WEBVTT

1
00:00:00.040 --> 00:00:00.200
Hello,

2
00:00:00.340 --> 00:00:02.502
welcome back to Papers with Backtest podcast.

3
00:00:03.023 --> 00:00:03.242
Today,

4
00:00:03.344 --> 00:00:05.844
we dive into another Algo trading research paper.

5
00:00:06.047 --> 00:00:06.484
Great.

6
00:00:06.805 --> 00:00:07.527
What have we got today?

7
00:00:07.789 --> 00:00:10.367
The study we're unpacking is called Active Collar Strategy,

8
00:00:10.867 --> 00:00:15.195
and it's actually an update to a previous look at collar strategies applied to the QQQ

9
00:00:15.594 --> 00:00:16.438
ETF. Ah,

10
00:00:16.703 --> 00:00:16.812
OK.

11
00:00:17.078 --> 00:00:18.062
Collars on the QQQ.

12
00:00:18.250 --> 00:00:19.172
I remember the original.

13
00:00:19.203 --> 00:00:20.344
So this one takes it further.

14
00:00:20.641 --> 00:00:21.078
It does.

15
00:00:21.625 --> 00:00:28.469
This updated research extends their analysis from March 1999 all the way through September 2010.

16
00:00:28.840 --> 00:00:31.003
even including some out-of-sample data up to that point.

17
00:00:31.182 --> 00:00:32.845
That's a really interesting timeframe.

18
00:00:32.864 --> 00:00:35.028
You've got the dot-com bubble bursting in there,

19
00:00:35.227 --> 00:00:36.188
a decent bull run,

20
00:00:36.227 --> 00:00:38.630
and then the 2008 credit crisis.

21
00:00:38.747 --> 00:00:39.270
Exactly.

22
00:00:39.552 --> 00:00:40.575
And as the authors point out,

23
00:00:40.576 --> 00:00:41.770
that 08 crisis really,

24
00:00:41.997 --> 00:00:42.231
you know,

25
00:00:42.395 --> 00:00:45.122
renewed interest in protective strategies like these equity callers.

26
00:00:45.950 --> 00:00:48.013
People were definitely looking for ways to manage downside.

27
00:00:48.106 --> 00:00:48.575
Makes sense.

28
00:00:48.825 --> 00:00:51.466
So maybe we should quickly recap what a caller actually is.

29
00:00:51.622 --> 00:00:52.075
Good idea.

30
00:00:52.309 --> 00:00:54.138
So imagine you hold shares of the QQQ.

31
00:00:54.966 --> 00:00:57.122
A basic caller involves buying a put option.

32
00:00:58.028 --> 00:00:58.908
Think of it like insurance.

33
00:00:59.528 --> 00:01:02.171
It gives you the right to sell at a certain price if things go south.

34
00:01:02.190 --> 00:01:03.448
It protects your downside.

35
00:01:03.729 --> 00:01:03.948
Right.

36
00:01:04.350 --> 00:01:05.108
And at the same time,

37
00:01:05.350 --> 00:01:06.393
you sell a call option.

38
00:01:06.932 --> 00:01:10.854
This gives someone else the right to buy your shares if the price goes up to a certain level.

39
00:01:11.089 --> 00:01:12.495
Which generates some income,

40
00:01:12.854 --> 00:01:14.198
helps pay for the put maybe,

41
00:01:14.339 --> 00:01:16.667
but it also caps your potential gains.

42
00:01:17.057 --> 00:01:17.651
Precisely.

43
00:01:18.026 --> 00:01:19.620
Your returns get kind of bracketed,

44
00:01:19.839 --> 00:01:22.354
limited on both the upside and the downside.

45
00:01:22.448 --> 00:01:24.854
And this paper looks at two main flavors of this,

46
00:01:25.448 --> 00:01:26.620
passive and active.

47
00:01:26.952 --> 00:01:27.092
Yeah.

48
00:01:27.132 --> 00:01:31.718
So the passive approach is where you just set fixed rules for picking the options,

49
00:01:31.777 --> 00:01:32.937
you know,

50
00:01:33.320 --> 00:01:34.199
how far out of the money,

51
00:01:34.742 --> 00:01:35.484
when they expire,

52
00:01:35.937 --> 00:01:36.781
and you just stick to them.

