[00:01.940 --> 00:02.640]  I'm Eric.
[00:02.920 --> 00:06.040]  This is my talk, Popping Shells, Popping Cells, it's a pun, right?
[00:06.160 --> 00:07.720]  Hunting for Evolence in the Stock Market.
[00:09.040 --> 00:15.560]  I'm here primarily with the Physical Security Village, formerly known as the Lock Bypass Village.
[00:16.400 --> 00:26.800]  And what we do is look at, you know, systems that are slightly outside the realm of, you know, your traditional cybersecurity, right?
[00:27.500 --> 00:30.320]  Physical infrastructure, doors, windows.
[00:31.940 --> 00:36.660]  And we really focus on the assumptions that are made in designing these systems, right?
[00:36.840 --> 00:46.820]  You've got your door latch that's designed to be able to allow the door to close and not, you know, open itself, but it also needs to be operable from the inside.
[00:47.560 --> 00:52.920]  So you can attack that maybe by slipping the latch under the door tool, pulling it down.
[00:53.440 --> 00:57.200]  And there's these edge cases that are made in designing physical systems.
[00:57.700 --> 01:07.520]  And there are assumptions that, you know, participants on the secure side of the physical system are all well-behaved and using this physical infrastructure as intended.
[01:08.640 --> 01:10.220]  But, you know, I've got something to admit.
[01:10.580 --> 01:12.140]  I'm a little bit ashamed of this.
[01:12.360 --> 01:13.640]  My day job is in finance.
[01:13.940 --> 01:15.060]  I know, I know.
[01:15.900 --> 01:18.640]  I'm not on Wall Street, I'm a Canadian, so I'm on Bay Street.
[01:18.640 --> 01:20.980]  But it's pretty much the same.
[01:22.500 --> 01:24.740]  But, you know, don't hold it against me.
[01:24.860 --> 01:26.740]  I really love my job.
[01:26.900 --> 01:35.620]  I work in a realm called early algorithmic trading, where we look at how stock markets work, how we interact with them, how we can make it better.
[01:37.080 --> 01:46.080]  And then, you know, naturally, as a follow-on to that, that I think is maybe of interest to this audience, is how market manipulation actually works.
[01:47.400 --> 01:49.820]  And that's what this talk is nominally about today.
[01:50.240 --> 01:52.660]  I think I wrote a pretty slick abstract.
[01:53.680 --> 02:02.760]  The reality is that when I started going through the literature on market manipulation, it's pretty sparse, right?
[02:03.400 --> 02:03.980]  Enforcement...
[02:03.980 --> 02:06.440]  And I think this should be a main takeaway for this talk.
[02:06.560 --> 02:08.360]  Enforcement is actually really good.
[02:08.360 --> 02:12.500]  You know, the SEC, the CFTC, they've got fantastic data.
[02:13.820 --> 02:15.900]  And, you know, years and years of data.
[02:16.140 --> 02:22.160]  And so, you know, today they're prosecuting cases for things that were done in like 2006 through 2015.
[02:23.160 --> 02:26.520]  So, you know, the wheels of justice turn slowly but grind finally.
[02:27.380 --> 02:30.920]  The end result of this is that really there isn't too much to talk about market manipulation.
[02:30.920 --> 02:36.080]  I'll talk through a few specific cases that you should not try at home.
[02:36.600 --> 02:40.900]  But I think I'll talk a bit more about how, you know, exchanges are built.
[02:41.020 --> 02:45.000]  And a bit more about the strategy of how folks interact with them.
[02:47.440 --> 02:48.680]  Two main disclaimers.
[02:48.820 --> 02:50.440]  One is that I'm on vacation right now.
[02:50.680 --> 02:53.220]  I really love my job and I love talking about it.
[02:53.340 --> 02:55.260]  But I'm not working right now.
[02:57.340 --> 03:02.180]  Nothing that I write was approved or screened in any way.
[03:02.440 --> 03:04.000]  I am here on my own.
[03:04.200 --> 03:05.080]  Paid my own travel.
[03:06.900 --> 03:11.720]  And second, naturally, you know, any instance of market manipulation is illegal.
[03:12.040 --> 03:19.820]  In the U.S., it will be under Title 15, Section 78, which deals specifically with manipulation of securities.
[03:21.240 --> 03:28.460]  You know, and even if your personal ethical framework, you know, objects to that and says, well, you know, that's fine.
[03:29.800 --> 03:35.920]  Again, the reality is that the regulator is exceptionally good and will fine you.
[03:36.060 --> 03:37.500]  There are multiple lines of defense, right?
[03:37.560 --> 03:38.280]  There's your broker.
[03:38.460 --> 03:39.260]  There's the regulator.
[03:39.560 --> 03:44.900]  And then at the end of the day, anyone who you seek to rip off is probably much more sophisticated than you.
[03:44.900 --> 03:50.500]  And is probably going to look at why they generated a loss, investigate that and report it.
[03:50.660 --> 03:53.080]  So, you know, don't try it at home.
[03:53.320 --> 03:56.040]  The other thing too, everything I say today is open-source.
[03:56.040 --> 03:59.460]  If you know where to look, you can find it easily.
[03:59.680 --> 04:04.080]  There's nothing, no IP here, no original research.
[04:06.120 --> 04:08.640]  So why am I doing this at great personal risk?
[04:09.140 --> 04:10.120]  Someone was wrong.
[04:11.520 --> 04:16.040]  I want to say what first gave me this idea, it was DEFCON.
[04:16.160 --> 04:17.900]  I want to say it was the 2021 con.
[04:18.260 --> 04:19.160]  Fantastic time.
[04:19.160 --> 04:21.620]  If any of you were there, the pool party was amazing.
