1
00:00:03,240 --> 00:00:08,280
Speaker 1: This is Masters in Business with Barry Ridholts on Bloomberg Radio. Hi,

2
00:00:08,440 --> 00:00:11,160
Speaker 1: welcome to the podcast. This is Barry rid Holts and

3
00:00:11,280 --> 00:00:15,920
Speaker 1: my guest today is Joe Saluzi. You probably know him

4
00:00:16,680 --> 00:00:20,720
Speaker 1: as co author of the book Broken Markets or a

5
00:00:20,840 --> 00:00:26,560
Speaker 1: blog poster at Famous Trading. He's actually UH an institutional trader,

6
00:00:26,960 --> 00:00:31,040
Speaker 1: agency trader, agency broker on behalf of big mutual funds,

7
00:00:31,040 --> 00:00:34,080
Speaker 1: hedge funds, what have you. Joe and his partner sal

8
00:00:34,240 --> 00:00:40,440
Speaker 1: were very early in identifying problems with markets with market structures.

9
00:00:41,080 --> 00:00:45,960
Speaker 1: They identified some glitches from algorithmic trading, from high frequency trading,

10
00:00:46,400 --> 00:00:50,040
Speaker 1: issues with dark pools, and they've been a pretty vocal

11
00:00:50,200 --> 00:00:55,320
Speaker 1: set of um more than critics, but participants in the

12
00:00:55,360 --> 00:00:59,880
Speaker 1: markets who want to make them better, fairer, less expense

13
00:01:00,040 --> 00:01:04,600
Speaker 1: of uh more even handed for for all participants. Uh

14
00:01:04,640 --> 00:01:08,520
Speaker 1: they've brought a lot of um let's call it malfeasance

15
00:01:08,880 --> 00:01:12,119
Speaker 1: to the public's attention. Their first white paper had come

16
00:01:12,120 --> 00:01:13,760
Speaker 1: out in two thousand eight, right in the middle of

17
00:01:13,840 --> 00:01:16,399
Speaker 1: financial crisis, when no one was thinking about things like

18
00:01:16,880 --> 00:01:19,679
Speaker 1: market structure. But once we had the two thousand and

19
00:01:19,680 --> 00:01:22,520
Speaker 1: ten flash crash, I think people started paying a whole

20
00:01:22,560 --> 00:01:26,720
Speaker 1: lot more attention to them. I know these guys probably

21
00:01:27,120 --> 00:01:31,080
Speaker 1: for five or six years, um, through the usual circuits, radio,

22
00:01:31,160 --> 00:01:34,039
Speaker 1: TV conferences, what have you. You know, it's a really

23
00:01:34,040 --> 00:01:37,160
Speaker 1: small world and you end up bumping into them. Today's

24
00:01:37,240 --> 00:01:40,760
Speaker 1: conversation is a little technical, a little wonky. If you

25
00:01:40,800 --> 00:01:43,600
Speaker 1: find this sort of stuff interesting, I think you'll find

26
00:01:43,600 --> 00:01:47,040
Speaker 1: it fascinating. Uh. There are very very few people who

27
00:01:47,040 --> 00:01:51,560
Speaker 1: are as as knowledgeable from a trading perspective about market

28
00:01:51,600 --> 00:01:57,280
Speaker 1: structures and potential flaws in what's going on. And you'll hear.

29
00:01:57,760 --> 00:02:01,040
Speaker 1: You know, things have gotten better for the flash crash,

30
00:02:01,400 --> 00:02:04,160
Speaker 1: but not nearly as as good as as Joe and

31
00:02:04,160 --> 00:02:07,640
Speaker 1: sal would like to see them. Um. So that's a positive,

32
00:02:07,800 --> 00:02:10,359
Speaker 1: but you know, we're still not quite there, and there's

33
00:02:10,360 --> 00:02:14,399
Speaker 1: still risk of another flash crash, not only in equities,

34
00:02:14,440 --> 00:02:19,680
Speaker 1: but as we've seen in treasuries, in currencies. It turns

35
00:02:19,680 --> 00:02:23,840
Speaker 1: out that commodities maybe the best structured market there is

36
00:02:23,919 --> 00:02:27,120
Speaker 1: because they have some inherent rules and inherent structures built

37
00:02:27,160 --> 00:02:30,040
Speaker 1: in that work better than some of these other markets.

38
00:02:30,080 --> 00:02:32,640
Speaker 1: So I could keep babbling about this for hours, but

39
00:02:32,800 --> 00:02:35,120
Speaker 1: rather than do that, let's just go right to my

40
00:02:35,240 --> 00:02:42,959
Speaker 1: conversation with Joe Saluzzi. Of Famous Trading. This is Masters

41
00:02:43,000 --> 00:02:47,519
Speaker 1: in Business with Barry Ridholts on Bloomberg Radio. My guest

42
00:02:47,600 --> 00:02:51,440
Speaker 1: this week is Joe Saluzzi of Famous Trading. Him and

43
00:02:51,520 --> 00:02:54,680
Speaker 1: his partner are the authors of Broken Markets. How high

44
00:02:54,680 --> 00:02:58,680
Speaker 1: frequency trading and predatory practices on Wall Street are destroying

45
00:02:58,720 --> 00:03:03,840
Speaker 1: investor confidences and hurting your portfolios. Joe, Welcome to Bloomberg.

46
00:03:04,000 --> 00:03:06,079
Speaker 1: Very thick, great, thanks for having me here today. So

47
00:03:06,200 --> 00:03:09,200
Speaker 1: I wanted to have you guys on, and by coincidence,

48
00:03:09,360 --> 00:03:11,760
Speaker 1: it works out that you follow a week where we

49
00:03:11,840 --> 00:03:15,760
Speaker 1: had a quant Cliff ast Nest fascinating conversation, but he's

50
00:03:15,800 --> 00:03:19,560
Speaker 1: been a fairly he calls it tepid defender of high

51
00:03:19,639 --> 00:03:24,240
Speaker 1: frequency trading before. So so it just turned out fortuitously

52
00:03:24,320 --> 00:03:27,160
Speaker 1: that we have you on the week after UM and

53
00:03:27,320 --> 00:03:30,600
Speaker 1: coming up we have Brad Katsuyama from i X and

54
00:03:30,639 --> 00:03:33,680
Speaker 1: he's also going to discuss market structures. But before we

55
00:03:33,720 --> 00:03:36,400
Speaker 1: get into the nitty gritty of this, a little background

56
00:03:36,520 --> 00:03:40,880
Speaker 1: about yourself. UM finance, bachelor's at n y U and

57
00:03:40,920 --> 00:03:44,800
Speaker 1: then an NBA at Chapel Hill down at North Carolina.

58
00:03:45,120 --> 00:03:48,600
Speaker 1: Before you, guys founded Famous Trading, you worked at instant

59
00:03:48,600 --> 00:03:51,520
Speaker 1: at But the question I really have to ask is

60
00:03:51,920 --> 00:03:56,720
Speaker 1: how did you find your way into investigating issues in

61
00:03:57,120 --> 00:03:59,880
Speaker 1: market structure. Yeah, that's a great question. And if you

62
00:04:00,040 --> 00:04:02,160
Speaker 1: to ask me, you know, five or eight years ago,

63
00:04:02,160 --> 00:04:04,360
Speaker 1: would we be here saying absolutely not. You know, our

64
00:04:04,440 --> 00:04:06,800
Speaker 1: our job is to trade stocks. That's what we do

65
00:04:06,840 --> 00:04:10,400
Speaker 1: for a living. Where agency brokers, we trade for institutional clients,

66
00:04:10,440 --> 00:04:12,760
Speaker 1: hadge funds, mutual funds, and so on. But we got

67
00:04:12,760 --> 00:04:14,840
Speaker 1: our start as you mentioned, Sal and I both worked

68
00:04:14,840 --> 00:04:18,599
Speaker 1: at Instant in the nineties, in early nineties before electronic trading,

69
00:04:18,640 --> 00:04:20,880
Speaker 1: just as electronic trading was really taken off in the

70
00:04:20,920 --> 00:04:23,560
Speaker 1: equities business, so we kind of we cut our teeth

71
00:04:23,640 --> 00:04:26,760
Speaker 1: trading electronically. We learned a lot there, We really saw

72
00:04:26,800 --> 00:04:29,760
Speaker 1: it progressing, and then started to see the rule changes

73
00:04:29,800 --> 00:04:32,840
Speaker 1: happening mainly at the SEC some things like regg ets

74
00:04:32,920 --> 00:04:35,400
Speaker 1: and a few other regulations. So let's let's get into

75
00:04:35,440 --> 00:04:37,320
Speaker 1: that in a little bit. But for for the lay

76
00:04:37,360 --> 00:04:39,400
Speaker 1: person who may be listening to this and not familiar

77
00:04:39,400 --> 00:04:42,600
Speaker 1: with Instant, I began my career as a trader, So

78
00:04:42,680 --> 00:04:44,760
Speaker 1: you had either a level two or level three depending

79
00:04:44,800 --> 00:04:47,120
Speaker 1: on the access, whether you were buyer or whether you're

80
00:04:47,200 --> 00:04:52,120
Speaker 1: making markets and Instant. It was a private network that institutions.

81
00:04:52,360 --> 00:04:56,000
Speaker 1: Institutional network used to buy and sell stocks with each

82
00:04:56,000 --> 00:04:59,120
Speaker 1: other when they maybe were trading a little outside of

83
00:04:59,160 --> 00:05:02,800
Speaker 1: the market, outside of the been asking moving real size

84
00:05:03,000 --> 00:05:05,480
Speaker 1: so as to not disrupt the market. Is that a

85
00:05:05,600 --> 00:05:09,440
Speaker 1: fair Yeah, that that was its original intent exactly you said. Um.

86
00:05:09,880 --> 00:05:13,440
Speaker 1: Eventually brokers came on the system, just like today, finding

87
00:05:13,480 --> 00:05:16,560
Speaker 1: liquidity was tough back then when you had two institutions

88
00:05:16,560 --> 00:05:19,640
Speaker 1: trying to cross stock, they didn't exactly always meet. Brokers

89
00:05:19,680 --> 00:05:21,599
Speaker 1: were then allowed on the system, but the real action

90
00:05:21,640 --> 00:05:25,640
Speaker 1: on Instant was between institutional investors was blocking up stock,

91
00:05:25,920 --> 00:05:28,239
Speaker 1: was trying to put up fifty a hundred thousand shares

92
00:05:28,480 --> 00:05:31,080
Speaker 1: without moving the stock. And and that's the trick and

93
00:05:31,120 --> 00:05:33,520
Speaker 1: even today there's still the trick to the trading stocks

94
00:05:33,560 --> 00:05:35,440
Speaker 1: is trying to find a piece of liquidity, trying to

95
00:05:35,440 --> 00:05:38,240
Speaker 1: find that block without having the noise come through and

96
00:05:38,279 --> 00:05:40,680
Speaker 1: a lot a lot of that excess leakage. So that's

97
00:05:40,720 --> 00:05:43,080
Speaker 1: what our job was. Our job was to facilitate trades.

98
00:05:43,120 --> 00:05:45,120
Speaker 1: We weren't prop traders or any agency guys just like

99
00:05:45,200 --> 00:05:47,200
Speaker 1: we are now. So so you traded on behalf of

100
00:05:47,520 --> 00:05:51,560
Speaker 1: endowments foundations truss hedge funds, which were much moller way

101
00:05:51,560 --> 00:05:54,080
Speaker 1: back then. Mutual funds were the big boys on the

102
00:05:54,080 --> 00:05:58,080
Speaker 1: block in those days, and so your business were commissioned brokers.

103
00:05:58,080 --> 00:06:01,200
Speaker 1: You would get paid for the trade you executed on

104
00:06:01,240 --> 00:06:03,840
Speaker 1: their behalf right and even better, actually the clients had

105
00:06:03,839 --> 00:06:06,960
Speaker 1: the system on their desk themselves, so we were actually

106
00:06:06,960 --> 00:06:09,440
Speaker 1: we were more coverage people. So we covered an account

107
00:06:09,760 --> 00:06:11,839
Speaker 1: and whether it's you know, you kind of helped that

108
00:06:11,880 --> 00:06:14,320
Speaker 1: account along, but they can actually hit the keys themselves.

109
00:06:14,320 --> 00:06:16,679
Speaker 1: They didn't need us. And the power that the buy side,

110
00:06:16,720 --> 00:06:19,520
Speaker 1: particularly the institutional investment community got, that was the first

111
00:06:19,560 --> 00:06:21,720
Speaker 1: time where they were really involved in the process and

112
00:06:21,720 --> 00:06:24,440
Speaker 1: they really liked it, which is why the instant revenue

113
00:06:24,480 --> 00:06:26,600
Speaker 1: stream went like a ski slope from the early nineties.

114
00:06:26,600 --> 00:06:30,719
Speaker 1: It really ramp. It was so from back in the nineties.

115
00:06:30,960 --> 00:06:35,480
Speaker 1: What has changed in the markets? A lot, uh, well,

116
00:06:35,520 --> 00:06:37,560
Speaker 1: a lot for the good, A lot for so you know,

117
00:06:37,600 --> 00:06:39,159
Speaker 1: we're known as critics and so on, but we don't

118
00:06:39,240 --> 00:06:40,800
Speaker 1: we see a lot of good that has come over

119
00:06:41,000 --> 00:06:44,839
Speaker 1: over the last ten fifteen years in the market. Narrowing spreads,

120
00:06:45,160 --> 00:06:48,679
Speaker 1: lower commission costs um and that came about really around

121
00:06:48,680 --> 00:06:51,640
Speaker 1: two thousand and two that's when decimalization came out, when

122
00:06:51,680 --> 00:06:54,280
Speaker 1: they changed those audits and quarters right that we were

123
00:06:54,320 --> 00:06:57,240
Speaker 1: trade actions and everybody back, everybody who was on a

124
00:06:57,279 --> 00:07:00,800
Speaker 1: trading desk knew what teenys were teeny the courts even

125
00:07:01,600 --> 00:07:05,120
Speaker 1: it was crazy stuff. It was sixteenth of a of

126
00:07:05,160 --> 00:07:07,520
Speaker 1: a cent. And I was good at doing math in

127
00:07:07,560 --> 00:07:09,120
Speaker 1: my head, but a lot of people had a hard

128
00:07:09,160 --> 00:07:12,080
Speaker 1: time with it. Decimalization was good for the people who

129
00:07:12,080 --> 00:07:15,640
Speaker 1: were a little fractionally challenged. Let's call it absolutely a

130
00:07:15,720 --> 00:07:18,200
Speaker 1: good thing. Um. One thing that they forgot about was,

131
00:07:18,400 --> 00:07:20,160
Speaker 1: you know, putting in some sort of minimum spread. But

132
00:07:20,160 --> 00:07:21,400
Speaker 1: we can get to that later when it comes to

133
00:07:21,440 --> 00:07:25,240
Speaker 1: small caps. But really, electronic trading is what brought down

134
00:07:25,320 --> 00:07:27,840
Speaker 1: the cost of trading right today. There could be this

135
00:07:27,880 --> 00:07:29,600
Speaker 1: whole argument, we can get into it later about high

136
00:07:29,600 --> 00:07:32,640
Speaker 1: frequency trading and whether or not that increased or decreased cost.

137
00:07:32,880 --> 00:07:34,560
Speaker 1: But the real fact, and you look at charts and

138
00:07:34,600 --> 00:07:38,720
Speaker 1: you can see this transaction cost for institutional investors dropped

139
00:07:39,040 --> 00:07:42,080
Speaker 1: around two thousand and two, two thousand and three four,

140
00:07:42,280 --> 00:07:44,680
Speaker 1: that's when the real drop and the real decline came in.

141
00:07:45,080 --> 00:07:47,560
Speaker 1: Since so oh seven, we're oh, wait, we've kind of

142
00:07:47,640 --> 00:07:49,760
Speaker 1: leveled off. We haven't. We've gone down and we've pretty

143
00:07:49,840 --> 00:07:52,280
Speaker 1: much stay down there. So that drop in trading costs,

144
00:07:52,360 --> 00:07:56,520
Speaker 1: did it imbute to the advantage of the investor the institution?

145
00:07:56,600 --> 00:08:00,200
Speaker 1: Where did that squeezing of margins end up falling? That

146
00:08:00,280 --> 00:08:02,160
Speaker 1: that should go directly to an investment because you're a

147
00:08:02,160 --> 00:08:03,960
Speaker 1: if you're a hedge fund or an investment advisor and

148
00:08:03,960 --> 00:08:07,440
Speaker 1: you have fees and expenses, transaction costs is certainly one

149
00:08:07,480 --> 00:08:10,400
Speaker 1: of them, or at least explicitly commissioned costs. You know,

150
00:08:10,440 --> 00:08:13,000
Speaker 1: there's more of an implicit transaction costs, which will you know,

151
00:08:13,040 --> 00:08:14,520
Speaker 1: we talk about all the time on the on the

152
00:08:14,600 --> 00:08:16,800
Speaker 1: institutional world, is how much does it costs you to

153
00:08:16,880 --> 00:08:19,080
Speaker 1: really get into it and out of that stock? How

154
00:08:19,160 --> 00:08:21,760
Speaker 1: much leakage was there? Did you move the stock a

155
00:08:21,800 --> 00:08:24,360
Speaker 1: certain amount, did you get the v wap the volume

156
00:08:24,360 --> 00:08:26,880
Speaker 1: weight at average price, or did you actually increase it

157
00:08:26,960 --> 00:08:28,800
Speaker 1: because you left a lot of trails and that that

158
00:08:28,920 --> 00:08:32,000
Speaker 1: is really what implicit transaction cost is, and that, you know,

159
00:08:32,520 --> 00:08:34,040
Speaker 1: that's one of those things that's kind of stayed the

160
00:08:34,040 --> 00:08:36,760
Speaker 1: same over the last few years. So the approach of

161
00:08:36,800 --> 00:08:39,160
Speaker 1: big institutions, or at least it used to be and

162
00:08:39,200 --> 00:08:42,200
Speaker 1: I'm assuming to some degree it still is is to

163
00:08:42,240 --> 00:08:44,719
Speaker 1: be a little stealthy and to try and get these

164
00:08:44,760 --> 00:08:48,560
Speaker 1: trades off without just you know, the signs in the

165
00:08:48,600 --> 00:08:51,760
Speaker 1: marina are at this point on it's five miles per hour,

166
00:08:51,880 --> 00:08:53,959
Speaker 1: no wake. They don't want to leave a giant wake

167
00:08:54,000 --> 00:08:56,000
Speaker 1: in the market. They want to sneak in and sneak out.