53
00:01:36.906 --> 00:01:37.664
Set it and forget it,

54
00:01:37.679 --> 00:01:38.085
basically.

55
00:01:38.101 --> 00:01:38.484
Pretty much.

56
00:01:38.726 --> 00:01:41.749
And then the active strategy is where those rules actually change.

57
00:01:41.929 --> 00:01:44.671
They adapt based on market conditions.

58
00:01:44.890 --> 00:01:45.093
Okay.

59
00:01:45.109 --> 00:01:46.171
And that's where it gets interesting.

60
00:01:46.249 --> 00:01:47.468
What conditions are they looking at?

61
00:01:47.796 --> 00:01:48.656
This is the core of it.

62
00:01:49.046 --> 00:01:51.640
They tested using signals like momentum,

63
00:01:52.312 --> 00:01:53.203
market volatility,

64
00:01:53.265 --> 00:01:53.843
think the VIX.

65
00:01:54.444 --> 00:01:56.925
and even some macroeconomic data like unemployment trends.

66
00:01:57.104 --> 00:01:57.385
Ah,

67
00:01:57.665 --> 00:02:02.466
so trying to be smarter about when to tighten the protection or when to allow for more upside?

68
00:02:02.848 --> 00:02:03.262
Exactly.

69
00:02:03.606 --> 00:02:05.106
The big question for you listening is,

70
00:02:05.966 --> 00:02:08.411
does this active approach actually work better?

71
00:02:08.489 --> 00:02:13.067
Does adapting the collar based on these signals improve performance compared to the simple passive rules?

72
00:02:13.208 --> 00:02:14.692
Or it's just holding the QQQ.

73
00:02:15.286 --> 00:02:16.677
That's what the back tests explore.

74
00:02:17.145 --> 00:02:17.395
All right,

75
00:02:17.396 --> 00:02:18.989
let's get into the passive ones first then.

76
00:02:19.223 --> 00:02:19.708
As you said,

77
00:02:19.817 --> 00:02:21.364
fixed rules for option selection,

78
00:02:21.427 --> 00:02:21.848
moneyness,

79
00:02:21.895 --> 00:02:22.630
and expiration.

80
00:02:22.817 --> 00:02:23.036
Right.

81
00:02:23.448 --> 00:02:24.709
They looked at a few combinations,

82
00:02:25.110 --> 00:02:28.115
different starting points for how far out of the money the puts and calls were,

83
00:02:28.493 --> 00:02:31.677
ranging from ATM at the money up to 5%

84
00:02:31.794 --> 00:02:32.201
OTM.

85
00:02:32.497 --> 00:02:34.161
And different lengths for the puts.

86
00:02:34.279 --> 00:02:34.396
Yeah,

87
00:02:34.419 --> 00:02:35.161
they tested one,

88
00:02:35.263 --> 00:02:35.583
three,

89
00:02:35.919 --> 00:02:36.841
and six-month puts,

90
00:02:37.341 --> 00:02:39.224
but always paired them with a one-month call.

91
00:02:39.966 --> 00:02:42.044
Keeps the upside cap resetting more frequently,

92
00:02:42.060 --> 00:02:42.544
I suppose.

93
00:02:42.669 --> 00:02:43.060
Makes sense.

94
00:02:43.091 --> 00:02:46.232
And they focused on one particular setup as a kind of benchmark?

95
00:02:46.388 --> 00:02:46.826
They did.

96
00:02:47.326 --> 00:02:49.951
A lot of the comparison revolves around a passive strategy,

97
00:02:50.372 --> 00:02:52.919
using a one-month call and a longer six-month put.

98
00:02:53.176 --> 00:02:54.177
With the put starting 2%

99
00:02:54.458 --> 00:02:54.897
out of the money.

100
00:02:55.057 --> 00:02:55.278
Okay.

101
00:02:55.458 --> 00:02:55.659
2%

102
00:02:55.919 --> 00:02:56.518
OTM put,

103
00:02:56.659 --> 00:02:57.479
six-month expiry,

104
00:02:57.620 --> 00:02:58.139
one-month call.

