidence stacks up.

345
00:11:09.561 --> 00:11:10.749
It seems like this isn't just,

346
00:11:10.764 --> 00:11:11.030
you know,

347
00:11:11.031 --> 00:11:12.186
an academic curiosity.

348
00:11:12.671 --> 00:11:15.124
It looks like it has real implications for market pricing.

349
00:11:15.514 --> 00:11:17.217
That's definitely the strong suggestion.

350
00:11:17.264 --> 00:11:20.077
And it gets even maybe more profound than that.

351
00:11:20.237 --> 00:11:20.597
How so?

352
00:11:20.878 --> 00:11:23.120
They explored this really intriguing idea.

353
00:11:24.020 --> 00:11:28.665
Could accrual volatility actually be a fundamental pricing factor in the market?

354
00:11:28.946 --> 00:11:29.907
A pricing factor,

355
00:11:30.610 --> 00:11:32.048
like alongside size,

356
00:11:32.751 --> 00:11:32.946
SMB,

357
00:11:33.813 --> 00:11:34.571
and value,

358
00:11:34.813 --> 00:11:35.094
HML,

359
00:11:35.149 --> 00:11:36.188
in the Fama French models?

360
00:11:36.594 --> 00:11:37.251
Kind of like that,

361
00:11:37.313 --> 00:11:37.469
yeah.

362
00:11:37.813 --> 00:11:44.204
Is it capturing some kind of systematic risk or characteristic that investors collectively demand compensation for?

363
00:11:45.188 --> 00:11:45.376
Okay.

364
00:11:45.594 --> 00:11:46.282
How would you test that?

365
00:11:47.061 --> 00:11:52.226
They created a sort of mimic portfolio designed to capture the returns you'd get from going long,

366
00:11:52.487 --> 00:11:54.167
low volatility stocks and short,

367
00:11:54.249 --> 00:11:55.405
high volatility stocks.

368
00:11:55.429 --> 00:11:55.569
Right,

369
00:11:55.648 --> 00:11:56.288
that 10%

370
00:11:56.327 --> 00:11:57.913
potential return strategy we mentioned.

371
00:11:57.929 --> 00:11:58.491
Exactly.

372
00:11:59.015 --> 00:12:05.124
And then they checked if the returns of that hypothetical portfolio could help explain the returns of other well-known strategies,

373
00:12:05.577 --> 00:12:08.780
like portfolios sorted by size and book-to-market value.

374
00:12:08.921 --> 00:12:11.390
Even after accounting for the standard factors themselves.

375
00:12:11.483 --> 00:12:11.765
Yes.

376
00:12:12.108 --> 00:12:14.421
And they found that this accrual volatility factor

377
00:12:14.601 --> 00:12:18.967
did seem to have some explanatory power for the returns of those other portfolios.

378
00:12:19.103 --> 00:12:19.346
Wow.

379
00:12:19.506 --> 00:12:22.350
So it's not just a potential source of alpha like outperformance.

380
00:12:22.490 --> 00:12:26.150
It might actually be part of the underlying structure of market returns itself,

381
00:12:26.775 --> 00:12:30.682
capturing some kind of risk or characteristic that the market prices across the board.

382
00:12:30.791 --> 00:12:32.603
That is that's a pretty deep thought.

383
00:12:32.838 --> 00:12:33.494
It really is.

384
00:12:33.994 --> 00:12:42.228
And it just highlights how these seemingly subtle details in accounting reports can potentially have quite significant ripples through the market.

385
00:12:42.361 --> 00:12:42.481
OK,

386
00:12:42.601 --> 00:12:44.363
let's try and pull this all together then for you,

387
00:12:44.664 --> 00:12:45.145
our listener.

388
00:12:45.324 --> 00:12:45.565
Right.

389
00:12:45.703 --> 00:12:46.586
The main takeaway.

390
00:12:46.727 --> 00:12:48.926
The key point from this deep dive seems clear.

391
00:12:49.809 --> 00:12:51.410
Higher accrual volatility,

392
00:12:52.067 --> 00:12:53.012
meaning a bigger,

393
00:12:53.153 --> 00:12:57.613
more consistent gap between reported earnings and actual cash flow,

394
00:12:58.238 --> 00:13:01.067
tends to be linked with lower future stock returns.

395
00:13:01.395 --> 00:13:03.645
And we touched on the likely reasons why.

396
00:13:03.770 --> 00:13:07.879
Maybe it's investors being too focused on earnings and mispricing the accrual part.

397
00:13:07.926 --> 00:13:09.473
That's the earnings fixation idea.

398
00:13:09.757 --> 00:13:12.861
Or maybe it's that high volatility acts like a warning sign,

399
00:13:13.341 --> 00:13:16.462
signaling more uncertainty about the company's real financial picture,

400
00:13:17.064 --> 00:13:18.568
the information uncertainty effect.

401
00:13:18.607 --> 00:13:20.130
And this isn't some fleeting thing.

402
00:13:20.208 --> 00:13:24.114
The research suggests this accrual volatility anomaly is pretty robust.

403
00:13:24.271 --> 00:13:24.388
Yeah,

404
00:13:24.450 --> 00:13:26.239
it holds up against common risk factors.

405
00:13:26.333 --> 00:13:28.583
It's distinct from the standard accrual anomaly.

406
00:13:29.161 --> 00:13:32.818
It seems to survive transaction costs and short selling limits,

407
00:13:33.239 --> 00:13:34.021
at least over time.

408
00:13:34.318 --> 00:13:35.114
And it persists,

409
00:13:35.239 --> 00:13:35.458
too.

410
00:13:36.005 --> 00:13:37.364
Not just over a few months,

411
00:13:37.411 --> 00:13:37.521
but...

412
00:13:37.597 --> 00:13:41.220
but potentially for years they found effects looking out up to five years.

413
00:13:41.742 --> 00:13:44.722
Plus that really intriguing possibility we just discussed,

414
00:13:45.343 --> 00:13:48.589
that accrual volatility might even be a broader pricing factor,

415
00:13:49.050 --> 00:13:52.394
reflecting something fundamental about how the market values companies.

416
00:13:52.714 --> 00:13:52.894
Yeah,

417
00:13:53.089 --> 00:13:54.488
shaping returns overall.

418
00:13:55.128 --> 00:13:56.691
So to leave you with something to chew on,

419
00:13:57.550 --> 00:13:59.519
now that we've gone deep into accrual volatility,

420
00:14:00.425 --> 00:14:01.097
what other...

421
00:14:01.453 --> 00:14:05.697
Maybe less obvious accounting metrics might hold hidden clues about where a company is headed.

422
00:14:06.197 --> 00:14:07.779
As someone who's now equipped with this insight,

423
00:14:08.759 --> 00:14:11.345
how might you start looking at financial statements differently?

424
00:14:11.384 --> 00:14:11.540
Yeah,

425
00:14:11.642 --> 00:14:13.845
looking beyond just the big headline numbers,

426
00:14:13.884 --> 00:14:16.150
trying to understand more about the story behind those numbers.

427
00:14:16.447 --> 00:14:18.306
What's really going on under the surface?

428
00:14:18.525 --> 00:14:20.837
It definitely encourages a more critical,

429
00:14:20.931 --> 00:14:23.759
maybe more insightful way of looking at financial data.