168
00:08:56,360 --> 00:08:58,520
Speaker 1: And really that's my job, my real job now, and

169
00:08:58,600 --> 00:09:01,199
Speaker 1: when we're not writing books and writing log posts, to

170
00:09:01,360 --> 00:09:03,680
Speaker 1: trade for clients and to leave is little weak and

171
00:09:03,679 --> 00:09:05,880
Speaker 1: that's a great way of looking at it as possible

172
00:09:05,920 --> 00:09:08,920
Speaker 1: so that others aren't pretty much attaching to what you're doing.

173
00:09:09,120 --> 00:09:11,520
Speaker 1: And it's difficult, even though everyone thinks that there's all

174
00:09:11,559 --> 00:09:14,680
Speaker 1: electronics and algorithms and whatnot, it is really difficult to

175
00:09:14,720 --> 00:09:17,400
Speaker 1: make sure that you're not being spotted and taken advantage of.

176
00:09:17,559 --> 00:09:19,760
Speaker 1: And that's how you lower your transaction costs. You lower

177
00:09:19,840 --> 00:09:22,240
Speaker 1: them by not being spotted. I'm Barry ridd Holts. You're

178
00:09:22,280 --> 00:09:25,640
Speaker 1: listening to Masters in Business on Bloomberg Radio. My guest

179
00:09:25,679 --> 00:09:29,720
Speaker 1: today Joe Saluzzi of Famous Trading, co author of the

180
00:09:29,760 --> 00:09:33,880
Speaker 1: book Broken Markets. Let's continue our conversation a bit about

181
00:09:34,040 --> 00:09:38,800
Speaker 1: what you've described as a market structure that's become broken.

182
00:09:39,160 --> 00:09:43,040
Speaker 1: So we were talking earlier about the nineties, how things

183
00:09:43,120 --> 00:09:45,800
Speaker 1: change today, They've changed a lot when it comes to

184
00:09:45,880 --> 00:09:49,320
Speaker 1: the actual who's trading lot and the types of players

185
00:09:49,320 --> 00:09:52,040
Speaker 1: that are out there, and the the diversity of liquidity

186
00:09:52,040 --> 00:09:54,959
Speaker 1: and diversity of market participants. So what we found out,

187
00:09:54,960 --> 00:09:57,200
Speaker 1: and I'll just back up in real quickly, we found

188
00:09:57,440 --> 00:09:59,920
Speaker 1: as we were trading throughout the middle of two thousands too,

189
00:10:00,000 --> 00:10:02,199
Speaker 1: thousand and six, two thousand and seven, two thousand and

190
00:10:02,280 --> 00:10:05,079
Speaker 1: eight really changed a lot. And there was a reason

191
00:10:05,120 --> 00:10:06,839
Speaker 1: why it changed a lot, and it was because of

192
00:10:06,840 --> 00:10:10,120
Speaker 1: a regulation called Reagan MS that the SEC approved. That's

193
00:10:10,200 --> 00:10:13,960
Speaker 1: the National market system. What what did that regulation do?

194
00:10:14,240 --> 00:10:17,920
Speaker 1: It pretty much exploded the equity universe. Were used to

195
00:10:18,000 --> 00:10:21,280
Speaker 1: have particularly listed names on the ux dock scene, names

196
00:10:21,320 --> 00:10:23,600
Speaker 1: where there was a specialist down on the floor and

197
00:10:23,640 --> 00:10:26,840
Speaker 1: a human was actually buying and selling and and setting

198
00:10:26,840 --> 00:10:29,120
Speaker 1: the price correct. And they had about eight percent market

199
00:10:29,160 --> 00:10:31,640
Speaker 1: share on the New York Stock Exchange pre oh seven,

200
00:10:31,640 --> 00:10:35,199
Speaker 1: pre Reagan MS. Now they're around and that dropped on

201
00:10:35,360 --> 00:10:37,080
Speaker 1: was overnight as soon as the rule was put in place.

202
00:10:37,200 --> 00:10:40,520
Speaker 1: Reagan MS, basically it was intended to help competition. It

203
00:10:40,600 --> 00:10:43,600
Speaker 1: wanted to allow other exchanges and other venues to post

204
00:10:43,640 --> 00:10:46,440
Speaker 1: their bids and offers and have people interact with them

205
00:10:46,520 --> 00:10:49,480
Speaker 1: and not ignore those quotes. Conceptually, that's a great idea.

206
00:10:49,640 --> 00:10:51,120
Speaker 1: You should have to trade with a venue that has

207
00:10:51,120 --> 00:10:53,280
Speaker 1: a better price, right well thinks, in other words, the

208
00:10:53,600 --> 00:10:57,200
Speaker 1: best bit. So if you're a seller, you should your

209
00:10:57,400 --> 00:11:00,200
Speaker 1: broker who's operating in your half, on your behalf, be

210
00:11:00,280 --> 00:11:03,680
Speaker 1: hitting the highest bid, the best bid as opposed to

211
00:11:03,760 --> 00:11:05,920
Speaker 1: just what's most convenient. That's exactly it. And and one

212
00:11:05,920 --> 00:11:07,360
Speaker 1: of the parts of the rule is called the Trade

213
00:11:07,360 --> 00:11:10,240
Speaker 1: through Protection or rule six eleven. And that's what it says.

214
00:11:10,240 --> 00:11:12,760
Speaker 1: If the best bid or offer before you can trade,

215
00:11:13,120 --> 00:11:15,280
Speaker 1: let's say you say the best offers ten cents before

216
00:11:15,320 --> 00:11:17,280
Speaker 1: you can trade a fifteen cents, you've got to clear

217
00:11:17,320 --> 00:11:19,400
Speaker 1: the ten cents. So that for a block trader that

218
00:11:19,440 --> 00:11:21,920
Speaker 1: becomes potentially a problem. You've got to kind of sweep

219
00:11:21,960 --> 00:11:24,440
Speaker 1: through those levels, which we can do through what they

220
00:11:24,440 --> 00:11:26,960
Speaker 1: call electronic give They have an ISO orders into market

221
00:11:27,000 --> 00:11:29,400
Speaker 1: sweep boarders with. But the bottom line was, all of

222
00:11:29,440 --> 00:11:33,640
Speaker 1: a sudden, once this this fragmentation with the sec called competition,

223
00:11:33,880 --> 00:11:38,199
Speaker 1: we call fragmentation it allowed noise and different types of

224
00:11:38,280 --> 00:11:42,280
Speaker 1: venues to start coming through. And that's where the games began.

225
00:11:42,360 --> 00:11:44,880
Speaker 1: And that's where the high speed traders, the high frequency

226
00:11:44,920 --> 00:11:47,600
Speaker 1: traders as their so called, so they invent they started

227
00:11:47,600 --> 00:11:49,760
Speaker 1: coming in. So let's define that, because I know people

228
00:11:49,800 --> 00:11:52,160
Speaker 1: are gonna say, I hear about h f T. I

229
00:11:52,200 --> 00:11:55,600
Speaker 1: hear about, but what are high frequency traders? Really? It's

230
00:11:55,600 --> 00:11:58,720
Speaker 1: a great question. And I was on a CFTC subcommittee

231
00:11:58,720 --> 00:12:00,679
Speaker 1: and the and the job of our sub committee was

232
00:12:00,720 --> 00:12:03,640
Speaker 1: to define high frequency trading, and we really didn't come

233
00:12:03,679 --> 00:12:05,679
Speaker 1: up with a very good definition. I actually dissented to

234
00:12:05,720 --> 00:12:07,439
Speaker 1: a public and when they finally came up with it,

235
00:12:07,520 --> 00:12:10,080
Speaker 1: because it's really hard to define. It's it's it's not

236
00:12:10,200 --> 00:12:13,360
Speaker 1: exactly you know, it's certainly obviously using speed and using

237
00:12:13,440 --> 00:12:16,040
Speaker 1: technology to you know, it's and we think it's certainly

238
00:12:16,240 --> 00:12:18,480
Speaker 1: a matter of when it comes to inventory and net

239
00:12:18,520 --> 00:12:21,200
Speaker 1: positions being close to zero at all times. They're not

240
00:12:21,280 --> 00:12:24,400
Speaker 1: holding huge positions, although they could be long one particular

241
00:12:24,559 --> 00:12:27,000
Speaker 1: options contract and short the stock, or a long an

242
00:12:27,040 --> 00:12:29,240
Speaker 1: ETF and short the future. But at the end of

243
00:12:29,280 --> 00:12:31,640
Speaker 1: the day, the net kind of neutral. They're not carrying

244
00:12:31,720 --> 00:12:35,280
Speaker 1: any inventory. They're in and out very very quickly, very

245
00:12:35,280 --> 00:12:38,240
Speaker 1: short holding periods exactly well, and now some of them

246
00:12:38,240 --> 00:12:40,480
Speaker 1: will hold and and actually now we're starting to see

247
00:12:40,520 --> 00:12:42,480
Speaker 1: some public statements. There's an s one out there from

248
00:12:42,480 --> 00:12:45,679
Speaker 1: one particular HIPT. They're holding positions, but they're usually being

249
00:12:45,720 --> 00:12:48,440
Speaker 1: netted out against others. And we don't necessarily have a

250
00:12:48,480 --> 00:12:52,000
Speaker 1: problem with a high frequency trading market maker. I don't

251
00:12:52,040 --> 00:12:53,960
Speaker 1: have a problem if there I want to trade with

252
00:12:54,040 --> 00:12:57,079
Speaker 1: any participant in a market, in a pool that's all

253
00:12:57,080 --> 00:12:59,520
Speaker 1: trading on a transparent and fair way. Why wouldn't I

254
00:12:59,520 --> 00:13:02,040
Speaker 1: want to trade? Diversity is a good thing. We take

255
00:13:02,160 --> 00:13:04,560
Speaker 1: issue with some of the high frequency traders who are

256
00:13:04,600 --> 00:13:07,880
Speaker 1: more when there's lots of pinging and sniffing, and and

257
00:13:07,920 --> 00:13:09,959
Speaker 1: we can get into exactly how this goes. So let

258
00:13:09,960 --> 00:13:12,839
Speaker 1: me ask you a series of questions, or rather than that,

259
00:13:13,040 --> 00:13:15,320
Speaker 1: I'm gonna give you one word and I want you

260
00:13:15,360 --> 00:13:17,960
Speaker 1: to define it. And I have about ten of these.

261
00:13:18,360 --> 00:13:21,120
Speaker 1: We'll see how far these go. Let's start with co

262
00:13:21,360 --> 00:13:24,880
Speaker 1: located servers. A co location is you can put your

263
00:13:24,920 --> 00:13:26,560
Speaker 1: server at the New York Let's use the New York

264
00:13:26,559 --> 00:13:29,280
Speaker 1: Stock Exchange facility right next to the main matching engine,

265
00:13:29,280 --> 00:13:31,840
Speaker 1: which sits in Mahwah, New Jersey, not on you with

266
00:13:31,960 --> 00:13:33,480
Speaker 1: you know, the stock change is not on Wall Street,

267
00:13:33,600 --> 00:13:36,880
Speaker 1: it's in Mahwah, New Jersey. You can rent a cabinet

268
00:13:36,880 --> 00:13:40,040
Speaker 1: they call it a cage and put your server in there.

269
00:13:40,200 --> 00:13:42,520
Speaker 1: It's got an equal length cord just like everybody else

270
00:13:42,559 --> 00:13:45,480
Speaker 1: in the facility to get to that engine. If the

271
00:13:45,520 --> 00:13:47,480
Speaker 1: trick is to be as close as possible so that

272
00:13:47,600 --> 00:13:50,640
Speaker 1: you can see the data coming at or process the

273
00:13:50,720 --> 00:13:54,280
Speaker 1: data as fast as possible before your competitors a process

274
00:13:54,320 --> 00:13:58,640
Speaker 1: so relative to co location, let's talk about latency arbitrage.

275
00:13:59,080 --> 00:14:01,840
Speaker 1: Great question, Okay, So lazy RB is basically the matter

276
00:14:01,880 --> 00:14:05,120
Speaker 1: of having two separate quotes. What I mean is if

277
00:14:05,120 --> 00:14:08,120
Speaker 1: I was a high speed trader and I can build

278
00:14:08,160 --> 00:14:10,680
Speaker 1: my own quote. So what I do is there's twelve,

279
00:14:10,679 --> 00:14:13,000
Speaker 1: there's eleven stock exchanges soon to be twelve again. But

280
00:14:13,160 --> 00:14:14,480
Speaker 1: what you would do is you would take all those

281
00:14:14,520 --> 00:14:17,360
Speaker 1: stock exchanges, get a direct data feed into them, co

282
00:14:17,480 --> 00:14:20,800
Speaker 1: locate your computers and basically take out any sort of

283
00:14:20,880 --> 00:14:24,040
Speaker 1: latency or time to little delay and create a quote.

284
00:14:24,040 --> 00:14:26,560
Speaker 1: Will create an NBBO, a national best bit or offer

285
00:14:26,840 --> 00:14:29,960
Speaker 1: that is most likely going to be faster than what

286
00:14:30,040 --> 00:14:32,520
Speaker 1: the public sees because they see it through something called

287
00:14:32,520 --> 00:14:35,840
Speaker 1: a SIP or a security information processor. So there's a

288
00:14:35,880 --> 00:14:38,240
Speaker 1: difference in time, and there will always be a difference

289
00:14:38,240 --> 00:14:41,040
Speaker 1: in time because the SIP has to aggregate all of

290
00:14:41,040 --> 00:14:44,720
Speaker 1: those various exchanges or eleven or twelve and then show it,

291
00:14:45,080 --> 00:14:46,920
Speaker 1: whereas if you're just doing one, you could be a

292
00:14:46,920 --> 00:14:49,280
Speaker 1: little well, you're doing your own and you're picking and choosing,

293
00:14:49,360 --> 00:14:51,480
Speaker 1: and you're creating your own quote. And the problem with

294
00:14:51,520 --> 00:14:55,440
Speaker 1: the SIP is that NASDAC, the for profit exchange, runs

295
00:14:55,480 --> 00:14:57,320
Speaker 1: the SIP. They're actually the provider of the SIP for

296
00:14:57,360 --> 00:14:59,000
Speaker 1: a couple of years. Last year I had a little

297
00:14:59,000 --> 00:15:01,640
Speaker 1: problem with the SIP, and the question became, are they

298
00:15:01,680 --> 00:15:05,040
Speaker 1: spending enough money upgrading the technology so that the SIP

299
00:15:05,080 --> 00:15:07,640
Speaker 1: is as fast as everyone else? And currently it used

300
00:15:07,640 --> 00:15:10,080
Speaker 1: to be around a millisecond time delay. Now it's at

301
00:15:10,080 --> 00:15:12,120
Speaker 1: a half of the millisecond. It sounds like nothing, right,

302
00:15:12,280 --> 00:15:14,480
Speaker 1: but a half of a millisecond, So I Frequency Trader

303
00:15:14,560 --> 00:15:16,520
Speaker 1: was all the time in the world. We'll talk a

304
00:15:16,520 --> 00:15:19,800
Speaker 1: little bit about flashboys later, but the book begins with

305
00:15:20,320 --> 00:15:23,840
Speaker 1: the story of creating a fiber op to cable in

306
00:15:23,960 --> 00:15:28,120
Speaker 1: a perfectly straight line from New York to Chicago in

307
00:15:28,240 --> 00:15:31,000
Speaker 1: order to shave a couple of minutes, literally a couple

308
00:15:31,000 --> 00:15:33,960
Speaker 1: of milliseconds off trading. And the guys who did that

309
00:15:34,360 --> 00:15:36,560
Speaker 1: spent a billion dollars on it, and it was an

310
00:15:36,680 --> 00:15:40,880
Speaker 1: enormous return on that investment. Let's keep going. What is

311
00:15:41,160 --> 00:15:44,520
Speaker 1: quote stuffing? Quote stuffing is a theory. It's never been

312
00:15:44,520 --> 00:15:46,520
Speaker 1: proven at the SEC, but it certainly does seem to

313
00:15:46,680 --> 00:15:49,480
Speaker 1: exist what we've seen evidence of it. Nan X by

314
00:15:49,520 --> 00:15:52,720
Speaker 1: the way, he puts out these fabulous graphics showing all

315
00:15:52,760 --> 00:15:56,000
Speaker 1: these giant surges of quotes that show up for a

316
00:15:56,080 --> 00:15:59,440
Speaker 1: millisecond and then disappear. Incredible work that NanoX eric over

317
00:15:59,440 --> 00:16:01,520
Speaker 1: And now I mean he could be he should be

318
00:16:01,560 --> 00:16:03,480
Speaker 1: working for the SEC. That the amount of data that

319
00:16:03,520 --> 00:16:06,560
Speaker 1: he processes, things that he sees, it's amazing that the

320
00:16:06,600 --> 00:16:08,800
Speaker 1: folks in that the SEC or the other regulators don't

321
00:16:08,800 --> 00:16:10,800
Speaker 1: see this, which kind of scares us a bit. But

322
00:16:10,880 --> 00:16:13,440
Speaker 1: quote stuffing is that if stocks trade on what they

323
00:16:13,480 --> 00:16:16,440
Speaker 1: call channels, is maybe there's eight is eight D in

324
00:16:16,440 --> 00:16:19,880
Speaker 1: one channel e F and another. If if you're looking

325
00:16:19,920 --> 00:16:21,880
Speaker 1: at stock in the A channel, you can kind of

326
00:16:21,880 --> 00:16:24,120
Speaker 1: go after another stock send through a whole bunch of

327
00:16:24,200 --> 00:16:26,600
Speaker 1: quotes and then kind of extract out that to make

328
00:16:26,640 --> 00:16:30,240
Speaker 1: your processing faster. It's an interesting theory and it's certainly

329
00:16:30,280 --> 00:16:32,520
Speaker 1: that it's been proven. It's actually been talked about in

330
00:16:32,560 --> 00:16:35,240
Speaker 1: some regulatory documents as well. I'm Barry rid Helts. She

331
00:16:35,320 --> 00:16:38,400
Speaker 1: listening to Masters in Business on Bloomberg Radio. My guest

332
00:16:38,440 --> 00:16:41,360
Speaker 1: today is Joe Saluzzi. He is one of the partners

333
00:16:41,360 --> 00:16:45,240
Speaker 1: of Famous Trading and the co author of Broken Markets

334
00:16:45,360 --> 00:16:49,680
Speaker 1: and a long standing critic of market structures, high frequency

335
00:16:49,760 --> 00:16:53,000
Speaker 1: trading and some of the other things that we discuss

336
00:16:53,120 --> 00:16:57,880
Speaker 1: that works uh to the investors disadvantage. Earlier we were

337
00:16:57,880 --> 00:17:01,320
Speaker 1: discussing um things like co okay ation and packet sniffing.