105
00:02:58.300 --> 00:02:58.663
Got it.

106
00:02:58.983 --> 00:03:02.647
So how did that specific passive caller do over the whole period,

107
00:03:03.163 --> 00:03:05.288
April 99 to September 2010,

108
00:03:05.772 --> 00:03:07.327
compared to just holding QQQ?

109
00:03:07.772 --> 00:03:07.889
Well,

110
00:03:07.890 --> 00:03:09.014
the difference was pretty stark,

111
00:03:09.108 --> 00:03:09.413
actually.

112
00:03:10.132 --> 00:03:11.210
If you just held QQQ,

113
00:03:11.397 --> 00:03:12.350
you basically broke even,

114
00:03:12.397 --> 00:03:14.991
maybe a tiny loss of 0.3%

115
00:03:15.163 --> 00:03:15.725
annualized,

116
00:03:16.054 --> 00:03:17.163
but with a lot of volatility,

117
00:03:17.194 --> 00:03:18.257
almost 30%.

118
00:03:18.288 --> 00:03:18.397
Oof.

119
00:03:18.819 --> 00:03:19.007
Yeah,

120
00:03:19.179 --> 00:03:21.522
that was a rough decade overall for QQQ.

121
00:03:21.780 --> 00:03:22.140
But the 2%

122
00:03:22.581 --> 00:03:23.781
OTM passive collar,

123
00:03:24.102 --> 00:03:25.623
it showed a 9.6%

124
00:03:25.664 --> 00:03:26.463
annualized gain.

125
00:03:27.084 --> 00:03:27.725
And crucially,

126
00:03:27.807 --> 00:03:29.147
the volatility was way lower,

127
00:03:29.229 --> 00:03:29.569
only about

128
00:03:30.010 --> 00:03:31.088
10.7%. Wow,

129
00:03:31.089 --> 00:03:31.291
okay.

130
00:03:31.651 --> 00:03:33.471
Better returns and much smoother ride.

131
00:03:34.190 --> 00:03:35.557
What about the worst case scenario,

132
00:03:35.651 --> 00:03:36.791
the maximum drawdown?

133
00:03:36.815 --> 00:03:38.119
That's where the protection really shows up.

134
00:03:38.182 --> 00:03:42.057
The max drawdown for QQQ was a brutal negative 81.1%.

135
00:03:42.058 --> 00:03:43.026
It's devastating.

136
00:03:43.088 --> 00:03:43.244
Yeah,

137
00:03:43.245 --> 00:03:43.776
I remember that.

138
00:03:44.026 --> 00:03:44.916
For the passive collar,

139
00:03:44.963 --> 00:03:45.182
though,

140
00:03:45.260 --> 00:03:47.901
it was only negative 17.6%.

141
00:03:47.932 --> 00:03:48.619
A huge,

142
00:03:48.713 --> 00:03:50.244
huge difference in capital preservation.

143
00:03:50.516 --> 00:03:52.917
That alone is a pretty compelling argument for the collar,

144
00:03:52.996 --> 00:03:54.236
especially during that period.

145
00:03:54.435 --> 00:03:54.818
Definitely.

146
00:03:55.017 --> 00:03:55.177
Yeah.

147
00:03:55.337 --> 00:03:56.556
Let's break down those periods a bit.

148
00:03:56.857 --> 00:03:58.376
What about the tech bubble bursting,

149
00:03:58.498 --> 00:03:58.662
say,

150
00:03:58.677 --> 00:04:00.201
99 to late 02?

151
00:04:00.396 --> 00:04:00.615
Okay,

152
00:04:00.662 --> 00:04:00.779
yeah,

153
00:04:00.841 --> 00:04:02.255
that was prime time for QQQP,

154
00:04:02.318 --> 00:04:04.380
and it lost about 23.3%

155
00:04:04.381 --> 00:04:05.419
annualized during that stretch,

156
00:04:05.443 --> 00:04:07.177
with volatility over 40%,

157
00:04:07.178 --> 00:04:08.318
just wild swings.

158
00:04:08.365 --> 00:04:09.052
And the passive collar?

159
00:04:09.193 --> 00:04:10.037
This is incredible.

160
00:04:10.083 --> 00:04:10.427
The 2%

161
00:04:10.974 --> 00:04:13.099
OTM passive collar actually made money.

162
00:04:13.615 --> 00:04:16.740
It had an annualized return of 21.6%,

163
00:04:16.741 --> 00:04:17.287
positive,

164
00:04:17.615 --> 00:04:19.490
with volatility around 13,

165
00:04:19.521 --> 00:04:19.818
14%.

166
00:04:20.024 --> 00:04:25.769
So it turned a massive loss into a significant gain just by capping the upside and protecting the downside.

167
00:04:25.949 --> 00:04:26.371
Effectively,

168
00:04:26.429 --> 00:04:26.671
yes.

169
00:04:26.910 --> 00:04:28.929
And that drawdown number during the tech bust,

170
00:04:29.015 --> 00:04:30.836
Q2Q hit that netable 81%.

171
00:04:31.312 --> 00:04:36.125
But the caller's worst drawdown in that period was only netable 7.5%.

172
00:04:36.126 --> 00:04:37.203
Minus 7.5%.

173
00:04:37.406 --> 00:04:38.468
That's amazing protection.

174
00:04:38.500 --> 00:04:41.078
It really highlights the buffering effect during a crash.

175
00:04:41.312 --> 00:04:41.531
Okay.

176
00:04:41.968 --> 00:04:43.203
But callers aren't always great,

177
00:04:43.453 --> 00:04:43.625
right?

178
00:04:44.375 --> 00:04:46.531
The paper mentioned an unfavorable period,

179
00:04:46.609 --> 00:04:48.515
October 2002 to September 2007.

180
00:04:48.828 --> 00:04:49.390
What happened then?

181
00:04:49.696 --> 00:04:49.917
Right.

182
00:04:49.937 --> 00:04:51.420
That was mostly a strong,

183
00:04:51.701 --> 00:04:53.385
steady bull market recovery phase.

184
00:04:53.465 --> 00:04:54.830
QQQ did really well,

185
00:04:54.850 --> 00:04:56.131
about 20.4%

186
00:04:56.132 --> 00:04:56.834
annualized return.

187
00:04:56.955 --> 00:04:57.455
And the caller.

188
00:04:57.801 --> 00:04:59.984
It lagged because the market was mostly going up.

189
00:05:00.023 --> 00:05:02.243
That call option you sold kept capping your gains.

190
00:05:02.646 --> 00:05:03.388
The passive 2%

191
00:05:03.868 --> 00:05:06.493
OTM caller only made about 5.2%

192
00:05:06.494 --> 00:05:07.150
annualized then.

193
00:05:07.314 --> 00:05:08.314
So lower volatility,

194
00:05:08.368 --> 00:05:09.876
but you missed out on a lot of the rally.

195
00:05:10.150 --> 00:05:10.712
Exactly.

196
00:05:10.915 --> 00:05:11.697
That's the tradeoff.

197
00:05:11.837 --> 00:05:13.056
You pay for that insurance,

198
00:05:13.134 --> 00:05:14.322
and when there are no accidents,

199
00:05:14.337 --> 00:05:15.822
you feel the cost of the premium,

200
00:05:15.837 --> 00:05:16.306
so to speak.

201
00:05:16.556 --> 00:05:16.993
Makes sense.

202
00:05:17.743 --> 00:05:18.697
Then came the next storm,

203
00:05:19.228 --> 00:05:20.165
the credit crisis,

204
00:05:20.650 --> 00:05:22.368
late 07 to September 2010.

205
00:05:23.009 --> 00:05:23.853
How did it play out there?

206
00:05:24.290 --> 00:05:25.509
Similar story to the tech...

207
00:05:25.565 --> 00:05:25.885
bust,

208
00:05:26.066 --> 00:05:27.407
just maybe less extreme,

209
00:05:27.908 --> 00:05:29.869
QQQ lost about 1%

210
00:05:30.291 --> 00:05:30.912
annualized.