338
00:17:01,680 --> 00:17:03,760
Speaker 1: Let's talk a little bit before we get to the

339
00:17:03,800 --> 00:17:08,000
Speaker 1: flash crash. Talk a little bit about spoofing and walking away.

340
00:17:08,359 --> 00:17:11,840
Speaker 1: What what is spoofing. Spoofing is placing bids are offers

341
00:17:11,960 --> 00:17:14,200
Speaker 1: where you have no intention of trading them, and there's

342
00:17:14,200 --> 00:17:17,520
Speaker 1: actually been a couple of cases just recently. The CFTC has,

343
00:17:17,600 --> 00:17:19,879
Speaker 1: by the way, has much more the Commodity Futures and

344
00:17:19,880 --> 00:17:22,520
Speaker 1: Trading Commission much more power now under Dodd Frank to

345
00:17:22,560 --> 00:17:25,200
Speaker 1: go after spoofers than the SEC does. It's kind of

346
00:17:25,200 --> 00:17:27,200
Speaker 1: a weird thing where how it was written in the rules,

347
00:17:27,200 --> 00:17:29,720
Speaker 1: But they've come after a couple of guys recently and

348
00:17:29,760 --> 00:17:32,440
Speaker 1: have settled a few cases where folks or these firms

349
00:17:32,520 --> 00:17:35,399
Speaker 1: high speed firms, because you really you could manually spoof,

350
00:17:35,400 --> 00:17:37,160
Speaker 1: but it's really not going to be effective. But they

351
00:17:37,160 --> 00:17:39,639
Speaker 1: were basically placed layering the book really, which is another

352
00:17:39,720 --> 00:17:41,879
Speaker 1: term layers, so at different prices, a whole bunch of

353
00:17:41,960 --> 00:17:44,240
Speaker 1: orders that look like there's a big buy or look

354
00:17:44,320 --> 00:17:46,440
Speaker 1: like there's a big seller to move the price, and

355
00:17:46,480 --> 00:17:48,440
Speaker 1: then of course as soon as someone comes in they

356
00:17:48,480 --> 00:17:51,359
Speaker 1: cancel all those orders gone. There's never an intention, and

357
00:17:51,440 --> 00:17:53,600
Speaker 1: that's the thing, and how do they prove that. It's

358
00:17:53,600 --> 00:17:56,439
Speaker 1: an intense se scenario. And the CFCC again has a

359
00:17:56,440 --> 00:17:58,800
Speaker 1: lot more leeway. So if you're a few, if you're

360
00:17:58,800 --> 00:18:00,760
Speaker 1: a guy in the in the future, his world, and

361
00:18:00,760 --> 00:18:02,640
Speaker 1: you're spoofing, you should be a little scared now because

362
00:18:02,640 --> 00:18:05,880
Speaker 1: the CFCC has a much much more powerful regular relation

363
00:18:05,920 --> 00:18:07,720
Speaker 1: that they can use against you. I mean, isn't that

364
00:18:07,760 --> 00:18:10,840
Speaker 1: really an attempt to manipulate markets with no intention to

365
00:18:11,400 --> 00:18:14,240
Speaker 1: actually execute a trade? Yes, And that's exactly one of

366
00:18:14,320 --> 00:18:16,800
Speaker 1: the problems that we have with certain high speed trading

367
00:18:16,880 --> 00:18:19,120
Speaker 1: or any trader for that matter. You know, we're we're

368
00:18:19,160 --> 00:18:22,040
Speaker 1: critics of of potential problems in the market of market structure.

369
00:18:22,240 --> 00:18:24,919
Speaker 1: Spoofing has no business being in the market, right, So

370
00:18:25,080 --> 00:18:27,800
Speaker 1: here's my proposal. Tell me what the result of this is.

371
00:18:28,000 --> 00:18:29,600
Speaker 1: If you want to put a bit of a quote

372
00:18:29,600 --> 00:18:32,040
Speaker 1: out there, you have to have it stay there for

373
00:18:32,080 --> 00:18:35,280
Speaker 1: a tenth of a second? Is am I being ridiculous

374
00:18:35,320 --> 00:18:37,640
Speaker 1: or what's the know? And and we actually had recommended

375
00:18:37,640 --> 00:18:39,800
Speaker 1: that once also, But the the high speed community the

376
00:18:39,840 --> 00:18:41,399
Speaker 1: other side will say that that's going to mess up

377
00:18:41,400 --> 00:18:44,560
Speaker 1: their models. They can't place bids and offers and therefore

378
00:18:44,600 --> 00:18:47,959
Speaker 1: they can't arbitrage certain scenarios. Let's let's say it's an

379
00:18:47,960 --> 00:18:50,119
Speaker 1: et F arbitrage ore and he's gonna under you know,

380
00:18:50,119 --> 00:18:52,280
Speaker 1: he's gonna trade five hundred stocks against the SMP five

381
00:18:52,359 --> 00:18:55,560
Speaker 1: hundreds if he's forced to place a bitter offer for

382
00:18:55,600 --> 00:18:57,399
Speaker 1: a tenth of a second, which you know, I'm blinking

383
00:18:57,400 --> 00:19:00,840
Speaker 1: our eyes in ridiculous. You know, we certainly wouldn't have

384
00:19:00,920 --> 00:19:03,359
Speaker 1: a problem with that. They would scream bloody murder. And

385
00:19:03,359 --> 00:19:05,080
Speaker 1: then here's the flip side. They're gonna say, well, if

386
00:19:05,080 --> 00:19:07,480
Speaker 1: you do that, I'm gonna take away my quote unquote

387
00:19:07,480 --> 00:19:10,080
Speaker 1: liquidity and I'll be gone. And that's where everyone gets scared.

388
00:19:10,160 --> 00:19:13,040
Speaker 1: So now let's let's talk about walking away, which is

389
00:19:13,040 --> 00:19:15,880
Speaker 1: going to lead us to ultimately to the flash crash.

390
00:19:16,000 --> 00:19:18,560
Speaker 1: You know, when you had a specialist down on the exchange,

391
00:19:19,040 --> 00:19:23,680
Speaker 1: they were there theoretically to provide an orderly market. They

392
00:19:23,680 --> 00:19:26,400
Speaker 1: were ready, willing and able to to buy or sell

393
00:19:26,480 --> 00:19:30,560
Speaker 1: stocks at an equilibrium price. Uh. The counter argument, as

394
00:19:30,600 --> 00:19:35,359
Speaker 1: you go back to and hey down one day, I

395
00:19:35,400 --> 00:19:38,159
Speaker 1: think that was a different era. You know, everything was

396
00:19:38,200 --> 00:19:43,719
Speaker 1: fairly manual back then. But that taking away the liquidly Uh.

397
00:19:44,520 --> 00:19:48,040
Speaker 1: Don't they have essentially no obligation to make a market?

398
00:19:48,080 --> 00:19:51,600
Speaker 1: Can't they just walk away anytime they want? Spot on exactly?

399
00:19:51,720 --> 00:19:55,679
Speaker 1: And that's the problem. Obligations are nowhere to be found

400
00:19:55,720 --> 00:19:58,760
Speaker 1: in the market making community anymore, although they will tell you.

401
00:19:58,880 --> 00:20:00,720
Speaker 1: I mean some of the exchanges have an obligation of

402
00:20:01,000 --> 00:20:03,240
Speaker 1: eight percent within the n b b O, which means

403
00:20:03,240 --> 00:20:05,080
Speaker 1: you can, quote, for a want to wait on a

404
00:20:05,119 --> 00:20:07,679
Speaker 1: hundred all of stock, Thank you very much, I'll take you. Know,

405
00:20:07,680 --> 00:20:09,639
Speaker 1: you can have that liquidity, But if you don't have

406
00:20:09,680 --> 00:20:13,440
Speaker 1: an oblimy. That's a touch more than the flash crash. Well,

407
00:20:13,440 --> 00:20:15,600
Speaker 1: we had stub quotes then they were trading mid zero, right,

408
00:20:15,840 --> 00:20:18,280
Speaker 1: that was how crazy was that? But you know, people

409
00:20:18,280 --> 00:20:20,119
Speaker 1: will say the specialist system, they were problems and they

410
00:20:20,200 --> 00:20:23,000
Speaker 1: land them with handcuffs and all. If there were problems,

411
00:20:23,000 --> 00:20:24,840
Speaker 1: and they were, we would call those out if we

412
00:20:24,840 --> 00:20:26,359
Speaker 1: were back in the nineties. Right now, sal and I

413
00:20:26,400 --> 00:20:28,560
Speaker 1: will be talking about problems back in the specialist world,

414
00:20:28,680 --> 00:20:32,000
Speaker 1: because that doesn't get rid of a hold of these problems. Right,

415
00:20:32,000 --> 00:20:34,440
Speaker 1: But let's think about it. At least the specialists had

416
00:20:34,480 --> 00:20:37,200
Speaker 1: what they call an affirmative and a negative obligation, which

417
00:20:37,240 --> 00:20:39,119
Speaker 1: means they had to buy stock at certain times and

418
00:20:39,160 --> 00:20:42,399
Speaker 1: they couldn't sell its other times. And the customer knew that,

419
00:20:42,400 --> 00:20:46,760
Speaker 1: and there was there was the confidence. So let's let's

420
00:20:46,800 --> 00:20:49,720
Speaker 1: talk about the flash crash. May six, two thou and ten.

421
00:20:50,080 --> 00:20:53,199
Speaker 1: All of a sudden, the down plummets. I remember this

422
00:20:53,200 --> 00:20:55,040
Speaker 1: because I was on a flight back from Dallas. I

423
00:20:55,080 --> 00:20:57,800
Speaker 1: missed it. But and every time I fly now people

424
00:20:58,240 --> 00:21:01,880
Speaker 1: point out that the market gets color crush, But um,

425
00:21:02,320 --> 00:21:04,919
Speaker 1: what happened. What happened on that day? That was an

426
00:21:05,119 --> 00:21:08,320
Speaker 1: evaporation in liquidity basically, right. And let me read you

427
00:21:08,640 --> 00:21:11,040
Speaker 1: just a real quick quote from after the flash. Crasty

428
00:21:11,080 --> 00:21:14,399
Speaker 1: sec CFTC Advisory Committee put out a report which we

429
00:21:14,440 --> 00:21:16,080
Speaker 1: thought was one of the best pieces that ever came

430
00:21:16,080 --> 00:21:18,760
Speaker 1: out of DC. Listen to these words they said talking

431
00:21:18,800 --> 00:21:21,879
Speaker 1: about the flash crash. Indeed, even in the absence of

432
00:21:21,920 --> 00:21:25,679
Speaker 1: extraordinary market events, limit order books can quickly empty and

433
00:21:25,720 --> 00:21:28,760
Speaker 1: prices can crash simply due to the speed and numbers

434
00:21:28,800 --> 00:21:30,879
Speaker 1: of orders flowing into the market and due to the

435
00:21:30,920 --> 00:21:35,000
Speaker 1: ability to instantly cancel orders. So the books emptied. There

436
00:21:35,080 --> 00:21:37,159
Speaker 1: was a situation, there was grease going on like there

437
00:21:37,240 --> 00:21:39,359
Speaker 1: is today, right, same, sorry. And then there was a

438
00:21:39,440 --> 00:21:42,840
Speaker 1: supposedly a large e mini order in the futures market.

439
00:21:42,920 --> 00:21:46,040
Speaker 1: Not it was an order, right, and by the way,

440
00:21:46,080 --> 00:21:49,040
Speaker 1: it was a seventy contract order, which half of it

441
00:21:49,080 --> 00:21:52,480
Speaker 1: wasn't even executed until after this was over, So it

442
00:21:52,520 --> 00:21:55,359
Speaker 1: really is kind of an it's a bad excuse, and

443
00:21:55,400 --> 00:21:57,720
Speaker 1: it wasn't the cause. The cause was the market was

444
00:21:57,760 --> 00:22:00,639
Speaker 1: on pins and needles. The bids and offers just evaporated.

445
00:22:00,680 --> 00:22:02,800
Speaker 1: And there was a professor, actually he was at the

446
00:22:02,800 --> 00:22:05,320
Speaker 1: CFCC that his name is Andre Carolinko. He turned the

447
00:22:05,400 --> 00:22:08,720
Speaker 1: phrase hot potato trading where they basically the h f

448
00:22:08,800 --> 00:22:11,080
Speaker 1: T guys started flipping back and forth between each other

449
00:22:11,240 --> 00:22:14,080
Speaker 1: until their position limits got exhausted, and then they just

450
00:22:14,119 --> 00:22:15,919
Speaker 1: walked away. And there was even one in the market

451
00:22:15,960 --> 00:22:18,480
Speaker 1: that said he hit the button h f T stop,

452
00:22:18,520 --> 00:22:21,280
Speaker 1: which meant get me out. Everything was done. Everybody just

453
00:22:21,320 --> 00:22:23,960
Speaker 1: walked away. I'm Barry rit Helps. You're listening to Masters

454
00:22:23,960 --> 00:22:27,840
Speaker 1: in Business on Bloomberg Radio. My guest today is Joe Saluzzi.

455
00:22:28,000 --> 00:22:31,119
Speaker 1: He is the co author of Broken Markets with his

456
00:22:31,200 --> 00:22:34,480
Speaker 1: partner Sal Arnuck. The two of them run Famous Trading,

457
00:22:34,880 --> 00:22:39,959
Speaker 1: and we were previously discussing the causes of the flash crash,

458
00:22:40,080 --> 00:22:43,200
Speaker 1: so let's let's go back to that. So it's May

459
00:22:43,240 --> 00:22:47,520
Speaker 1: two thousand and ten, and suddenly stocks that normally trade

460
00:22:47,560 --> 00:22:52,600
Speaker 1: at eight are a penny. The Dow loses some crazy

461
00:22:53,359 --> 00:22:55,199
Speaker 1: I almost want to say ten percent. Is that a

462
00:22:55,200 --> 00:22:58,840
Speaker 1: bullpark number? In a matter of minutes and a half

463
00:22:58,920 --> 00:23:01,679
Speaker 1: hour later, everything come snapping back as if it was

464
00:23:02,400 --> 00:23:05,800
Speaker 1: no big deal. How on earth does that happen in

465
00:23:05,840 --> 00:23:08,520
Speaker 1: the twenty one century in America? Yeah, I think they're

466
00:23:08,520 --> 00:23:12,040
Speaker 1: still looking into that, the regulators, right, But it happened,

467
00:23:12,200 --> 00:23:14,040
Speaker 1: you know. But basically the problem a lot of the

468
00:23:14,080 --> 00:23:17,280
Speaker 1: retail investor's face was folks who had stop loss orders

469
00:23:17,320 --> 00:23:19,200
Speaker 1: in there. If a stock got to a certain level,

470
00:23:19,240 --> 00:23:21,760
Speaker 1: they wanted to sell it. Those stop losses became market

471
00:23:21,840 --> 00:23:24,520
Speaker 1: orders once they activate, well they really what happens with

472
00:23:24,560 --> 00:23:27,080
Speaker 1: a market order. Normally it'll get sent over to what

473
00:23:27,119 --> 00:23:30,160
Speaker 1: they call an internalizing broker. They will normally pick up

474
00:23:30,160 --> 00:23:32,639
Speaker 1: this market order. They'll trade against it because they consider

475
00:23:32,720 --> 00:23:36,000
Speaker 1: retail flow dumb. I'm sorry, that's that's their word's not mine,

476
00:23:36,400 --> 00:23:38,920
Speaker 1: and they want to basically give it a subpenny price improvement,

477
00:23:39,000 --> 00:23:40,919
Speaker 1: and then they're off they and they're trying to trade against.

478
00:23:41,160 --> 00:23:44,680
Speaker 1: In this scenario, the internalizers tend they walked away as well.

479
00:23:44,760 --> 00:23:47,879
Speaker 1: So these market orders got exhausted to the let exchanges

480
00:23:47,920 --> 00:23:50,119
Speaker 1: over New York and to the NASDAC and there was

481
00:23:50,160 --> 00:23:52,359
Speaker 1: no bids or offers there. So what does a market

482
00:23:52,440 --> 00:23:54,159
Speaker 1: order do. It goes down and starts looking for the

483
00:23:54,160 --> 00:23:57,120
Speaker 1: best bit or offer until it finds something. And even

484
00:23:58,160 --> 00:24:01,400
Speaker 1: and even at selling Johnson and John's in it, Accentura

485
00:24:01,520 --> 00:24:03,160
Speaker 1: was traded at a penny of share, so I think

486
00:24:03,200 --> 00:24:05,400
Speaker 1: was Sam Adams too, which is whatever's stock price now

487
00:24:05,640 --> 00:24:07,720
Speaker 1: a penny of share. It was absurd and it was

488
00:24:07,760 --> 00:24:10,159
Speaker 1: basically it was called a stub quote, where a market

489
00:24:10,200 --> 00:24:12,560
Speaker 1: maker could place a one sided quote. So he'll put

490
00:24:12,560 --> 00:24:14,840
Speaker 1: a penny bid offered at a hundred dollars whatever it

491
00:24:14,880 --> 00:24:17,080
Speaker 1: may be. Say, you know, the SEC has since bann

492
00:24:17,119 --> 00:24:19,240
Speaker 1: stub quotes. But that wasn't the problem. The problem was

493
00:24:19,400 --> 00:24:21,159
Speaker 1: why did that market or to come into the system

494
00:24:21,160 --> 00:24:23,400
Speaker 1: and have no bids or offers. Why did the limit

495
00:24:23,520 --> 00:24:26,320
Speaker 1: order books that we talked about before just disappear? Well,

496
00:24:26,359 --> 00:24:28,840
Speaker 1: it's not really real liquidity. It was kind of a

497
00:24:28,840 --> 00:24:31,600
Speaker 1: phantom liquidity. It's there when you don't need it, and

498
00:24:31,640 --> 00:24:34,800
Speaker 1: when you really need it it disappears. That's the big crew.

499
00:24:34,960 --> 00:24:37,280
Speaker 1: That's one of my big complaints bearing him on sooon

500
00:24:37,320 --> 00:24:40,679
Speaker 1: and disappears in the draft. So so so, there are

501
00:24:40,720 --> 00:24:43,399
Speaker 1: some people who I've actually had on the show and

502
00:24:43,440 --> 00:24:47,440
Speaker 1: had conversations with. Jack Brennan is chairman of the Vanguard Group.

503
00:24:47,520 --> 00:24:50,399
Speaker 1: Cliff Astness, the c i O of of a q R,

504
00:24:50,480 --> 00:24:54,920
Speaker 1: who have said high frequency trading and the modern competition

505
00:24:54,960 --> 00:24:59,800
Speaker 1: have made costs of execution lower. They've tightened spreads, they've

506
00:24:59,840 --> 00:25:02,399
Speaker 1: done the number of things, although the reality is, you know,

507
00:25:02,440 --> 00:25:04,480
Speaker 1: the spread is tightened, but it's for forty seven shares.