211
00:05:31.091 --> 00:05:32.072
Not catastrophic,

212
00:05:32.111 --> 00:05:33.052
but still negative.

213
00:05:33.173 --> 00:05:33.650
And the collar.

214
00:05:33.814 --> 00:05:35.431
The passive collar was positive again.

215
00:05:35.455 --> 00:05:37.017
It returned 3.8%

216
00:05:37.056 --> 00:05:40.041
annualized with less than half the volatility of QQQ.

217
00:05:40.220 --> 00:05:40.580
So again,

218
00:05:40.581 --> 00:05:42.494
it provided that cushion during market turmoil.

219
00:05:42.634 --> 00:05:42.837
Okay.

220
00:05:42.869 --> 00:05:44.400
So passive works for protection,

221
00:05:44.806 --> 00:05:47.259
maybe enhances returns overall in volatile times,

222
00:05:47.447 --> 00:05:49.041
but can lag in strong bull markets.

223
00:05:49.728 --> 00:05:51.087
Now the active strategies,

224
00:05:51.462 --> 00:05:52.478
they tried to improve on this.

225
00:05:52.634 --> 00:05:53.212
That was the goal.

226
00:05:53.409 --> 00:05:54.590
Instead of fixed rules,

227
00:05:54.630 --> 00:05:59.634
the active callers adjusted the option choices based on three main types of signals,

228
00:05:59.814 --> 00:06:00.294
momentum,

229
00:06:00.357 --> 00:06:01.779
volatility and macro factors.

230
00:06:01.794 --> 00:06:02.740
Trying to be more dynamic.

231
00:06:02.779 --> 00:06:03.396
Let's break those down.

232
00:06:03.443 --> 00:06:04.724
How did the momentum signal work?

233
00:06:04.982 --> 00:06:10.365
They use pretty standard simple moving average or SMA crossovers on the NASDAQ 100 index,

234
00:06:10.411 --> 00:06:10.943
the NDFs,

235
00:06:11.271 --> 00:06:15.958
things like the one day crossing the 50 day or the five day crossing the 150,

236
00:06:16.036 --> 00:06:16.818
different time frames.

237
00:06:16.896 --> 00:06:17.005
OK,

238
00:06:17.130 --> 00:06:18.490
short term versus longer term trends.

239
00:06:18.491 --> 00:06:20.115
And what did those signals do to the caller?

240
00:06:20.474 --> 00:06:20.974
Generally,

241
00:06:21.552 --> 00:06:21.677
a...

242
00:06:21.841 --> 00:06:22.702
buy signal,

243
00:06:22.982 --> 00:06:23.963
faster average,

244
00:06:24.004 --> 00:06:25.304
crossing above slower,

245
00:06:25.785 --> 00:06:27.304
would lead them to widen the collar.

246
00:06:27.527 --> 00:06:31.812
So they'd choose a put that was further out of the money and a pull that was also further out of the money.

247
00:06:31.929 --> 00:06:33.875
Giving it more room to run on the upside,

248
00:06:34.054 --> 00:06:35.351
maybe less immediate protection.

249
00:06:35.414 --> 00:06:35.672
Right.

250
00:06:36.015 --> 00:06:37.844
And a sell signal would do the opposite.

251
00:06:37.845 --> 00:06:38.844
It would tighten the collar,

252
00:06:39.359 --> 00:06:40.719
bring the put strike closer,

253
00:06:40.844 --> 00:06:43.875
bring the call streak closer for more immediate protection,

254
00:06:43.969 --> 00:06:45.359
but less upside potential.

255
00:06:45.531 --> 00:06:45.750
Okay.

256
00:06:45.969 --> 00:06:46.562
That makes sense.

257
00:06:47.000 --> 00:06:47.922
Adapting to the trend.

258
00:06:48.172 --> 00:06:50.937
What about the volatility signal based on the VIX?

259
00:06:51.047 --> 00:06:51.172
Yep.

260
00:06:51.297 --> 00:06:51.998
Based on the VIX,

261
00:06:52.218 --> 00:06:55.760
they looked at where the VIX was relative to its own moving average,

262
00:06:55.780 --> 00:06:56.421
like 50-day,

263
00:06:57.042 --> 00:06:57.182
150-day,

264
00:06:57.264 --> 00:06:57.382
etc.