508
00:25:04,520 --> 00:25:06,679
Speaker 1: If you have to buy five thousand shares, is the

509
00:25:06,680 --> 00:25:10,080
Speaker 1: spread is still ten or twenty or thirty cents. But

510
00:25:10,280 --> 00:25:13,560
Speaker 1: when when the chairman of Vanguard said, hey, this has

511
00:25:13,560 --> 00:25:16,600
Speaker 1: made our costs a lot lower, how do we argue

512
00:25:16,640 --> 00:25:19,280
Speaker 1: against that? Well, that's wonderful for Vanguard, who makes money

513
00:25:19,359 --> 00:25:21,480
Speaker 1: on you know how many trillions of ETFs they have

514
00:25:21,640 --> 00:25:25,479
Speaker 1: right now, three trillion dollars. It's you know, I wouldn't

515
00:25:25,480 --> 00:25:27,760
Speaker 1: want local, you know. But for them, it's a different

516
00:25:27,760 --> 00:25:30,480
Speaker 1: business model, right they're running something and for a quant

517
00:25:30,520 --> 00:25:32,480
Speaker 1: it's a different business mode because they're flipping in and

518
00:25:32,520 --> 00:25:36,080
Speaker 1: out of numerous, numerous positions throughout the day. But we're

519
00:25:36,080 --> 00:25:39,840
Speaker 1: talking about a traditional institutional and investor who's trying to

520
00:25:39,840 --> 00:25:41,760
Speaker 1: buy a hundred thousand shares of a stock that trades

521
00:25:41,800 --> 00:25:45,240
Speaker 1: a million shares. Things have changed dramatically for them because

522
00:25:45,240 --> 00:25:48,760
Speaker 1: that footprint they're leaving is basically the one that everyone's

523
00:25:48,760 --> 00:25:52,040
Speaker 1: trying to pick off. That liquidity that's there disappears quickly.

524
00:25:52,080 --> 00:25:55,040
Speaker 1: The obligations of a market maker are no longer. They're

525
00:25:55,280 --> 00:25:58,879
Speaker 1: the average trade size now on both lit and dark venues.

526
00:25:59,000 --> 00:26:02,520
Speaker 1: Is less than two hundred chairs, less than two Trade says,

527
00:26:02,560 --> 00:26:04,960
Speaker 1: it's incredible. How do you get anything done? So when

528
00:26:04,960 --> 00:26:08,679
Speaker 1: we look at the flash crash, what's to prevent that

529
00:26:08,800 --> 00:26:13,159
Speaker 1: from happening again? Nothing? Okay? What about that on a

530
00:26:13,359 --> 00:26:15,879
Speaker 1: D do stop? Well, you won't have the O and

531
00:26:16,080 --> 00:26:18,520
Speaker 1: U and A ninety two. So in other words, they'll disappear.

532
00:26:18,560 --> 00:26:20,399
Speaker 1: But they don't have to be there. They don't have

533
00:26:20,440 --> 00:26:22,000
Speaker 1: to know that well they would, you know, to be

534
00:26:22,119 --> 00:26:23,760
Speaker 1: considered a market maker, you have to be there. But

535
00:26:23,760 --> 00:26:25,080
Speaker 1: if you can easily just shut it off and say

536
00:26:25,200 --> 00:26:26,560
Speaker 1: I'm not going to be a market maker right now,

537
00:26:26,840 --> 00:26:28,719
Speaker 1: but get to do that for the day, all right,

538
00:26:30,440 --> 00:26:32,240
Speaker 1: And that means that the exchange won't give you the

539
00:26:32,320 --> 00:26:34,879
Speaker 1: special pricing that they give to these market makers. But

540
00:26:34,960 --> 00:26:38,360
Speaker 1: just as an example, October fifteenth of last year, there

541
00:26:38,400 --> 00:26:40,960
Speaker 1: was a flash crash in the Treasury mark Remember, okay

542
00:26:41,040 --> 00:26:44,040
Speaker 1: that that day should be on everybody's on the blackboard,

543
00:26:44,040 --> 00:26:46,680
Speaker 1: just like May six was. That was a very very

544
00:26:46,680 --> 00:26:48,760
Speaker 1: disturbing event because that's the most liquid acid in the

545
00:26:48,760 --> 00:26:51,639
Speaker 1: world and it crashed in a matter of minutes and

546
00:26:51,640 --> 00:26:54,040
Speaker 1: then it's zapped right back again. Just like the equity mark.

547
00:26:54,200 --> 00:26:56,920
Speaker 1: Bigger move was the yields are from my two tents

548
00:26:56,920 --> 00:27:00,840
Speaker 1: and one. I think it was huge in minutes and

549
00:27:00,880 --> 00:27:03,040
Speaker 1: apparently no, we know it's a Ukraine situation. Whatever. It

550
00:27:03,080 --> 00:27:05,879
Speaker 1: was the same scenario, But why did the treasury market

551
00:27:05,880 --> 00:27:08,080
Speaker 1: now that it's a known fact that high frequency traders

552
00:27:08,080 --> 00:27:10,760
Speaker 1: are now in the treasury market. Electronic trading has taken

553
00:27:10,800 --> 00:27:13,840
Speaker 1: over the treasury market. So it's the same scenario that,

554
00:27:14,080 --> 00:27:16,959
Speaker 1: like I read before, limit order books can quickly evaporate,

555
00:27:17,200 --> 00:27:19,080
Speaker 1: and that's what we're dealing with. So our our issues.

556
00:27:19,200 --> 00:27:21,119
Speaker 1: How do we build those limit order books again? How

557
00:27:21,119 --> 00:27:23,760
Speaker 1: do we put diverse liquidity back into there where people

558
00:27:23,800 --> 00:27:26,719
Speaker 1: feel comfortable that they're not gonna get picked off all day?

559
00:27:26,760 --> 00:27:31,119
Speaker 1: So you're sided. In the book Flash Boys, which was

560
00:27:31,280 --> 00:27:34,880
Speaker 1: Michael Lewis's book, I actually read it on vacation last year,

561
00:27:34,920 --> 00:27:37,760
Speaker 1: and his books are perfect beach reading. I find them

562
00:27:38,119 --> 00:27:42,120
Speaker 1: both entertaining and and informative. And we'll have to get

563
00:27:42,160 --> 00:27:45,320
Speaker 1: Michael on the show one of these days. But you

564
00:27:45,359 --> 00:27:47,520
Speaker 1: guys are mentioned in there. Tell us a little bit

565
00:27:47,560 --> 00:27:50,399
Speaker 1: about what took place in that book and what some

566
00:27:50,520 --> 00:27:55,040
Speaker 1: of the solutions to high frequency trading are coming down

567
00:27:55,040 --> 00:27:58,399
Speaker 1: the pipe because we've been hearing h f T is

568
00:27:58,480 --> 00:28:00,399
Speaker 1: less profitable than it used to be. Either, there's a

569
00:28:00,440 --> 00:28:04,560
Speaker 1: ton of competition. The rule changes have affected them somewhat.

570
00:28:04,840 --> 00:28:07,320
Speaker 1: What what's happening now? Well, the book was usually the

571
00:28:07,320 --> 00:28:10,560
Speaker 1: book was tremendous just to get to debate really moving forward.

572
00:28:10,640 --> 00:28:12,760
Speaker 1: And we've been struggling with this since oh eight, right,

573
00:28:12,800 --> 00:28:14,200
Speaker 1: so two thousand and eight we were at our first

574
00:28:14,200 --> 00:28:15,879
Speaker 1: white paper. We've been you know, we were on sixty

575
00:28:15,920 --> 00:28:17,760
Speaker 1: minutes and two thousand intent. It's been a long time,

576
00:28:17,760 --> 00:28:19,920
Speaker 1: a long struggle. Michael Lewis came in like a steam

577
00:28:20,000 --> 00:28:22,879
Speaker 1: roller and everyone is talking about which is fantastic, but

578
00:28:22,960 --> 00:28:25,040
Speaker 1: you know, he used the word rigged, which really got

579
00:28:25,040 --> 00:28:28,560
Speaker 1: people from sixty minutes. Michael Lewis said, the markets are rigged.

580
00:28:29,000 --> 00:28:32,520
Speaker 1: Do you agree with that? The structure of the market

581
00:28:33,160 --> 00:28:37,160
Speaker 1: is ragged. Okay, they're they're are forces. The stock exchange

582
00:28:37,200 --> 00:28:40,600
Speaker 1: model is broken there, they're poor profit companies that have

583
00:28:40,680 --> 00:28:43,760
Speaker 1: different incentives than they should have now. They used, for

584
00:28:44,080 --> 00:28:46,120
Speaker 1: people who may not be familiar with this, for most

585
00:28:46,160 --> 00:28:49,200
Speaker 1: of their histories, The New York Stock Exchange was a

586
00:28:49,360 --> 00:28:54,920
Speaker 1: nonprofit entity up until the whole situation with their former

587
00:28:55,000 --> 00:28:58,520
Speaker 1: chairman getting paid bonuses that they wasn't supposed to when

588
00:28:58,560 --> 00:29:02,760
Speaker 1: you're running a nonprofit, that ultimately lead um okay, well,

589
00:29:02,760 --> 00:29:05,080
Speaker 1: if you don't want to take a bonus as a nonprofit,

590
00:29:05,400 --> 00:29:07,800
Speaker 1: let's become for profit company. Right, and it changed their

591
00:29:07,840 --> 00:29:10,240
Speaker 1: philosophy a change. Okay, we're now for the bottom line.

592
00:29:10,240 --> 00:29:12,280
Speaker 1: We have to return money to investors because that's what

593
00:29:12,320 --> 00:29:15,360
Speaker 1: public companies do. Right. NASDAC, for instance, has less than

594
00:29:15,360 --> 00:29:19,040
Speaker 1: ten percent of their money generated from equity cash transactions.

595
00:29:19,600 --> 00:29:22,280
Speaker 1: That means stock market So where's old money coming from?

596
00:29:22,480 --> 00:29:27,160
Speaker 1: Data sales? Various things related to co location data feeds,

597
00:29:27,160 --> 00:29:30,120
Speaker 1: which is a huge profit generator for stock exchanges. So

598
00:29:30,160 --> 00:29:33,360
Speaker 1: they're and even one stock exchange is owned by brokers. Okay,

599
00:29:33,400 --> 00:29:36,920
Speaker 1: there's used huge conflicts of interest in the stock exchange model,

600
00:29:37,080 --> 00:29:39,800
Speaker 1: which has created the problems that we see in what

601
00:29:39,840 --> 00:29:42,040
Speaker 1: they call these smart order routers from the broker level,

602
00:29:42,120 --> 00:29:44,959
Speaker 1: where they go to one exchange first because that offers

603
00:29:45,000 --> 00:29:48,040
Speaker 1: a better rebate than it does from another exchange, or

604
00:29:48,040 --> 00:29:50,479
Speaker 1: they go to a dark pool because it doesn't charge

605
00:29:50,520 --> 00:29:53,840
Speaker 1: them anything. Define dark pools for listeners. Sure, dark pools

606
00:29:53,880 --> 00:29:57,080
Speaker 1: are off exchange venues that don't need to publish their quotes.

607
00:29:57,360 --> 00:29:58,880
Speaker 1: Back in the day, when I was an instant that

608
00:29:58,880 --> 00:30:00,680
Speaker 1: we called it a crossing net work at night because

609
00:30:00,680 --> 00:30:03,240
Speaker 1: we wanted the cross chunks of stock. Now they operate

610
00:30:03,440 --> 00:30:05,480
Speaker 1: throughout the day. And an average trade size on a

611
00:30:05,560 --> 00:30:07,240
Speaker 1: dark pool is no bigger than a lip pool now

612
00:30:07,360 --> 00:30:09,320
Speaker 1: less than two hundred shares. But but back then it

613
00:30:09,400 --> 00:30:12,600
Speaker 1: was big and fifty blocks. And there's still a couple

614
00:30:12,600 --> 00:30:14,960
Speaker 1: of dark pools now liquid that in particular that's still

615
00:30:14,960 --> 00:30:17,600
Speaker 1: over forty shares on the average trade size. Those are

616
00:30:17,680 --> 00:30:21,840
Speaker 1: institutions crossing stock, reducing transaction costs. And some may say, well,

617
00:30:21,880 --> 00:30:24,480
Speaker 1: you're shutting out the retail investor. No you're not, because

618
00:30:24,480 --> 00:30:27,720
Speaker 1: those institutions are trading on behalf of retail fund, that

619
00:30:28,840 --> 00:30:33,920
Speaker 1: retail money in an individual stock exactly. So that argument

620
00:30:33,960 --> 00:30:36,720
Speaker 1: doesn't hold any water. But dark pools themselves have been

621
00:30:36,720 --> 00:30:40,520
Speaker 1: so perverted and so over the years. What the Barkley's

622
00:30:40,560 --> 00:30:42,360
Speaker 1: case that the a g and York Attorney General is

623
00:30:42,400 --> 00:30:44,920
Speaker 1: brought against the Barkley's dark Pool, which is still pending,

624
00:30:45,240 --> 00:30:48,760
Speaker 1: exposed a lot of what actually goes on inside and

625
00:30:48,800 --> 00:30:51,959
Speaker 1: the lack of disclosion, the lack of transparency that the

626
00:30:52,000 --> 00:30:53,960
Speaker 1: clients of the dark pools don't have no idea, They

627
00:30:54,000 --> 00:30:56,280
Speaker 1: have no idea who's getting teared and who gets the

628
00:30:56,360 --> 00:30:59,680
Speaker 1: order first. These are the problems inside the market structure

629
00:30:59,680 --> 00:31:02,320
Speaker 1: when you talk about the term rigged, that do exist,

630
00:31:02,560 --> 00:31:05,800
Speaker 1: that do advantage one class of investor over another, that

631
00:31:05,880 --> 00:31:07,840
Speaker 1: need to be taken out of the market. If we're

632
00:31:07,880 --> 00:31:10,560
Speaker 1: playing on a level playing field, we've got absolutely no

633
00:31:10,600 --> 00:31:13,120
Speaker 1: problem with any high frequency trading low I don't care

634
00:31:13,120 --> 00:31:15,480
Speaker 1: who you are, low frequency guy. I want to interact

635
00:31:15,480 --> 00:31:17,680
Speaker 1: with every piece of liquidity, but I want to do

636
00:31:17,680 --> 00:31:19,760
Speaker 1: it on a fair and level playing field. So let's

637
00:31:19,760 --> 00:31:23,840
Speaker 1: talk about the new exchange i e X, which solves

638
00:31:24,280 --> 00:31:27,320
Speaker 1: some of these problems. Absolutely. We're a trading We were

639
00:31:27,320 --> 00:31:29,120
Speaker 1: one of the first trading partners. You know, we think

640
00:31:29,120 --> 00:31:31,720
Speaker 1: i X. They nailed it with their model. They got

641
00:31:31,800 --> 00:31:34,520
Speaker 1: rid of rebates. Okay, there's no there's a flat fee

642
00:31:34,600 --> 00:31:36,640
Speaker 1: on the I X exchange. You pay what whether you

643
00:31:36,680 --> 00:31:39,560
Speaker 1: make the liquidity or take a liquidity one feet. They

644
00:31:39,600 --> 00:31:41,920
Speaker 1: got rid of the speed advantage by putting in something

645
00:31:42,000 --> 00:31:43,320
Speaker 1: you know, Brad will tell you more than I need

646
00:31:43,360 --> 00:31:47,320
Speaker 1: to buffer. And essentially it's a mile of spooled uh.

647
00:31:47,360 --> 00:31:52,080
Speaker 1: Fiber optics, nobody has any advantage of of speed. The

648
00:31:52,160 --> 00:31:54,280
Speaker 1: order comes in and then it gets buffered for a

649
00:31:54,400 --> 00:31:57,880
Speaker 1: millisecond and now everybody's arriving at the same time. That's right.

650
00:31:57,920 --> 00:32:00,880
Speaker 1: And they took away the gaming that goes on in

651
00:32:00,960 --> 00:32:03,720
Speaker 1: other systems. And they'll tell you about who types of

652
00:32:03,720 --> 00:32:05,400
Speaker 1: clients they have. But a lot of h f T

653
00:32:05,480 --> 00:32:08,320
Speaker 1: guys don't want to go there because it's not profitable.

654
00:32:08,560 --> 00:32:11,840
Speaker 1: But for us, as institutional traders, we love systems like that.

655
00:32:11,880 --> 00:32:13,840
Speaker 1: We'll get our average trade size is much bigger on

656
00:32:13,880 --> 00:32:15,960
Speaker 1: I X than is on any other venue. That that

657
00:32:16,120 --> 00:32:19,120
Speaker 1: is a very very good system. You get clean liquidity,

658
00:32:19,320 --> 00:32:22,120
Speaker 1: you're crossing things up nice. That's what you're supposed to have.

659
00:32:22,200 --> 00:32:24,600
Speaker 1: So the market here's a great example. The market is

660
00:32:24,600 --> 00:32:27,680
Speaker 1: now starting to solve its own problems, wonderful without the

661
00:32:27,720 --> 00:32:30,680
Speaker 1: regulators mucking it up more than they've done already. So

662
00:32:31,160 --> 00:32:35,240
Speaker 1: before we um uh and the radio portion of our interview,

663
00:32:35,360 --> 00:32:37,480
Speaker 1: if people want to read some of your white papers,

664
00:32:37,800 --> 00:32:39,880
Speaker 1: where where can they find out more about this? Sure?

665
00:32:39,880 --> 00:32:42,320
Speaker 1: We run a bog actually femist trading dot com. We

666
00:32:42,440 --> 00:32:44,160
Speaker 1: have a blog that we'll put out maybe three or

667
00:32:44,160 --> 00:32:46,920
Speaker 1: four posts a week that we normally every topical issues

668
00:32:46,920 --> 00:32:51,320
Speaker 1: on market structure. Obviously our book Broken Markets, white papers

669
00:32:51,320 --> 00:32:53,360
Speaker 1: are on the website as well image trading and and

670
00:32:53,480 --> 00:32:55,800
Speaker 1: you know we don't charge for these. We don't put this.

671
00:32:55,800 --> 00:32:57,600
Speaker 1: This is not a business model for us. Well, this

672
00:32:57,640 --> 00:32:59,520
Speaker 1: is this is a passion. This is what we do.