265
00:06:57.803 --> 00:07:00.049
So is volatility high or low compared to usual?

266
00:07:00.206 --> 00:07:00.807
Exactly.

267
00:07:00.987 --> 00:07:02.448
If the VIX was really high,

268
00:07:02.510 --> 00:07:07.409
like more than one standard deviation above its average indicating high anxiety or fear,

269
00:07:07.768 --> 00:07:09.815
they'd actually write fewer call options.

270
00:07:09.893 --> 00:07:14.221
Maybe only 0.75 calls per QQQ share instead of the usual one.

271
00:07:14.518 --> 00:07:14.940
Interesting.

272
00:07:15.299 --> 00:07:17.128
Why fewer calls when fear is high?

273
00:07:17.299 --> 00:07:17.784
The thinking,

274
00:07:17.799 --> 00:07:18.206
I believe,

275
00:07:18.207 --> 00:07:20.049
is that high VIX often precedes

276
00:07:20.337 --> 00:07:21.358
sharp market reversals,

277
00:07:21.438 --> 00:07:22.699
including potential bounces.

278
00:07:23.340 --> 00:07:28.603
So writing fewer calls leaves a bit more potential upside open if the market snaps back quickly.

279
00:07:29.408 --> 00:07:29.963
Conversely,

280
00:07:30.166 --> 00:07:31.549
in low anxiety periods,

281
00:07:31.611 --> 00:07:33.228
they might write more calls,

282
00:07:33.330 --> 00:07:34.994
maybe 1.25 per share.

283
00:07:35.416 --> 00:07:36.275
Adjusting the ratio.

284
00:07:36.619 --> 00:07:38.681
And the third signal was macroeconomic.

285
00:07:38.791 --> 00:07:38.963
Yeah,

286
00:07:39.103 --> 00:07:40.478
this combined two things.

287
00:07:40.994 --> 00:07:44.978
The trend in initial unemployment claims using moving average crossovers again,

288
00:07:45.041 --> 00:07:46.416
like one week versus 10 week,

289
00:07:46.697 --> 00:07:48.385
and the official NBER

290
00:07:48.577 --> 00:07:49.797
business cycle status,

291
00:07:49.857 --> 00:07:51.699
whether we're in expansion or contraction.

292
00:07:51.978 --> 00:07:53.197
How did that influence the caller?

293
00:07:53.997 --> 00:07:55.158
It depended on the cycle.

294
00:07:55.458 --> 00:07:55.958
For example,

295
00:07:55.997 --> 00:07:56.900
during an expansion,

296
00:07:56.939 --> 00:07:59.517
if unemployment claims started rising a bad sign,

297
00:07:59.556 --> 00:08:00.478
they'd widen the collar,

298
00:08:00.681 --> 00:08:02.376
move the put and call further OTM.

299
00:08:02.900 --> 00:08:04.103
But during a contraction,

300
00:08:04.322 --> 00:08:07.064
rising claims might actually lead them to tighten the collar.

301
00:08:07.486 --> 00:08:11.376
The idea is the market might react differently depending on the overall economic backdrop.

302
00:08:11.845 --> 00:08:13.783
That's quite sophisticated layering those signals.

303
00:08:14.158 --> 00:08:15.861
So on roll dates monthly for calls,

304
00:08:16.048 --> 00:08:18.001
six monthly for the puts in the main strategy.

305
00:08:18.325 --> 00:08:20.985
they'd calculate these signals and use formulas to pick the options.

306
00:08:21.386 --> 00:08:21.725
That's right.

307
00:08:21.946 --> 00:08:27.308
The formulas directly use the signal values to determine the percentage OTM for the call and the put,

308
00:08:27.784 --> 00:08:30.972
and also that call writing ratio based on the mixed signal.

309
00:08:31.409 --> 00:08:31.948
For instance,

310
00:08:32.245 --> 00:08:38.448
positive momentum and positive macro signals might both add to how far OTM the call strike was set.

311
00:08:39.011 --> 00:08:39.386
Got it.