673
00:32:59,560 --> 00:33:01,719
Speaker 1: We invent, stickate things, and then we trade based on

674
00:33:01,720 --> 00:33:05,040
Speaker 1: what we learned. Okay, welcome back to the podcast. If

675
00:33:05,040 --> 00:33:07,960
Speaker 1: you're listening to this, it means you already heard the

676
00:33:08,080 --> 00:33:11,040
Speaker 1: radio portion of our show, and now I get to

677
00:33:11,520 --> 00:33:14,800
Speaker 1: stop worrying about the time segments and actually just relax

678
00:33:14,880 --> 00:33:18,280
Speaker 1: and enjoy chatting with my guests. I know Joe for

679
00:33:18,640 --> 00:33:20,280
Speaker 1: we know each other for a good couple of years.

680
00:33:20,280 --> 00:33:22,160
Speaker 1: I'm trying to figure out where I met you and

681
00:33:22,200 --> 00:33:24,360
Speaker 1: Sal could have been at a conference one of the

682
00:33:24,920 --> 00:33:27,440
Speaker 1: could have been certainly spoke at your conference years back,

683
00:33:27,640 --> 00:33:29,320
Speaker 1: which was great. And we end up we've done a

684
00:33:29,320 --> 00:33:31,360
Speaker 1: couple of shows together. We always you know, it's a

685
00:33:31,400 --> 00:33:35,400
Speaker 1: small universe that you end up bumping into each other. Um,

686
00:33:35,680 --> 00:33:40,680
Speaker 1: I recall it was before the flash crash, it was

687
00:33:40,680 --> 00:33:43,480
Speaker 1: when h f T really first started rearing its head.

688
00:33:44,480 --> 00:33:47,560
Speaker 1: I want to say it was before the financial crisis.

689
00:33:47,640 --> 00:33:49,920
Speaker 1: I want to say, oh seven could have been when

690
00:33:49,920 --> 00:33:53,360
Speaker 1: did you first notice things were getting a little, you know,

691
00:33:53,520 --> 00:33:56,240
Speaker 1: funky when you were trading? We we noticed it's certainly

692
00:33:56,240 --> 00:33:57,960
Speaker 1: like an oh five, oh six. Our first white paper

693
00:33:58,080 --> 00:33:59,680
Speaker 1: was the end of oh eight, so that's when we

694
00:33:59,680 --> 00:34:01,960
Speaker 1: publish something called Toxic Equity Trading, which kind of sat

695
00:34:02,000 --> 00:34:03,840
Speaker 1: on the shelf for a while. Everybody missed it because

696
00:34:03,960 --> 00:34:05,960
Speaker 1: we're in the mid financial crisis, and yeah, no one

697
00:34:06,000 --> 00:34:08,440
Speaker 1: really cared about market structure. It really popped up in

698
00:34:08,520 --> 00:34:11,840
Speaker 1: July of oh nine when the Goldman Sachs programmer Sergey Lichnikov,

699
00:34:12,880 --> 00:34:14,800
Speaker 1: and all of a sudden people started googling what's this

700
00:34:14,880 --> 00:34:17,520
Speaker 1: high frequency trading? And they popped up our white paper,

701
00:34:17,560 --> 00:34:21,640
Speaker 1: and then our phones lit up. There was a phenomenal um.

702
00:34:21,760 --> 00:34:24,239
Speaker 1: There was a phenomenal blog post. And I'm trying to

703
00:34:24,239 --> 00:34:26,920
Speaker 1: remember who first round it. I wouldn't be surprised if

704
00:34:26,920 --> 00:34:30,839
Speaker 1: it was Zero Hedge about who is this guy that

705
00:34:30,920 --> 00:34:34,160
Speaker 1: Goldman Sachs says that has escaped with the keys of

706
00:34:34,160 --> 00:34:38,240
Speaker 1: the castle and if unchecked, he could bring the market.

707
00:34:39,440 --> 00:34:41,719
Speaker 1: It was that the FBI. So where the hell did

708
00:34:41,760 --> 00:34:44,080
Speaker 1: these guys get that from? They had to have someone

709
00:34:44,120 --> 00:34:46,680
Speaker 1: from Goldman whispering in there, and and and Sergey is

710
00:34:46,680 --> 00:34:51,279
Speaker 1: still being prosecuted. He's still going through criminal eventually. And

711
00:34:51,320 --> 00:34:55,120
Speaker 1: you read the story, he essentially took open source stuff

712
00:34:55,400 --> 00:34:58,200
Speaker 1: that he had that basically was just a thing of convenience.

713
00:34:58,800 --> 00:35:00,799
Speaker 1: And uh and by the way, a lot of this

714
00:35:00,920 --> 00:35:05,880
Speaker 1: is actually in Flashboys. But the prosecution for that sham

715
00:35:05,920 --> 00:35:08,520
Speaker 1: on Guldman Sacks for for what they did. Don't cross

716
00:35:08,560 --> 00:35:12,000
Speaker 1: the squid apparently not or the Matt Taibi has is

717
00:35:12,000 --> 00:35:15,040
Speaker 1: still around Nobody's uh no, but nobody's He's the guy

718
00:35:15,080 --> 00:35:19,560
Speaker 1: who coined the phrase a jamming. It's blood funnel into

719
00:35:19,840 --> 00:35:23,000
Speaker 1: the throat of America of the vampire squid. If you

720
00:35:23,040 --> 00:35:26,200
Speaker 1: remember that very very famous article. It stayed with us.

721
00:35:26,400 --> 00:35:28,400
Speaker 1: So we we blew through a lot of stuff due

722
00:35:28,440 --> 00:35:31,440
Speaker 1: to time constraints earlier. Let me let me ask you

723
00:35:31,480 --> 00:35:34,360
Speaker 1: a few more questions that that we missed. So we

724
00:35:34,440 --> 00:35:38,200
Speaker 1: talked about improved trading costs and spreads and liquidly. The

725
00:35:38,600 --> 00:35:41,560
Speaker 1: one thing that we really didn't get to um that

726
00:35:41,680 --> 00:35:45,000
Speaker 1: I wanted was the difference between h f T s

727
00:35:45,440 --> 00:35:49,120
Speaker 1: and specialists or market makers. For that that fact, what

728
00:35:49,320 --> 00:35:53,080
Speaker 1: is the legal obligation what is the differences between what

729
00:35:53,680 --> 00:35:57,200
Speaker 1: they do and what they used to be required to do? Well,

730
00:35:57,239 --> 00:35:59,600
Speaker 1: that's to say, what exactly what is their obligation to

731
00:35:59,719 --> 00:36:02,359
Speaker 1: buy and sell a certain amount of shares? There is none,

732
00:36:02,360 --> 00:36:04,000
Speaker 1: And actually there's a couple of h f T s.

733
00:36:04,040 --> 00:36:07,800
Speaker 1: Now the bigger market makers are arguing for more obligations.

734
00:36:08,040 --> 00:36:10,560
Speaker 1: They want the guys who really understand that the h

735
00:36:10,600 --> 00:36:12,239
Speaker 1: f T market makers, you really know what they're doing,

736
00:36:12,360 --> 00:36:14,480
Speaker 1: who have really figured this out and are not necessarily

737
00:36:14,480 --> 00:36:17,640
Speaker 1: gaming the system, want more obligations. Give me some names.

738
00:36:18,040 --> 00:36:19,960
Speaker 1: I can't get into all the specifics, but there was

739
00:36:20,000 --> 00:36:24,080
Speaker 1: a what else do you put in that category? Well,

740
00:36:24,080 --> 00:36:25,680
Speaker 1: I don't know if I would put them in those

741
00:36:25,800 --> 00:36:27,880
Speaker 1: in the positive, I haven't seen those two guys asking

742
00:36:27,880 --> 00:36:30,800
Speaker 1: from more obligations. I did see a letter coming and

743
00:36:30,880 --> 00:36:33,480
Speaker 1: the tick size pilot program, which has been proposed by

744
00:36:33,480 --> 00:36:36,600
Speaker 1: the SEC. Virtue actually wrote a piece in their letter

745
00:36:36,600 --> 00:36:39,319
Speaker 1: which I thought was really impressive, asking them to put

746
00:36:39,400 --> 00:36:43,399
Speaker 1: more obligations on electronic market makers that, okay, let's step

747
00:36:43,480 --> 00:36:45,279
Speaker 1: up to the plate. How about the rest of those guys?

748
00:36:45,360 --> 00:36:47,719
Speaker 1: Do they want obligations? Now? Does this give a big

749
00:36:47,719 --> 00:36:50,319
Speaker 1: company like Virtue and uh An advantage. Is it a

750
00:36:50,320 --> 00:36:53,399
Speaker 1: competitive advantage or are they saying, hey, let's not kill

751
00:36:53,480 --> 00:36:56,200
Speaker 1: the golden goose here. I think they're they're pretty tight

752
00:36:56,280 --> 00:36:58,719
Speaker 1: on their market making. They pretty much. I mean they look,

753
00:36:58,719 --> 00:37:00,719
Speaker 1: they just published the next on and in there is

754
00:37:00,960 --> 00:37:02,520
Speaker 1: you know, a second time around for them going trying

755
00:37:02,560 --> 00:37:04,480
Speaker 1: to get that I p O out. But they've exposed

756
00:37:04,480 --> 00:37:06,440
Speaker 1: pretty much everything they do. They're not afraid to talk

757
00:37:06,480 --> 00:37:08,439
Speaker 1: about it. I would love to have some of these,

758
00:37:08,440 --> 00:37:10,680
Speaker 1: more of these one name h f T s that

759
00:37:10,680 --> 00:37:12,840
Speaker 1: you never heard of, that have a web page of

760
00:37:12,840 --> 00:37:15,520
Speaker 1: a half a page. Let them talk about their business model.

761
00:37:15,560 --> 00:37:17,600
Speaker 1: Are they making markets or they just you know, a

762
00:37:17,600 --> 00:37:19,560
Speaker 1: bunch of guys in a few different pods who are

763
00:37:19,600 --> 00:37:22,120
Speaker 1: looking at what what we call front running, which is

764
00:37:22,120 --> 00:37:23,920
Speaker 1: not really what they're doing. But what they're doing is

765
00:37:24,000 --> 00:37:26,239
Speaker 1: because they have no customers. By the way, all h

766
00:37:26,320 --> 00:37:28,319
Speaker 1: f t s are prop traders with no customers. They're

767
00:37:28,320 --> 00:37:31,839
Speaker 1: not what I describe as front running. You're actually right

768
00:37:31,880 --> 00:37:36,040
Speaker 1: because I'm mischaracterizing it. What they're doing is they're buying

769
00:37:36,160 --> 00:37:39,600
Speaker 1: a data service from the exchange which allows them to

770
00:37:39,880 --> 00:37:44,200
Speaker 1: packet sniff. It allows them to see orders not after

771
00:37:44,239 --> 00:37:47,200
Speaker 1: they hit the exchange. After they are traded. They like

772
00:37:47,520 --> 00:37:52,120
Speaker 1: it lets them see customer orders, not their customers, someone

773
00:37:52,120 --> 00:37:56,680
Speaker 1: else's customers, those orders before they're executed, so they get

774
00:37:56,680 --> 00:38:00,080
Speaker 1: to run ahead of them. If we all go the

775
00:38:00,200 --> 00:38:03,239
Speaker 1: front running, it's running ahead. But we were actually what

776
00:38:03,280 --> 00:38:05,759
Speaker 1: they're doing. And here's what they'll I'll put my h

777
00:38:05,800 --> 00:38:08,239
Speaker 1: f T ahead and they'll say that no, no one

778
00:38:08,280 --> 00:38:11,080
Speaker 1: can see an order before anyone else. It gets released

779
00:38:11,080 --> 00:38:14,239
Speaker 1: from the exchange server at the exact same time, and

780
00:38:14,400 --> 00:38:18,120
Speaker 1: anybody who spends the amount of money, the millions necessary

781
00:38:18,120 --> 00:38:20,920
Speaker 1: to process it can see that order. But guess what not.

782
00:38:21,200 --> 00:38:23,520
Speaker 1: You know, most people aren't doing what they do, So

783
00:38:23,719 --> 00:38:26,000
Speaker 1: we think what they're doing is not front running because

784
00:38:26,040 --> 00:38:28,920
Speaker 1: it's a customer but it's front running of demand and

785
00:38:28,960 --> 00:38:31,479
Speaker 1: front running of supply. They're building the book. They're looking

786
00:38:31,480 --> 00:38:33,560
Speaker 1: for what they call book pressure, and when they see

787
00:38:33,560 --> 00:38:35,960
Speaker 1: a stocks about to move based on their speed advantage

788
00:38:36,320 --> 00:38:39,080
Speaker 1: off and going right, so they're taking that offer before you.

789
00:38:39,200 --> 00:38:41,480
Speaker 1: They're canceling that bid before they get hit so they

790
00:38:41,520 --> 00:38:44,040
Speaker 1: don't get run over. Because essentially they're just trading getting

791
00:38:44,040 --> 00:38:47,200
Speaker 1: out of positions throughout the day. Their front running demand

792
00:38:47,280 --> 00:38:51,239
Speaker 1: and front running supply. Technically, nothing legal about that. Now,

793
00:38:51,280 --> 00:38:52,880
Speaker 1: as a broker, if I had a client order for

794
00:38:52,920 --> 00:38:54,840
Speaker 1: ten thousand shares and I decided to buy ahead of that,

795
00:38:54,920 --> 00:38:57,320
Speaker 1: I go to jail, as I should because that's front running.

796
00:38:57,760 --> 00:39:00,719
Speaker 1: That's that's legal front But Mike takeaway has always been,

797
00:39:01,400 --> 00:39:04,200
Speaker 1: you know, it's a zero sum game. And if these

798
00:39:04,239 --> 00:39:08,240
Speaker 1: guys are are capturing a penny on a hundred thousand shares,

799
00:39:08,680 --> 00:39:12,800
Speaker 1: that means the quote unquote natural buyer is not getting

800
00:39:12,800 --> 00:39:16,760
Speaker 1: that penny, and you multiply it by all of the transactions,

801
00:39:16,760 --> 00:39:19,600
Speaker 1: and that means tens of billions of dollars of profit

802
00:39:19,719 --> 00:39:24,799
Speaker 1: each year that should be in Grandma's IRA or your

803
00:39:24,920 --> 00:39:28,120
Speaker 1: four oh one k simply aren't there. It's going to

804
00:39:28,200 --> 00:39:31,000
Speaker 1: the h f T farms. Yes, it's going to proprietary

805
00:39:31,000 --> 00:39:34,120
Speaker 1: traders who are basically extracting that for their own and

806
00:39:34,160 --> 00:39:37,080
Speaker 1: not providing a social value. That's the key question. Are

807
00:39:37,120 --> 00:39:40,960
Speaker 1: they providing They're not going back into the ecosystem of

808
00:39:41,000 --> 00:39:43,000
Speaker 1: the trading world. Think about the old days when you

809
00:39:43,040 --> 00:39:45,440
Speaker 1: had an investment bank that had a trading department, a

810
00:39:45,520 --> 00:39:48,960
Speaker 1: research department, I p o s all sorts of different areas,

811
00:39:49,000 --> 00:39:51,399
Speaker 1: and in the trading department actually was a loss leader.

812
00:39:51,400 --> 00:39:54,399
Speaker 1: At times they would lose money to facilitate the other

813
00:39:54,440 --> 00:39:56,719
Speaker 1: parts of the business. But there you had the real

814
00:39:56,800 --> 00:39:59,680
Speaker 1: purpose of a stock market was to help companies go

815
00:39:59,760 --> 00:40:02,160
Speaker 1: pub right, bring them out, put them on road shows.

816
00:40:02,360 --> 00:40:04,920
Speaker 1: That's kind of falling aside because the economics aren't there

817
00:40:04,960 --> 00:40:07,840
Speaker 1: any right, the economics of the research department. Now you

818
00:40:08,120 --> 00:40:11,880
Speaker 1: talked about earlier how great decimalization is. I've read a

819
00:40:12,000 --> 00:40:14,880
Speaker 1: number of people complaining that when we went from fractions

820
00:40:14,880 --> 00:40:18,320
Speaker 1: to decimalizations, you you took a lot of the profit

821
00:40:18,320 --> 00:40:22,080
Speaker 1: out of trading, and that profit was what subsidized research.

822
00:40:22,400 --> 00:40:24,320
Speaker 1: Is that true? Yeah, And actually there's something called a

823
00:40:24,360 --> 00:40:26,959
Speaker 1: tick size pilot which has just been proposed. We wrote

824
00:40:27,000 --> 00:40:29,279
Speaker 1: a comment letter, the industry wrote tons of comment letters

825
00:40:29,320 --> 00:40:32,760
Speaker 1: on it. But the theory there is decimalization was great,

826
00:40:32,920 --> 00:40:35,160
Speaker 1: but you never put in a minimum spread. In other ways,

827
00:40:35,160 --> 00:40:37,440
Speaker 1: you collapse the spreads down to a penny. Wouldn't it

828
00:40:37,480 --> 00:40:41,080
Speaker 1: make sense for stocks that or even dying, for say

829
00:40:41,080 --> 00:40:43,880
Speaker 1: a small cap that doesn't trade. We're trying to attract

830
00:40:43,920 --> 00:40:46,560
Speaker 1: real liquidity, and real liquidity providers are not going to

831
00:40:46,640 --> 00:40:48,720
Speaker 1: hang around when they're gonna get picked off by somebody

832
00:40:48,719 --> 00:40:50,759
Speaker 1: who has a penny to come in play in the game.

833
00:40:51,000 --> 00:40:53,479
Speaker 1: Where now, if it was a nickel and you wanted

834
00:40:53,520 --> 00:40:55,399
Speaker 1: to get in this game and pick somebody off, it's

835
00:40:55,400 --> 00:40:57,520
Speaker 1: gonna cost you more and your risk is gonna go higher.

836
00:40:57,840 --> 00:40:59,520
Speaker 1: It's a really good theory. We don't know if it

837
00:40:59,560 --> 00:41:01,680
Speaker 1: will work, which is why it's called the pilot program.

838
00:41:01,760 --> 00:41:03,719
Speaker 1: The SEC wants to put out, but the industry is

839
00:41:03,760 --> 00:41:06,080
Speaker 1: going nuts already because they think it's gonna translate them

840
00:41:06,120 --> 00:41:08,120
Speaker 1: to what they call a trade at rule, which would

841
00:41:08,160 --> 00:41:09,840
Speaker 1: mean that you can't trade in the dark until you

842
00:41:09,880 --> 00:41:12,080
Speaker 1: trade on lid venues first. So they haven't a fit

843
00:41:12,120 --> 00:41:14,400
Speaker 1: because it would destroy their business model. But that's not

844
00:41:14,440 --> 00:41:16,759
Speaker 1: what the pilot program is calling for it. So, so

845
00:41:16,800 --> 00:41:19,720
Speaker 1: if we went to a nickel spread with size on

846
00:41:19,719 --> 00:41:22,760
Speaker 1: on bid and ask what would that do the trading costs?

847
00:41:22,760 --> 00:41:25,880
Speaker 1: What would that do to to a company like Vanguard

848
00:41:26,040 --> 00:41:28,960
Speaker 1: or any of the big mutual funds. How would then

849
00:41:29,000 --> 00:41:32,680
Speaker 1: impact they're they're trading right? And this would be only

850
00:41:32,719 --> 00:41:35,440
Speaker 1: on small to mid cap. So we proposed the privot

851
00:41:35,440 --> 00:41:37,839
Speaker 1: program says five billion market cap and less. We said

852
00:41:37,840 --> 00:41:40,000
Speaker 1: that's too high, go to like two billion in less

853
00:41:40,200 --> 00:41:42,720
Speaker 1: and and and that's still a couple of thousand stocks

854
00:41:42,800 --> 00:41:46,400
Speaker 1: under two billion, none of which really trade. You know,

855
00:41:46,400 --> 00:41:49,160
Speaker 1: they trade a few hundred thousand shiars a day. Their orphans.

856
00:41:49,200 --> 00:41:51,440
Speaker 1: No one trades them, right, nobody follows them. There's no

857
00:41:51,480 --> 00:41:54,479
Speaker 1: Wall Street coverage on them. And you know, last week

858
00:41:54,640 --> 00:41:58,440
Speaker 1: we were talking about the small cap um premium and

859
00:41:58,560 --> 00:42:01,719
Speaker 1: why small cap stocks have a tendency to outperform big

860
00:42:01,760 --> 00:42:05,040
Speaker 1: gap stocks over time, especially if you control for junk,

861
00:42:05,520 --> 00:42:08,640
Speaker 1: and that was one of the factors is their orphan stocks.

862
00:42:08,920 --> 00:42:12,000
Speaker 1: There's not a lot of liquidly, there's almost no coverage

863
00:42:12,120 --> 00:42:15,120
Speaker 1: there certainly isn't big firm coverage. You're not gonna get

864
00:42:15,320 --> 00:42:19,080
Speaker 1: Goldman and more Maryland Morgan covering, you know, and even

865
00:42:19,200 --> 00:42:22,920
Speaker 1: scrap the small firms. The regional brokers. Remember some of

866
00:42:22,920 --> 00:42:26,200
Speaker 1: the names like Robinson, Humphrey McDonald and all of these

867
00:42:26,280 --> 00:42:27,880
Speaker 1: names that would be the been eating up by other

868
00:42:27,920 --> 00:42:29,840
Speaker 1: brokers have went out of business. They don't have the

869
00:42:29,880 --> 00:42:32,719
Speaker 1: economics to support that trading anymore. Those market makers are gone,

870
00:42:32,760 --> 00:42:35,040
Speaker 1: they've been replaced by algorithms and so on. So how

871
00:42:35,040 --> 00:42:36,840
Speaker 1: do you bring them back? And and the theory and

872
00:42:37,040 --> 00:42:39,560
Speaker 1: Jeff Solomon over account and who's the president of account.

873
00:42:39,800 --> 00:42:42,360
Speaker 1: He's big on this. That is to how do I

874
00:42:42,400 --> 00:42:44,480
Speaker 1: get these market makers back? And you have to put

875
00:42:44,520 --> 00:42:47,239
Speaker 1: an economic incentive. You have to add otherwise no one's

876
00:42:47,239 --> 00:42:49,520
Speaker 1: gonna do it. And the economic incentive would be a

877
00:42:49,600 --> 00:42:52,120
Speaker 1: nicola a dime spread. And that's how. But you also

878
00:42:52,160 --> 00:42:54,080
Speaker 1: have to have an obligation come in there and put

879
00:42:54,120 --> 00:42:55,920
Speaker 1: some real size in there, not just a hundred chair

880
00:42:56,040 --> 00:42:57,920
Speaker 1: that that's been one of the complaints. You see this

881
00:42:57,960 --> 00:42:59,759
Speaker 1: penny spread and you go to buy something and you

882
00:42:59,800 --> 00:43:02,120
Speaker 1: get forty seven shares, And what's the point I need?

883
00:43:03,920 --> 00:43:07,520
Speaker 1: The penny spread is really the liquidity is illusory, the

884
00:43:07,600 --> 00:43:09,759
Speaker 1: spread is illusory. A lot of things that look like

885
00:43:09,800 --> 00:43:13,520
Speaker 1: one thing really aren't. Although I'm starting to get a sense,

886
00:43:13,800 --> 00:43:17,640
Speaker 1: a hint from you that your position has softened over

887
00:43:17,680 --> 00:43:20,359
Speaker 1: the past few years, because two or two years ago,

888
00:43:20,520 --> 00:43:22,480
Speaker 1: it was like we're all going to heck and this

889
00:43:22,640 --> 00:43:25,440
Speaker 1: is a disaster in the market's gonna collapse under its

890
00:43:25,480 --> 00:43:28,719
Speaker 1: own weight. To Hey, there have been some incremental improvements

891
00:43:28,960 --> 00:43:31,840
Speaker 1: and some market based improvements like I E X that

892
00:43:31,920 --> 00:43:37,040
Speaker 1: are actually bringing back some positive things to the market structure.

893
00:43:37,239 --> 00:43:39,120
Speaker 1: I would agree with that, and and and I think

894
00:43:39,160 --> 00:43:41,520
Speaker 1: it's because of the efforts of a bunch of folks,

895
00:43:41,520 --> 00:43:43,120
Speaker 1: and you know, we certainly were one of the first

896
00:43:43,120 --> 00:43:45,600
Speaker 1: out of the water. But if this market structure debate

897
00:43:45,680 --> 00:43:48,759
Speaker 1: didn't get pushed forward, we wouldn't be talking about improvements

898
00:43:48,800 --> 00:43:52,320
Speaker 1: right now. We wouldn't be talking about market market based solutions.

899
00:43:52,400 --> 00:43:55,000
Speaker 1: We'd be talking about the same problem. So systemic risks

900
00:43:55,040 --> 00:43:57,359
Speaker 1: that you guys have written a lot about, they'd still

901
00:43:57,360 --> 00:43:59,239
Speaker 1: be there, right and then they are still there, But

902
00:43:59,280 --> 00:44:02,000
Speaker 1: we're moving long in the right through. We're getting exchanges now,

903
00:44:02,040 --> 00:44:04,360
Speaker 1: like Jeff Sprecker of the New York Stock Exchange coming

904
00:44:04,360 --> 00:44:06,840
Speaker 1: in saying, hey, this make or take a model is ridiculous.

905
00:44:06,880 --> 00:44:08,879
Speaker 1: We gotta get rid of it. A couple of years ago,

906
00:44:09,440 --> 00:44:11,839
Speaker 1: you know, Duncanita Howard would never would have said that.

907
00:44:12,000 --> 00:44:14,360
Speaker 1: So things are different, and in the market structure debate,

908
00:44:14,520 --> 00:44:17,560
Speaker 1: it's a sophisticated one. It's moving forward, and this is

909
00:44:17,560 --> 00:44:20,520
Speaker 1: what we always wanted. So on a scale of one

910
00:44:20,560 --> 00:44:23,640
Speaker 1: to ten, if it was a disaster in two thousand ten,

911
00:44:23,719 --> 00:44:27,919
Speaker 1: at zero and ten is the optimal market structure, where

912
00:44:27,920 --> 00:44:33,160
Speaker 1: are we today? Four or not even not even halfway there? Wow,

913
00:44:33,200 --> 00:44:37,040
Speaker 1: that's uh, that's a really fascinating number. What what other

914
00:44:37,160 --> 00:44:41,919
Speaker 1: areas of the markets were you cop stock market wise?

915
00:44:41,960 --> 00:44:44,000
Speaker 1: I mean the biggest we have issues with, like I

916
00:44:44,040 --> 00:44:48,920
Speaker 1: mentioned before, fragmentation, dark pools. The non exchange off exchange

917
00:44:49,360 --> 00:44:53,280
Speaker 1: percentage of volume is close to which means that almost

918
00:44:53,320 --> 00:44:55,560
Speaker 1: four out of every ten shares is not traded on

919
00:44:55,600 --> 00:44:59,720
Speaker 1: an exchange. So now when we look at really soft volume,

920
00:44:59,760 --> 00:45:03,560
Speaker 1: which we've seen since you know, MARCHO nine lows how

921
00:45:03,640 --> 00:45:06,280
Speaker 1: much of that is just mom and pop walking away?

922
00:45:06,600 --> 00:45:09,720
Speaker 1: How much of it is dark pools? Why has volume

923
00:45:10,160 --> 00:45:12,560
Speaker 1: been so low? And what does that mean? Volume is

924
00:45:12,600 --> 00:45:15,080
Speaker 1: low right now because volatility is low. But if ilum

925
00:45:15,120 --> 00:45:18,719
Speaker 1: has been fairly suppressed for five years, has uh it

926
00:45:18,760 --> 00:45:20,600
Speaker 1: well off the peak of it. We wereround ten billion

927
00:45:20,640 --> 00:45:23,920
Speaker 1: shares across all exchanges back when the financial crisis was

928
00:45:23,960 --> 00:45:26,760
Speaker 1: going on. Now we're closely to six billions. So it's

929
00:45:26,600 --> 00:45:28,840
Speaker 1: a big drop. Yeah, because we got really lot of

930
00:45:28,840 --> 00:45:31,359
Speaker 1: the noisy. You know, if it volatility is not there,

931
00:45:31,400 --> 00:45:35,359
Speaker 1: you can't profit from latency arbitrarge situations. You can't pick

932
00:45:35,400 --> 00:45:38,120
Speaker 1: off someone as easily. Where is some of that volume going?

933
00:45:38,120 --> 00:45:41,480
Speaker 1: It's going to other markets. The currency market has exploded

934
00:45:41,520 --> 00:45:45,160
Speaker 1: with the electronic activity because it's so it's much more volatile,

935
00:45:45,200 --> 00:45:48,200
Speaker 1: and it's very liquid, and you know these trillions of

936
00:45:48,239 --> 00:45:52,120
Speaker 1: dollars literally, and it's it's not centralized. The they don't

937
00:45:52,120 --> 00:45:53,719
Speaker 1: even have at least in the equity world, we have

938
00:45:53,760 --> 00:45:56,239
Speaker 1: a linkage that's SIP the security. They haven't only got

939
00:45:56,239 --> 00:45:58,080
Speaker 1: a linkage in the in the foreign exchange world. So

940
00:45:58,080 --> 00:46:00,279
Speaker 1: it's a lot easier for you to develop your own

941
00:46:00,360 --> 00:46:03,319
Speaker 1: quote and to see something better than somebody else. The

942
00:46:03,360 --> 00:46:06,000
Speaker 1: banks have their own problems there as well. There's an article,

943
00:46:06,040 --> 00:46:09,040
Speaker 1: yes they'm on Bloomberg about the Last Look and things

944
00:46:09,080 --> 00:46:10,759
Speaker 1: like that, that there's a lot of issues in the

945
00:46:10,760 --> 00:46:13,399
Speaker 1: currency world, which we in the equity world. Actually we're

946
00:46:13,400 --> 00:46:16,120
Speaker 1: a lot more advanced when it comes to market structure

947
00:46:16,280 --> 00:46:18,920
Speaker 1: than something in the currency. People have talked about the

948
00:46:18,920 --> 00:46:22,880
Speaker 1: currency markets and and four x as the broadest, deepest

949
00:46:22,920 --> 00:46:26,640
Speaker 1: market with the most sophisticated systems, and you never can

950
00:46:26,680 --> 00:46:30,719
Speaker 1: you see a flash crash and concert absolutely, I absolutely

951
00:46:30,719 --> 00:46:33,400
Speaker 1: think you can it. Is it inevitable? Is it possible?

952
00:46:33,480 --> 00:46:35,440
Speaker 1: Is it likely? I think it's probably likely because with

953
00:46:35,520 --> 00:46:37,960
Speaker 1: the with the global event, you know, if if the

954
00:46:38,040 --> 00:46:40,720
Speaker 1: Swedish bang on to Swiss, Swiss Bank decides to change

955
00:46:40,719 --> 00:46:43,960
Speaker 1: the way they're going to set things which ago and

956
00:46:44,000 --> 00:46:46,640
Speaker 1: you did see a rapid moving currencies because you know,

957
00:46:46,800 --> 00:46:49,399
Speaker 1: is it really institutional volume, is it real bona FID

958
00:46:49,440 --> 00:46:52,719
Speaker 1: hedgers coming in there trading that, No, it's noise. It's

959
00:46:52,719 --> 00:46:55,000
Speaker 1: guys who are using I call it noise. But it's

960
00:46:55,000 --> 00:46:58,880
Speaker 1: basically a multi asset model. So you're trading stocks, equity,

961
00:46:58,920 --> 00:47:01,880
Speaker 1: you're trading stocks, opts, and futures, currencies, bonds. You're putting

962
00:47:01,880 --> 00:47:06,040
Speaker 1: it all together into this little magic sausage factory yours,

963
00:47:06,200 --> 00:47:08,560
Speaker 1: and you're developing a strategy. So if one all of

964
00:47:08,600 --> 00:47:10,680
Speaker 1: a sudden, there's an event in one area, you're gonna

965
00:47:10,680 --> 00:47:12,960
Speaker 1: move out. So yeah, it could happen in any one

966
00:47:13,000 --> 00:47:16,120
Speaker 1: of those markets. So zero to ten is a four,

967
00:47:16,440 --> 00:47:19,319
Speaker 1: So that's a big improvement, but that still means we're

968
00:47:19,320 --> 00:47:21,440
Speaker 1: sticks away from ten and and there's a lot of

969
00:47:21,440 --> 00:47:25,759
Speaker 1: potential problem. So you had mentioned um currencies, you had

970
00:47:25,760 --> 00:47:29,600
Speaker 1: mentioned treasuries. What about commodity markets. Commodity markets a little

971
00:47:29,600 --> 00:47:31,920
Speaker 1: different because they operate more in a vertical silo, so

972
00:47:31,960 --> 00:47:35,120
Speaker 1: you don't have the multi exchange, multi venue model, and

973
00:47:35,239 --> 00:47:38,960
Speaker 1: you have your trading oil, you're trading natures, or you're

974
00:47:39,040 --> 00:47:41,960
Speaker 1: you're not. Everything isn't as it's but they use a

975
00:47:42,040 --> 00:47:45,160
Speaker 1: huge leverage though they do and they have, but they're

976
00:47:45,160 --> 00:47:47,120
Speaker 1: they're a lot more control that they seem to have

977
00:47:47,160 --> 00:47:48,879
Speaker 1: their their grips on it better. I don't know, I'm

978
00:47:48,880 --> 00:47:51,400
Speaker 1: not certainly not an expert. You know, people used to

979
00:47:51,440 --> 00:47:53,759
Speaker 1: talk about commodity markets like it was the wild West

980
00:47:53,840 --> 00:47:55,720
Speaker 1: and all the lock limit down and all this crazy

981
00:47:55,760 --> 00:47:59,160
Speaker 1: stuff happens, and you're saying, the structure is better than

982
00:47:59,200 --> 00:48:02,320
Speaker 1: what we see else. Ironically, the the the flash crash

983
00:48:02,320 --> 00:48:04,000
Speaker 1: of two thousand and ten that we spoke about earlier

984
00:48:04,160 --> 00:48:07,160
Speaker 1: was stopped because of a piece of software at the CME,

985
00:48:07,280 --> 00:48:10,960
Speaker 1: the Chicago Mercantile Exchange called stop logic, which said that

986
00:48:11,000 --> 00:48:14,040
Speaker 1: if it got too volatile, halted for five seconds. I

987
00:48:14,080 --> 00:48:16,400
Speaker 1: believe it was five seconds. And that's really what the

988
00:48:16,400 --> 00:48:21,239
Speaker 1: bottom of the flash five second quote hall. And that

989
00:48:21,360 --> 00:48:23,759
Speaker 1: tells you that every asset is linked, because when the

990
00:48:23,960 --> 00:48:26,600
Speaker 1: mini stopped trading, everybody said, hold on a second, what's

991
00:48:26,600 --> 00:48:28,960
Speaker 1: going on, and that guys like us who are trading stocks,

992
00:48:28,960 --> 00:48:31,359
Speaker 1: but guys who have multi asset strategies who were all

993
00:48:31,360 --> 00:48:34,200
Speaker 1: electronically linked, then realized, hey, there really isn't something going

994
00:48:34,400 --> 00:48:36,320
Speaker 1: on here For people who don't know E Minis or

995
00:48:36,360 --> 00:48:41,200
Speaker 1: the SMP futures huge leverage with that you're called minis,

996
00:48:41,239 --> 00:48:43,759
Speaker 1: but you're putting up a million dollars and trading like

997
00:48:44,000 --> 00:48:45,640
Speaker 1: I don't remember what it is, but it's tens of

998
00:48:45,680 --> 00:48:48,160
Speaker 1: millions of dollars of stock. It doesn't take a lot

999
00:48:48,160 --> 00:48:51,279
Speaker 1: of money to swing a boatload of stock in those

1000
00:48:51,440 --> 00:48:53,879
Speaker 1: and it's supposedly the most liquid instrument when it comes

1001
00:48:53,880 --> 00:48:56,560
Speaker 1: to commodities. Yet that was also the one blamed for

1002
00:48:56,600 --> 00:48:59,399
Speaker 1: the initiation of the flash crash because they couldn't handle

1003
00:48:59,440 --> 00:49:01,880
Speaker 1: seventy five housing contracts, which makes you kind of scratch

1004
00:49:01,880 --> 00:49:03,360
Speaker 1: your head. Why could the most liquid security in the

1005
00:49:03,440 --> 00:49:05,719
Speaker 1: world not handled. Sure, it's a big order, but it

1006
00:49:05,840 --> 00:49:08,879
Speaker 1: doesn't the reason that that side put out doesn't hold

1007
00:49:08,920 --> 00:49:11,719
Speaker 1: water at all. Right, So, first, the first rumor we

1008
00:49:11,840 --> 00:49:14,120
Speaker 1: heard was it was a fat thumb, which is trade

1009
00:49:14,160 --> 00:49:17,320
Speaker 1: or talk for somebody accidentally throwing an extra zero on

1010
00:49:17,400 --> 00:49:20,360
Speaker 1: an order, or or you know, just leaning on a keyboard.