312
00:08:39.495 --> 00:08:41.651
So the million dollar question,

313
00:08:43.308 --> 00:08:45.276
did this active approach actually work?

314
00:08:45.393 --> 00:08:48.177
better over the full 1999-2010 period?

315
00:08:48.276 --> 00:08:49.257
It seems like it did.

316
00:08:49.458 --> 00:08:50.536
The active strategies,

317
00:08:50.599 --> 00:08:53.622
particularly the one using the shorter term momentum and macro signals,

318
00:08:54.044 --> 00:08:58.028
generally outperformed both the QQQ itself and the benchmark passive caller.

319
00:08:58.224 --> 00:08:58.825
By how much?

320
00:08:58.942 --> 00:08:59.645
The best one,

321
00:08:59.685 --> 00:09:00.864
the short-term active strategy,

322
00:09:01.310 --> 00:09:03.130
delivered about 12.5%

323
00:09:03.131 --> 00:09:06.192
annualized return with volatility just over 11%.

324
00:09:06.380 --> 00:09:08.177
So higher return than the passive caller,

325
00:09:08.333 --> 00:09:09.333
slightly higher vol,

326
00:09:09.442 --> 00:09:11.536
but better risk-adjusted performance overall.

327
00:09:11.755 --> 00:09:14.708
Measures like the Stutzer index and Leland's alpha were higher.

328
00:09:14.945 --> 00:09:15.846
Better bang for your buck,

329
00:09:16.166 --> 00:09:16.627
risk-wise.

330
00:09:16.628 --> 00:09:17.127
Pretty much.

331
00:09:17.367 --> 00:09:19.408
And how did Active do in those specific periods,

332
00:09:19.428 --> 00:09:20.349
like the tech bubble?

333
00:09:20.490 --> 00:09:23.013
It significantly outperformed the passive caller during the tech bubble.

334
00:09:23.193 --> 00:09:23.834
Better returns,

335
00:09:23.873 --> 00:09:24.732
better risk metrics.

336
00:09:25.076 --> 00:09:27.381
And the unfavorable period,

337
00:09:27.396 --> 00:09:27.974
the bull run,

338
00:09:29.279 --> 00:09:30.943
did Active help at the lag?

339
00:09:32.030 --> 00:09:33.711
It seemed to mitigate it somewhat.

340
00:09:33.831 --> 00:09:38.052
The active strategies still lag the QQQ during that strong uptrend,

341
00:09:38.411 --> 00:09:40.091
but less so than the passive one.

342
00:09:40.114 --> 00:09:42.130
They didn't sacrifice quite as much upside.

343
00:09:42.169 --> 00:09:43.216
And the credit crisis.

344
00:09:43.411 --> 00:09:45.294
Active outperformed passive again,

345
00:09:45.770 --> 00:09:49.434
and both obviously crushed the QQQ during that downturn.

346
00:09:49.872 --> 00:09:54.216
It really suggests that adapting the collar structure based on these signals added value,

347
00:09:54.278 --> 00:09:55.559
especially in volatile times.

348
00:09:55.778 --> 00:09:57.622
And they even had a little bit of out-of-sample data.

349
00:09:57.809 --> 00:09:57.981
Yeah,

350
00:09:57.997 --> 00:10:00.591
a short period from June 2009 to September 10th.

351
00:10:00.858 --> 00:10:02.139
after the main study period.

352
00:10:02.580 --> 00:10:03.400
And in that window,

353
00:10:03.541 --> 00:10:06.324
the active callers also outperformed the passive ones.

354
00:10:06.746 --> 00:10:07.984
It's not definitive proof,

355
00:10:08.207 --> 00:10:09.090
but it's encouraging.

356
00:10:09.347 --> 00:10:09.769
Interesting.

357
00:10:10.066 --> 00:10:14.793
They also mentioned applying this to a mutual fund where you can't directly buy options on the fund itself.

358
00:10:14.894 --> 00:10:15.175
Right.

359
00:10:15.254 --> 00:10:16.636
They showed how you could adapt it.