1011
00:49:20,440 --> 00:49:22,560
Speaker 1: And you know, if you ever leading on a keyboard

1012
00:49:22,600 --> 00:49:25,759
Speaker 1: and here, just picture that on a trading desk and

1013
00:49:25,840 --> 00:49:28,040
Speaker 1: what that means when you're shooting you know, millions of

1014
00:49:28,200 --> 00:49:30,839
Speaker 1: orders all at once. It it's hey, fat thumb, that's

1015
00:49:30,880 --> 00:49:33,000
Speaker 1: really a problem. I'll tell you one of the problem

1016
00:49:33,080 --> 00:49:35,680
Speaker 1: that they have right now, they don't have the surveillance

1017
00:49:36,000 --> 00:49:39,760
Speaker 1: they SEC in particular, does not have a surveillance system

1018
00:49:39,960 --> 00:49:44,400
Speaker 1: to keep up with today's modern markets, today's micro second trading,

1019
00:49:44,719 --> 00:49:46,680
Speaker 1: you know, and a million of a second. These guys

1020
00:49:46,719 --> 00:49:49,000
Speaker 1: are doing things that we can't. There's a world underneath it, right.

1021
00:49:49,400 --> 00:49:52,640
Speaker 1: The SEC doesn't have a system tomorrow. They actually had

1022
00:49:52,680 --> 00:49:55,000
Speaker 1: to sub out to an h f T guy to

1023
00:49:55,160 --> 00:49:57,239
Speaker 1: build a system called Midas that COSTOM two and a

1024
00:49:57,280 --> 00:49:59,400
Speaker 1: half million a year, which is barely a bunch of

1025
00:49:59,480 --> 00:50:03,040
Speaker 1: direct fee being aggregated. How can nanex do this? And

1026
00:50:03,120 --> 00:50:07,680
Speaker 1: if you haven't seen the beautiful artwork that how do

1027
00:50:07,840 --> 00:50:11,200
Speaker 1: I pronounce his last name, Eric hunt Setter's work on

1028
00:50:11,600 --> 00:50:15,080
Speaker 1: nanax dot net? Is that right? Uh? He creates these

1029
00:50:15,160 --> 00:50:21,560
Speaker 1: beautiful mosaic showing all these quotes and trades and their

1030
00:50:21,680 --> 00:50:26,000
Speaker 1: artwork their fantastic all they are visual representation of actual trades.

1031
00:50:26,560 --> 00:50:29,600
Speaker 1: How can he do it? But the SEC can't. It's incredible.

1032
00:50:29,640 --> 00:50:31,719
Speaker 1: He could have built and I've spoken to Erica and

1033
00:50:32,000 --> 00:50:33,680
Speaker 1: we speak to him often but he could have built

1034
00:50:33,760 --> 00:50:35,959
Speaker 1: Midas for a lot cheaper than they spent on it. Okay,

1035
00:50:36,200 --> 00:50:38,400
Speaker 1: he runs it. He basically takes in the feeds and

1036
00:50:38,480 --> 00:50:41,640
Speaker 1: he analyzes this stuff. But the problem with the secs

1037
00:50:41,840 --> 00:50:43,759
Speaker 1: and and the problem with our regulatory system is we've

1038
00:50:43,840 --> 00:50:47,120
Speaker 1: got futures on one side, which is the CFCC, We've

1039
00:50:47,160 --> 00:50:49,640
Speaker 1: got the SEC we're doing equities and options. They're not

1040
00:50:49,800 --> 00:50:53,319
Speaker 1: aggregating everything, and in some counts over their treasuries. Over here,

1041
00:50:53,400 --> 00:50:56,359
Speaker 1: it's really very much fragment. And if if you're if

1042
00:50:56,400 --> 00:50:59,440
Speaker 1: you're spoofing in the currency market to affect something in

1043
00:50:59,480 --> 00:51:01,759
Speaker 1: the bond market, you're never gonna be caught. Is that

1044
00:51:01,920 --> 00:51:04,000
Speaker 1: going on today? I don't know, because we don't have

1045
00:51:04,040 --> 00:51:06,560
Speaker 1: the surveillance systems to watch it, but theoretically that I

1046
00:51:06,600 --> 00:51:08,560
Speaker 1: would think so absolutely. So let me ask you a

1047
00:51:08,640 --> 00:51:11,600
Speaker 1: hypothetical question. One of the guys. So, whenever we do

1048
00:51:11,760 --> 00:51:14,960
Speaker 1: these these conversations, my office is always so, who are

1049
00:51:14,960 --> 00:51:17,239
Speaker 1: we speaking with this week? And I always tell them so,

1050
00:51:17,680 --> 00:51:20,080
Speaker 1: and I get some fascinating questions from from some of

1051
00:51:20,120 --> 00:51:23,000
Speaker 1: the staff members. And my head of research said, Hey,

1052
00:51:23,200 --> 00:51:26,160
Speaker 1: do me a favor. Ask Joe, what would happen if

1053
00:51:26,239 --> 00:51:29,960
Speaker 1: tonight every h f T firm was shut down, what

1054
00:51:30,080 --> 00:51:33,719
Speaker 1: would happen to markets going forward? Much rejoicing, No, I'm

1055
00:51:33,719 --> 00:51:38,080
Speaker 1: just I'm just no. If they shut down, sure, you're

1056
00:51:38,080 --> 00:51:41,840
Speaker 1: gonna widen spreads, right naturally, If if the bigils in

1057
00:51:41,920 --> 00:51:44,800
Speaker 1: that gap, who will replaces them? It's a good question.

1058
00:51:44,920 --> 00:51:47,759
Speaker 1: Do market forces that overnight? It's not gonna happen, right,

1059
00:51:47,920 --> 00:51:50,640
Speaker 1: But eventually, Let's take for instance, that tick size pilot

1060
00:51:50,680 --> 00:51:52,480
Speaker 1: we were talking about. If the h f T s

1061
00:51:52,520 --> 00:51:54,360
Speaker 1: don't want to play in the stocks because of a

1062
00:51:54,440 --> 00:51:58,120
Speaker 1: nickel spread, there will be someone coming along. The vacuum

1063
00:51:58,160 --> 00:52:00,160
Speaker 1: will be there's money to be made, there's it to

1064
00:52:00,239 --> 00:52:02,640
Speaker 1: be made. Real traditional market makers will come back, Real

1065
00:52:02,719 --> 00:52:05,640
Speaker 1: liquidity will come back. Maybe the spread is a little

1066
00:52:05,640 --> 00:52:07,480
Speaker 1: bit wider, but we would argue that to spread is

1067
00:52:07,560 --> 00:52:09,640
Speaker 1: really today's market. Like you was saying before, if I

1068
00:52:11,719 --> 00:52:13,600
Speaker 1: want to buy five thousand shares, how do I get

1069
00:52:13,640 --> 00:52:15,600
Speaker 1: five And if I can do it with a nickel spread,

1070
00:52:15,640 --> 00:52:19,440
Speaker 1: that's actually a better thing, right the hundred fifty to forty.

1071
00:52:19,719 --> 00:52:23,040
Speaker 1: By the time you feel thousand shares your fifteen cents away,

1072
00:52:23,120 --> 00:52:25,440
Speaker 1: that's right. And we've heard all sorts of stories about

1073
00:52:25,719 --> 00:52:28,719
Speaker 1: about those sort of sweeps and that penny spread is

1074
00:52:28,760 --> 00:52:30,480
Speaker 1: nonsense now it is, it isn't. And that's where you

1075
00:52:30,520 --> 00:52:32,239
Speaker 1: get picked off by the way that people will see

1076
00:52:32,280 --> 00:52:34,799
Speaker 1: you coming. And that's where brad over they move out

1077
00:52:34,840 --> 00:52:37,160
Speaker 1: of the way. They will, they'll disappear if you're coming

1078
00:52:37,200 --> 00:52:39,080
Speaker 1: through a slower router, and the next thing you know,

1079
00:52:39,120 --> 00:52:40,840
Speaker 1: they'll take the stock in front of you, and you

1080
00:52:40,880 --> 00:52:42,839
Speaker 1: have now paying a nickel higher and you've got two

1081
00:52:43,360 --> 00:52:44,920
Speaker 1: eighty seven shares, Like, what the heck are you doing?

1082
00:52:45,040 --> 00:52:46,880
Speaker 1: That's we have to be when you're trading now, especially

1083
00:52:46,960 --> 00:52:49,680
Speaker 1: with larger pieces, you better be sophisticated. You better know

1084
00:52:49,719 --> 00:52:51,879
Speaker 1: how the routes work. You better know what's going into

1085
00:52:51,960 --> 00:52:54,520
Speaker 1: every route, what type of order is going out. We

1086
00:52:54,520 --> 00:52:56,560
Speaker 1: didn't even get into these crazy order types. But you

1087
00:52:56,680 --> 00:52:59,520
Speaker 1: need to know the plumbing right or or like if

1088
00:52:59,560 --> 00:53:01,120
Speaker 1: you ran a race car, if you're a race car drive,

1089
00:53:01,120 --> 00:53:02,919
Speaker 1: you want to know what's in that engine. You better

1090
00:53:03,000 --> 00:53:05,080
Speaker 1: know the inside of this engine. If you're trading size

1091
00:53:05,200 --> 00:53:08,319
Speaker 1: right now, you will get picked off. And then unfortunately,

1092
00:53:08,719 --> 00:53:10,000
Speaker 1: at the end of the day, you may think you

1093
00:53:10,080 --> 00:53:13,120
Speaker 1: got what they call a v WAPP or volume weighted

1094
00:53:13,200 --> 00:53:16,520
Speaker 1: average price, that that's gonna be across all these different

1095
00:53:17,200 --> 00:53:20,279
Speaker 1: bids and offers where you actually paid you match the

1096
00:53:20,400 --> 00:53:22,840
Speaker 1: daily v So everybody. There's a lot of managers that

1097
00:53:22,880 --> 00:53:25,080
Speaker 1: will try to benchmark the v WAP and they'll say,

1098
00:53:25,239 --> 00:53:26,880
Speaker 1: if I got as good as the VWAP, I did

1099
00:53:26,920 --> 00:53:28,920
Speaker 1: a good job today. And we would argue that you're

1100
00:53:28,920 --> 00:53:31,960
Speaker 1: not over paying through the trades correct, correct, because you're

1101
00:53:31,960 --> 00:53:34,680
Speaker 1: doing an average job, because that's the average price. And

1102
00:53:34,880 --> 00:53:37,239
Speaker 1: we would argue that if you're not sophisticated with your routes,

1103
00:53:37,440 --> 00:53:40,080
Speaker 1: you're actually causing that VWAP to be higher than where

1104
00:53:40,080 --> 00:53:42,239
Speaker 1: it should if you're a buyer, or a lower if

1105
00:53:42,320 --> 00:53:44,719
Speaker 1: you're a seller, because you're not really paying attention to

1106
00:53:44,840 --> 00:53:47,200
Speaker 1: the intricacies of why did that hundred shairs go up

1107
00:53:47,400 --> 00:53:49,440
Speaker 1: on the bat's inverted exchange, Why did it go up

1108
00:53:49,440 --> 00:53:51,600
Speaker 1: in a dark pool? Why is that route going here

1109
00:53:51,840 --> 00:53:55,759
Speaker 1: constantly in the same low rebate venue or high rebate venue.

1110
00:53:55,960 --> 00:53:57,960
Speaker 1: These are the questions that we ask all day long

1111
00:53:58,200 --> 00:54:00,600
Speaker 1: why our trading is different than others out there. But

1112
00:54:00,719 --> 00:54:02,520
Speaker 1: once you learn the inside of it, then you'll be

1113
00:54:02,880 --> 00:54:04,480
Speaker 1: armed a little bit better. It's still gonna be tough.

1114
00:54:04,680 --> 00:54:08,080
Speaker 1: It's not easy, So you've been doing this for twenty

1115
00:54:08,200 --> 00:54:11,560
Speaker 1: plus years. I remember when I began around that same

1116
00:54:11,640 --> 00:54:14,000
Speaker 1: time as a trader and as if I'm a if

1117
00:54:14,000 --> 00:54:16,640
Speaker 1: I'm a seller, and suddenly Goldman shows up on the

1118
00:54:16,719 --> 00:54:19,160
Speaker 1: top of you know, it was mostly NASDAX stocks on

1119
00:54:19,200 --> 00:54:22,160
Speaker 1: the top of the box as the bidder for size.

1120
00:54:22,640 --> 00:54:24,600
Speaker 1: I would get out of the way. But I wasn't

1121
00:54:24,600 --> 00:54:27,680
Speaker 1: a market maker. I was just executing orders um either

1122
00:54:27,800 --> 00:54:32,160
Speaker 1: for proprietary trades or on behalf of clients. This is

1123
00:54:32,280 --> 00:54:35,880
Speaker 1: really very, very different than those days. That's almost simple

1124
00:54:36,000 --> 00:54:40,680
Speaker 1: and quaint compared to the complications you're described today. It's

1125
00:54:40,760 --> 00:54:42,359
Speaker 1: it's you know, that was like the axe. They used

1126
00:54:42,360 --> 00:54:44,239
Speaker 1: to call it, right, that's right, and the ax came

1127
00:54:44,280 --> 00:54:46,640
Speaker 1: in everybody, and you knew in different stocks it was

1128
00:54:46,719 --> 00:54:49,279
Speaker 1: different different how you know, there were certain there was

1129
00:54:49,400 --> 00:54:52,600
Speaker 1: certain stocks that when bear Sterns showed up, hey, everybody

1130
00:54:52,640 --> 00:54:54,440
Speaker 1: out of the pool. They were starting to suck up

1131
00:54:54,440 --> 00:54:57,160
Speaker 1: a million shares or certain hedge funds, let's just call it.

1132
00:54:57,200 --> 00:55:00,160
Speaker 1: That everyone knew was quote unquote March smart money. They

1133
00:55:00,200 --> 00:55:02,320
Speaker 1: started ripping through his stock, but that was you know,

1134
00:55:02,400 --> 00:55:04,560
Speaker 1: his intelligence. Sure there was issues back then of of

1135
00:55:05,120 --> 00:55:08,640
Speaker 1: transparency as well. But the market now is so much different,

1136
00:55:08,640 --> 00:55:10,879
Speaker 1: and when you think about it. Scott Patterson, the Wall

1137
00:55:10,880 --> 00:55:13,480
Speaker 1: Street Journal reporter, wrote a book called Dark Pools, which

1138
00:55:13,560 --> 00:55:16,080
Speaker 1: I think is tremendous, by the way, and I love now.

1139
00:55:16,200 --> 00:55:19,320
Speaker 1: He also wrote The Quants Fantastic, which is one of

1140
00:55:19,440 --> 00:55:22,359
Speaker 1: my favorite books describing what's taking place over the past

1141
00:55:22,400 --> 00:55:25,680
Speaker 1: ten years. But dark Pools is almost like The Quants

1142
00:55:25,719 --> 00:55:30,080
Speaker 1: Part two. It's very sophisticated, it's very fascinating. And he

1143
00:55:30,200 --> 00:55:32,759
Speaker 1: goes through the the kind of the the lineage of

1144
00:55:32,840 --> 00:55:34,960
Speaker 1: what HFC, where they came from, and he goes back

1145
00:55:35,000 --> 00:55:37,360
Speaker 1: to this firm called Island. It was J C. And

1146
00:55:37,600 --> 00:55:39,120
Speaker 1: and when we were at instant that that was one

1147
00:55:39,120 --> 00:55:41,279
Speaker 1: of our competitors. They were this small kind of firm,

1148
00:55:41,520 --> 00:55:42,920
Speaker 1: but a lot of the guys who were at Island

1149
00:55:43,000 --> 00:55:44,680
Speaker 1: kind of morphed out and they went to these h

1150
00:55:44,800 --> 00:55:47,360
Speaker 1: F T firms along the way. And it's almost the

1151
00:55:47,480 --> 00:55:49,680
Speaker 1: Island came out first and came up with this i'll

1152
00:55:49,800 --> 00:55:52,760
Speaker 1: pay you to add liquidity theory. It was their original

1153
00:55:52,880 --> 00:55:54,640
Speaker 1: thing that said I'll pay you to add liquidity, and

1154
00:55:54,680 --> 00:55:56,560
Speaker 1: now it's in every major stock exchange around the world.

1155
00:55:56,640 --> 00:55:59,520
Speaker 1: That was their competitive advantage. That's how they attracted all

1156
00:55:59,600 --> 00:56:03,040
Speaker 1: this volu The other thing about the book Dark Pools

1157
00:56:03,239 --> 00:56:06,040
Speaker 1: was I think I may be confusing this. I think

1158
00:56:06,200 --> 00:56:10,040
Speaker 1: that's where I first learned that Renaissance Technologies one of

1159
00:56:10,080 --> 00:56:13,960
Speaker 1: the most successful hedge funds of all time a year

1160
00:56:14,120 --> 00:56:18,839
Speaker 1: for thirty years. So successful they said to their outside clients, Hey, here,

1161
00:56:19,320 --> 00:56:21,759
Speaker 1: thanks for your money, but we're just gonna invest our

1162
00:56:21,800 --> 00:56:25,560
Speaker 1: own money. That's how you know it's not a nonsense return.

1163
00:56:26,400 --> 00:56:28,960
Speaker 1: We'll we'll, we'll take it from here. Um that they

1164
00:56:29,040 --> 00:56:31,919
Speaker 1: were one of the earliest high frequency traders and had

1165
00:56:32,400 --> 00:56:34,640
Speaker 1: before anyone in Wall Street even had a clue what

1166
00:56:34,719 --> 00:56:36,839
Speaker 1: they were doing. That needs to talk it. These guys

1167
00:56:36,920 --> 00:56:40,080
Speaker 1: had come up with a series of just ingenious ways,

1168
00:56:41,040 --> 00:56:43,560
Speaker 1: long long before everybody else had figured this out. And

1169
00:56:43,680 --> 00:56:45,480
Speaker 1: that's when it was easy pickings for them, right, it

1170
00:56:45,560 --> 00:56:48,120
Speaker 1: was easy to shoot, have a latency, our situation going on,

1171
00:56:48,200 --> 00:56:49,400
Speaker 1: and no one even knew what the heck it was.

1172
00:56:49,840 --> 00:56:53,280
Speaker 1: And now I mean to the lots of competition much harder.

1173
00:56:53,400 --> 00:56:56,680
Speaker 1: It's changed, and so we increase their cost so so

1174
00:56:56,880 --> 00:57:00,239
Speaker 1: they're not as profitable, No one's Jim Simmons retired, they're

1175
00:57:00,280 --> 00:57:03,680
Speaker 1: not as profitable, and I keep although they're still wildly successful,

1176
00:57:04,160 --> 00:57:07,000
Speaker 1: and I continue to read that h f T firms

1177
00:57:07,640 --> 00:57:12,600
Speaker 1: are just becoming less and less profitable, less competition and

1178
00:57:12,840 --> 00:57:17,040
Speaker 1: and a handful of big guys are now dominating the volumes.

1179
00:57:17,440 --> 00:57:21,240
Speaker 1: And whether that's Virtue or Get Go or a handful

1180
00:57:21,280 --> 00:57:24,600
Speaker 1: of other companies um which are also which you don't

1181
00:57:24,640 --> 00:57:27,960
Speaker 1: go into details of names and in broken markets, but

1182
00:57:28,120 --> 00:57:33,480
Speaker 1: Flashboys does so. So is this just naturally on the

1183
00:57:33,600 --> 00:57:36,440
Speaker 1: wane or what's going on? It's the term they call

1184
00:57:36,520 --> 00:57:39,280
Speaker 1: the race to zero. This is by far the race

1185
00:57:39,360 --> 00:57:41,680
Speaker 1: they're trying to zero out that that latency has gotten

1186
00:57:41,720 --> 00:57:44,120
Speaker 1: so squeezed. But here's the problem for an h f T.

1187
00:57:44,600 --> 00:57:45,920
Speaker 1: If I'm an h f T and you're an h

1188
00:57:46,000 --> 00:57:48,640
Speaker 1: f T and you decide to build a new microwave network,

1189
00:57:48,880 --> 00:57:51,960
Speaker 1: that's gonna get you one micro second fast or some

1190
00:57:52,120 --> 00:57:55,800
Speaker 1: data and the thousandths of a second million, a million, millions,

1191
00:57:55,960 --> 00:57:58,040
Speaker 1: a million, and whatever it may be. I'm just throwing

1192
00:57:58,120 --> 00:57:59,800
Speaker 1: numbers out there. But you're gonna be faster because you're

1193
00:57:59,840 --> 00:58:02,920
Speaker 1: gonn to spend ten million dollars, say, building a microwave

1194
00:58:03,000 --> 00:58:08,480
Speaker 1: between here in Chicago over the is gone microwave. It's

1195
00:58:08,520 --> 00:58:11,440
Speaker 1: now over the lasers, whatever else to exact. But if

1196
00:58:11,520 --> 00:58:13,480
Speaker 1: you do it, I have no choice but to do

1197
00:58:13,600 --> 00:58:15,800
Speaker 1: it if I want to remain competitive to you, which

1198
00:58:15,840 --> 00:58:18,080
Speaker 1: means now it's gonna cause me ten million, which means

1199
00:58:18,120 --> 00:58:20,160
Speaker 1: that's gonna be sucked right out of my profits. That's

1200
00:58:20,160 --> 00:58:22,680
Speaker 1: why you see in the margin start to collapse. So eventually,

1201
00:58:22,720 --> 00:58:24,360
Speaker 1: I think what will happen is the smaller ones are

1202
00:58:24,360 --> 00:58:26,600
Speaker 1: going to drift off. They will not be able to compete.

1203
00:58:26,760 --> 00:58:29,440
Speaker 1: The bigger ones will get bigger, which does become an

1204
00:58:29,480 --> 00:58:31,680
Speaker 1: issue of how much power they have. And but that's

1205
00:58:31,680 --> 00:58:33,640
Speaker 1: where you're going to have to have some regulations coming

1206
00:58:33,640 --> 00:58:35,880
Speaker 1: and say, okay, let's get some obligations on these guys

1207
00:58:35,920 --> 00:58:38,840
Speaker 1: as well. They're not going away, They're so h f

1208
00:58:38,880 --> 00:58:40,320
Speaker 1: T is here to stay. We're not. We're not going

1209
00:58:40,360 --> 00:58:43,080
Speaker 1: to see them disappear. But it's morphing. In other words,

1210
00:58:43,120 --> 00:58:46,720
Speaker 1: it's becoming something different. I guess they're trying to legitimatize themselves, right.

1211
00:58:47,040 --> 00:58:49,400
Speaker 1: It's it's like, well, is that is that gonna happen?

1212
00:58:49,440 --> 00:58:53,880
Speaker 1: Are they going to become obligated legitimized players or they

1213
00:58:53,960 --> 00:58:57,480
Speaker 1: still just snipers picking off you know, millipennies will be different.

1214
00:58:57,520 --> 00:58:59,360
Speaker 1: There are different ones. There's you know, h F T

1215
00:58:59,560 --> 00:59:01,800
Speaker 1: is different and it's really hard to bucket. And we've

1216
00:59:01,840 --> 00:59:04,000
Speaker 1: learned we were not the smartest when we first started

1217
00:59:04,040 --> 00:59:06,440
Speaker 1: doing this, and we've learned a lot a lot over

1218
00:59:06,480 --> 00:59:08,960
Speaker 1: the years, and we've learned that you can't just bucket

1219
00:59:09,080 --> 00:59:12,160
Speaker 1: HFC into one bucket. They are different types of players, right,

1220
00:59:12,320 --> 00:59:15,080
Speaker 1: and they do different things. But the smaller players who

1221
00:59:15,120 --> 00:59:17,280
Speaker 1: are there to pick off and just kind of kind

1222
00:59:17,320 --> 00:59:20,000
Speaker 1: of sniff out a little of what we call venue arbitrage,

1223
00:59:20,120 --> 00:59:21,920
Speaker 1: trying to pick you off because one venue slower it

1224
00:59:21,960 --> 00:59:24,080
Speaker 1: in the other that business is going away. They they're

1225
00:59:24,080 --> 00:59:26,760
Speaker 1: squeezing that out, that that that latency are that used

1226
00:59:26,760 --> 00:59:29,000
Speaker 1: to see when you trade bats versus nasadac or a

1227
00:59:29,080 --> 00:59:31,240
Speaker 1: dark pool, and the dark pools are now being more

1228
00:59:31,320 --> 00:59:34,240
Speaker 1: you know, investigated more like two attorney in general, that's changing.

1229
00:59:34,440 --> 00:59:37,040
Speaker 1: But market making, if you call it market making, you know,

1230
00:59:37,080 --> 00:59:39,320
Speaker 1: we call something different like you know, specialists and so on.

1231
00:59:39,640 --> 00:59:41,840
Speaker 1: But if they're market making a stock versus an e

1232
00:59:41,960 --> 00:59:44,240
Speaker 1: t F or an e t F versus the future,

1233
00:59:44,640 --> 00:59:45,920
Speaker 1: and they would still want to do that and they

1234
00:59:45,960 --> 00:59:50,280
Speaker 1: think they're profitable. Congratulations, good luck, but let's play by

1235
00:59:50,320 --> 00:59:53,000
Speaker 1: the same rules. That's amazing. I can remember as a

1236
00:59:53,120 --> 00:59:58,800
Speaker 1: trader being able to visually see different prices, or throwing

1237
00:59:58,840 --> 01:00:02,439
Speaker 1: out a quote an eighth below the bid, suddenly seeing

1238
01:00:02,480 --> 01:00:05,120
Speaker 1: I'm filled, and immediately selling it to somebody else an

1239
01:00:05,160 --> 01:00:08,919
Speaker 1: eighth higher. The markets were at least on an eighth level,

1240
01:00:09,160 --> 01:00:12,680
Speaker 1: so inefficient. And I'm talking the nineties, we're not talking

1241
01:00:12,880 --> 01:00:16,400
Speaker 1: post two thousand, um. But you could literally do that

1242
01:00:16,680 --> 01:00:18,280
Speaker 1: if you were quick on the keyboard and you had

1243
01:00:18,320 --> 01:00:21,720
Speaker 1: a couple of little macro set up. But you didn't

1244
01:00:21,760 --> 01:00:25,200
Speaker 1: even know it. If you call that high frequency trading,

1245
01:00:25,280 --> 01:00:30,160
Speaker 1: it's it's a human manual. Uh, just a quickly picking

1246
01:00:30,200 --> 01:00:33,640
Speaker 1: off an eighth on a thousand shares. Not exactly high Fenci.

1247
01:00:34,240 --> 01:00:36,600
Speaker 1: By the way, whenever you whenever you did that, it

1248
01:00:36,720 --> 01:00:39,240
Speaker 1: was it was just a rare. It was on a slow,

1249
01:00:39,320 --> 01:00:41,840
Speaker 1: boring day and you had nothing to do, and let's

1250
01:00:41,880 --> 01:00:44,280
Speaker 1: throw a few quotes out and see what happens. Suddenly

1251
01:00:44,320 --> 01:00:47,000
Speaker 1: you find yourself filled and you're you know, at worst

1252
01:00:47,080 --> 01:00:48,920
Speaker 1: you're out for a break even or an eighth loss.

1253
01:00:49,240 --> 01:00:51,360
Speaker 1: But it was just one of those things that and

1254
01:00:51,520 --> 01:00:54,280
Speaker 1: you always wondered, why can't you do this automatically. Why

1255
01:00:54,400 --> 01:00:57,320
Speaker 1: wasn't there technology for it? Never in a million years

1256
01:00:57,640 --> 01:01:00,880
Speaker 1: imagining it would be like it, it's got and it's

1257
01:01:00,920 --> 01:01:03,360
Speaker 1: gotten today. Yeah, and there is a have and have

1258
01:01:03,560 --> 01:01:06,400
Speaker 1: not in the world of trading now. The halves are

1259
01:01:06,400 --> 01:01:08,960
Speaker 1: the ones who can pay enough money to buy that

1260
01:01:09,120 --> 01:01:11,960
Speaker 1: co location facility, to rent the co location facility, to

1261
01:01:12,120 --> 01:01:14,920
Speaker 1: purchase the data feeds, to set up a microwave network,

1262
01:01:15,120 --> 01:01:17,600
Speaker 1: to do all the things they do right, So you

1263
01:01:17,720 --> 01:01:20,240
Speaker 1: have created a tiered process. And in the in the

1264
01:01:20,320 --> 01:01:22,360
Speaker 1: trading world, there's no doubt about it, there are those

1265
01:01:22,400 --> 01:01:24,560
Speaker 1: that do and those that don't. So you know, you're

1266
01:01:24,560 --> 01:01:26,240
Speaker 1: really not competing on the same level playing for it

1267
01:01:26,280 --> 01:01:28,520
Speaker 1: when it comes to technology. But look, I'm not going

1268
01:01:28,600 --> 01:01:30,240
Speaker 1: to argue against certain things, but i am going to

1269
01:01:30,360 --> 01:01:32,680
Speaker 1: argue when there's a dark pool out there and he's

1270
01:01:32,680 --> 01:01:35,880
Speaker 1: advantaging one client because that client is you know, an

1271
01:01:36,000 --> 01:01:37,600
Speaker 1: h F T and he's gonna say, I'm gonna give

1272
01:01:37,600 --> 01:01:40,120
Speaker 1: you a special look or something like that. Here we

1273
01:01:40,200 --> 01:01:42,360
Speaker 1: have a lot of issues. Or you know, couple aren't

1274
01:01:42,360 --> 01:01:45,720
Speaker 1: exchanges doing something similar by selling that data they are

1275
01:01:46,040 --> 01:01:48,800
Speaker 1: and data feeds what they called the proprietary data feeds

1276
01:01:48,840 --> 01:01:51,720
Speaker 1: we think should absolutely be looked at by the SEC

1277
01:01:51,880 --> 01:01:55,640
Speaker 1: should be why they have you. Here's what the exchange

1278
01:01:55,680 --> 01:01:57,880
Speaker 1: will say, Hey, Barry, you want to buy a data feed,

1279
01:01:57,960 --> 01:02:00,280
Speaker 1: I love to sell you one, buddy, ten thousand a month.

1280
01:02:00,320 --> 01:02:02,959
Speaker 1: Here you go purchase and anyway anybody can buy. They're

1281
01:02:02,960 --> 01:02:06,120
Speaker 1: going to turn around. But realistically a month doesn't mean

1282
01:02:06,200 --> 01:02:08,520
Speaker 1: that if I'm not buying a hundred chairs I'm gonna need.

1283
01:02:08,560 --> 01:02:10,640
Speaker 1: It doesn't help you. It doesn't help you. And then

1284
01:02:10,680 --> 01:02:12,880
Speaker 1: you also got to build the entire infrastructures to make

1285
01:02:12,920 --> 01:02:14,720
Speaker 1: that data. The data feeds just a fuel that runs

1286
01:02:14,760 --> 01:02:17,160
Speaker 1: the engine. You gotta still build the engine. Okay, But

1287
01:02:17,320 --> 01:02:19,120
Speaker 1: what they're giving in these data feeds is not just

1288
01:02:19,480 --> 01:02:21,480
Speaker 1: there's a lot of information and a lot of content.

1289
01:02:21,560 --> 01:02:23,560
Speaker 1: In fact, a couple of years ago we actually wrote

1290
01:02:23,560 --> 01:02:25,439
Speaker 1: a white paper where we talked two of the major

1291
01:02:25,520 --> 01:02:28,240
Speaker 1: exchanges giving out information on what it was supposedly a

1292
01:02:28,320 --> 01:02:31,280
Speaker 1: hidden order flow, so an institution could put in fifty

1293
01:02:31,600 --> 01:02:33,640
Speaker 1: shares hidden I don't want anybody seeing it. But they

1294
01:02:33,680 --> 01:02:37,000
Speaker 1: were putting tags on this flow, which basically allowed others

1295
01:02:37,080 --> 01:02:39,560
Speaker 1: to re engineer it. People who bought the data feeds

1296
01:02:39,720 --> 01:02:41,520
Speaker 1: and figured out that there was a buyer there. So

1297
01:02:41,680 --> 01:02:44,480
Speaker 1: let me ask you a question. Knowing the exchanges for

1298
01:02:44,720 --> 01:02:47,880
Speaker 1: profit are willing to do that, why would anybody trust

1299
01:02:47,960 --> 01:02:52,440
Speaker 1: anything they say? Aren't these just shameless salespeople who are

1300
01:02:52,520 --> 01:02:55,080
Speaker 1: basically going to say, I will sell this client down

1301
01:02:55,120 --> 01:02:57,800
Speaker 1: the road to that client because I can. Is that

1302
01:02:57,960 --> 01:03:02,120
Speaker 1: what actually took place? It their penalties where they're fine,

1303
01:03:03,080 --> 01:03:05,720
Speaker 1: the SEC didn't even look into. But the exchanges after

1304
01:03:05,800 --> 01:03:09,360
Speaker 1: they call grief from the institutional community, changed it voluntarily,

1305
01:03:09,760 --> 01:03:11,920
Speaker 1: just like the direct Edge exchange a couple of years

1306
01:03:11,960 --> 01:03:14,560
Speaker 1: back with doing flash orders where they were showing orders

1307
01:03:14,880 --> 01:03:17,680
Speaker 1: to a certain subset of clients for half a second,

1308
01:03:17,840 --> 01:03:20,360
Speaker 1: which is an eternity, and they were half a second,

1309
01:03:20,400 --> 01:03:21,880
Speaker 1: would you like to buy us? No? Okay, you go

1310
01:03:21,920 --> 01:03:23,920
Speaker 1: out in front and front of yourself. It was ridiculous.

1311
01:03:24,120 --> 01:03:26,400
Speaker 1: But but but but here's what they'll in their defense.

1312
01:03:26,400 --> 01:03:28,040
Speaker 1: They're gonna say, well, wait a second, show everything we

1313
01:03:28,120 --> 01:03:30,080
Speaker 1: do with public everything that we do has to be

1314
01:03:30,120 --> 01:03:33,000
Speaker 1: filed with the SEC after the fact three months later

1315
01:03:33,200 --> 01:03:36,720
Speaker 1: in forty pages of six point legal ease. But buddy,

1316
01:03:36,880 --> 01:03:40,640
Speaker 1: it's approved by the SEC. Why would they ever allow

1317
01:03:40,840 --> 01:03:43,760
Speaker 1: certain order types to be approved? Who is that the SEC?

1318
01:03:44,000 --> 01:03:46,120
Speaker 1: Looking at the exchange documents saying, yeah, that looks like

1319
01:03:46,160 --> 01:03:48,760
Speaker 1: it's fair to everyone. It's insane. So here's why we

1320
01:03:48,840 --> 01:03:51,280
Speaker 1: call it a broken market. Because you've got regulators who

1321
01:03:51,280 --> 01:03:53,880
Speaker 1: are allowing this, exchanges who are for profit h f

1322
01:03:53,960 --> 01:03:55,760
Speaker 1: t s who take advantage of it, and in between

1323
01:03:55,840 --> 01:03:57,880
Speaker 1: is the retail and institutional guy going what the hell

1324
01:03:58,000 --> 01:04:01,680
Speaker 1: is going on? That? That's unbelieva. I'm gonna end this

1325
01:04:01,920 --> 01:04:05,480
Speaker 1: segment with um a quote from my colleague Josh Brown,

1326
01:04:06,040 --> 01:04:08,720
Speaker 1: who said, and I love this quote, the best way

1327
01:04:08,800 --> 01:04:12,480
Speaker 1: to defeat high frequency trading is with low frequency trading.

1328
01:04:13,160 --> 01:04:15,400
Speaker 1: Meaning if if you're not in there, if you're not

1329
01:04:15,520 --> 01:04:18,160
Speaker 1: in there as a retail investor, if you're not in

1330
01:04:18,240 --> 01:04:20,840
Speaker 1: there trying to compete with these guys, you're not gonna

1331
01:04:20,920 --> 01:04:24,320
Speaker 1: lose them. That that's that's exactly right. Well, well, this

1332
01:04:24,440 --> 01:04:27,160
Speaker 1: has been really a ton of fun. Thank you so much. Joe, Say,

1333
01:04:27,200 --> 01:04:30,320
Speaker 1: how do your partner salve for us? We've been speaking

1334
01:04:30,400 --> 01:04:33,480
Speaker 1: with Joe. I was gonna say, Joe Arnock, we've been

1335
01:04:33,480 --> 01:04:37,520
Speaker 1: speaking with Joe so Lousy, author of Broken Markets and

1336
01:04:38,080 --> 01:04:42,560
Speaker 1: partner with Sal Arnuck of Famous Trading. If you enjoy

1337
01:04:42,760 --> 01:04:45,400
Speaker 1: this conversation, be sure to look up or down an

1338
01:04:45,440 --> 01:04:49,040
Speaker 1: inch or two on iTunes or on bloomberg View dot

1339
01:04:49,080 --> 01:04:51,959
Speaker 1: com or on Bloomberg dot com to see the rest

1340
01:04:52,080 --> 01:04:55,320
Speaker 1: of our podcasts. I'm Barry Ridults. You've been listening to

1341
01:04:55,480 --> 01:05:00,720
Speaker 1: Masters in Business on Bloomberg Radio. You're listening to Master's

1342
01:05:00,800 --> 01:05:04,040
Speaker 1: in Business with Barry Redholts on Bloomberg Radio.