360
00:10:16.652 --> 00:10:18.269
Say you have a small cap mutual fund,

361
00:10:18.300 --> 00:10:20.925
you could calculate its rolling beta relative to the QQQ,

362
00:10:21.675 --> 00:10:23.144
look at the relative price levels,

363
00:10:23.394 --> 00:10:26.425
and then scale the QQQ caller position accordingly.

364
00:10:26.660 --> 00:10:28.972
So you'd use QQQ options as a proxy hedge.

365
00:10:29.314 --> 00:10:29.974
Exactly.

366
00:10:30.094 --> 00:10:37.415
You'd have to rebalance the size of the QQQ caller daily to maintain the right hedge ratio as the fund's price and its beta changed.

367
00:10:37.634 --> 00:10:38.798
And did it work,

368
00:10:39.134 --> 00:10:41.259
applying active and passive callers this way?

369
00:10:41.697 --> 00:10:46.415
They said the results were broadly similar to when they applied the callers directly to the QQQ,

370
00:10:47.165 --> 00:10:49.275
so it suggests the approach can be adapted.

371
00:10:49.712 --> 00:10:49.837
Okay,

372
00:10:49.838 --> 00:10:50.618
so wrapping this up,

373
00:10:50.697 --> 00:10:53.384
what are the main takeaways for someone thinking about callers?

374
00:10:53.822 --> 00:10:53.947
Well,

375
00:10:54.040 --> 00:10:54.415
first,

376
00:10:54.662 --> 00:10:57.986
even simple passive callers showed real value in reducing risk,

377
00:10:58.085 --> 00:10:59.667
especially big drawdowns.

378
00:11:00.046 --> 00:11:05.234
And they actually proved returns over that whole volatile period compared to just holding QQQ.

379
00:11:05.632 --> 00:11:07.296
Definitely valuable during market stress.

380
00:11:07.515 --> 00:11:10.039
But the really compelling part is the active strategy.

381
00:11:10.460 --> 00:11:12.781
By adapting the caller rules based on momentum,

382
00:11:12.921 --> 00:11:14.468
volatility and macro signals,

383
00:11:14.765 --> 00:11:18.062
they demonstrated the potential to outperform the passive approach,

384
00:11:18.328 --> 00:11:20.312
both in and out of sample in this study.

385
00:11:20.546 --> 00:11:22.203
So being dynamic seems to pay off.

386
00:11:22.453 --> 00:11:23.109
It appears so.

387
00:11:23.870 --> 00:11:24.431
But of course,

388
00:11:24.451 --> 00:11:29.595
a key thing to remember is that the effectiveness of any caller strategy really depends on the market environment.

389
00:11:30.095 --> 00:11:31.915
They shine in choppy or falling markets,

390
00:11:32.056 --> 00:11:33.860
but can be a drag in strong,

391
00:11:34.040 --> 00:11:35.102
steady uptrends.

392
00:11:35.360 --> 00:11:36.962
That trade-off is always there.

393
00:11:37.204 --> 00:11:37.767
Absolutely.

394
00:11:38.446 --> 00:11:38.985
And finally,

395
00:11:39.024 --> 00:11:43.149
the research suggests these strategies aren't just limited to assets with direct options.

396
00:11:43.415 --> 00:11:46.509
They can potentially be adapted using correlated index options,

397
00:11:46.540 --> 00:11:48.352
like they showed with the mutual fund example.

398
00:11:48.431 --> 00:11:49.009
Great summary.

399
00:11:49.274 --> 00:11:51.821
Lots to think about there regarding dynamic risk management.

400
00:11:51.915 --> 00:11:52.259
For sure.

401
00:11:52.498 --> 00:11:55.140
Thank you for tuning in to Papers with Backtests podcast.

402
00:11:55.582 --> 00:12:00.664
We hope today's deep dive gave you useful insights into the world of active and passive collar strategies.

403
00:12:01.625 --> 00:12:03.992
Join us next time as we break down more research.

404
00:12:04.773 --> 00:12:06.414
And for more papers and backtests,

405
00:12:06.789 --> 00:12:10.117
find us at https.paperswithbacktests.com.

406
00:12:11.070 --> 00:12:11.601
Happy trading!