[04:23.400 --> 04:29.980]  And I was ripping around Las Vegas in a rented U-Haul pickup truck, mostly moving doors around.
[04:30.480 --> 04:33.960]  But someone who I just met needed a ride to Walmart to pick something up.
[04:33.980 --> 04:34.900]  So I gave him a ride.
[04:35.600 --> 04:39.760]  We make small talk and I make the mistake of what I, mentioning what I do for a living.
[04:39.920 --> 04:40.820]  And I say I work in finance.
[04:41.060 --> 04:51.160]  So, well, that's kind of, you know, you say you work in cybersecurity and probably your parents or your friends or whoever you just met ask you for, you know, help making your printer work.
[04:52.560 --> 04:57.780]  It's somewhere between that and, like, trying to talk to your uncle who gets all their news from, like, Truth Social.
[04:58.320 --> 05:07.120]  Because the conspiracy theories start coming and, you know, there's no way to correct someone as you're driving them around in a U-Haul pickup truck.
[05:08.140 --> 05:23.740]  So, you know, I want to try and give you some introduction to how the market works, some resources that you can look into yourself to hopefully gain, you know, a deeper appreciation for this fascinating part of late-stage capitalism.
[05:24.560 --> 05:40.940]  But, like any, you know, esoteric strategy game, you go to a party, it's a Tuesday night, you meet up at 6, your friend spends the first two hours explaining the rules of the board game, and you don't actually end up playing until, you know, everyone's tired and it's about time to go home.
[05:42.140 --> 05:44.320]  So that's largely the structure of the stock.
[05:46.960 --> 05:51.000]  A lot of folks, sort of, when they picture the stock market, they have this belief.
[05:52.040 --> 06:02.920]  You know, you go on CNBC and you see the trading floor, like the one in the picture here, you know, where there are people shouting instructions and giving gang signs.
[06:04.700 --> 06:12.060]  Today, it's a great television soundstage, and they keep it open, I think, truthfully, because, you know, they like having it on television and issuers.
[06:12.660 --> 06:14.880]  If you're the CEO of a company, it's really cool to ring the bell.
[06:16.320 --> 06:19.600]  But, you know, at some point, there was some truth to this.
[06:19.740 --> 06:32.260]  There would be, you know, a specialist standing at one of those posts you see there, you know, probably wearing a top hat and, you know, a monocle, who is the specialist in a given security.
[06:33.000 --> 06:38.920]  You know, Mr. Monopoly here might be the specialist in, you know, General Consolidated Corporation.
[06:39.560 --> 06:44.580]  And he's always willing, you know, to sell General Consolidated at $10 and to buy at $8.
[06:45.020 --> 06:53.580]  So if you, you know, call up your, you know, broker in the 1980s, you know, Gordon Gekko on the telephone, and you ask him to buy stock in General Consolidated.
[06:54.100 --> 06:59.440]  You know, he'll write that down on a ticket, put it in, you know, a vacuum pipe, send it down to the floor broker.
[06:59.640 --> 07:06.420]  The floor broker will go out, go to that specialist's post and say, you know, they want to buy at $10.
[07:06.520 --> 07:08.480]  And it's that person who's making the market.
[07:11.080 --> 07:13.620]  Today, again, it's television soundstage.
[07:13.980 --> 07:18.840]  This all actually happens at data centers throughout the New Jersey Equity Triangle.
[07:18.840 --> 07:27.060]  So the data actually occurs, or the trading actually occurs, primarily in Carteret, New Jersey, Secaucus, New Jersey, and Malwa, New Jersey.
[07:27.240 --> 07:37.140]  Three places that I've never been to, but it seems to be my obsession in life, these three cities.
[07:38.440 --> 07:40.840]  You know, stop doing electronic trading.
[07:41.000 --> 07:43.100]  Computers were not meant to trade stocks.
[07:43.460 --> 07:45.700]  Dreamed up by the utterly deranged.
[07:46.920 --> 07:52.460]  And what's happening on these computers, essentially they're managing a FIFOQ.
[07:53.480 --> 08:05.220]  At each price level, there are, you know, a list of orders that are willing to sell at a given price level, the list of orders that are willing to buy at a price level.
[08:08.520 --> 08:13.940]  So, you know, you go in and you say, I want to buy stock.
[08:14.120 --> 08:22.560]  You have the option kind of of either joining the queue at the back at $5 and then hoping that, you know, someone comes down and trades with you.
[08:22.560 --> 08:25.520]  Or you can send a market order and buy the stock at $5 and 10.
[08:26.200 --> 08:33.460]  And the exchange matching engine is, you know, pairing off the buy and sell orders at a given price.
[08:33.960 --> 08:35.460]  And working through the queue.
[08:37.260 --> 08:45.460]  The actual way that you enter orders into the system is through something called the fixed protocol, which is like an absolute chimera.
[08:45.660 --> 08:46.700]  It's a terrible protocol.
[08:47.460 --> 08:50.200]  It uses a tag value system.
[08:50.800 --> 09:03.860]  So you define, you've got like a key that defines what that value corresponds to, an equal sign, the actual value, and then 0x01 is the delimiter.
[09:06.300 --> 09:11.600]  There's a always 8, which denotes the beginning of the message.
[09:12.220 --> 09:14.480]  9, which is the character length.
[09:14.480 --> 09:17.180]  35 there is the instruction type.
[09:17.280 --> 09:18.620]  So in this case, it's a new order.
[09:22.280 --> 09:24.740]  49, sender ID, so that's me.
[09:25.600 --> 09:29.320]  52, timestamp 56 is where I'm sending the order to.
[09:31.160 --> 09:32.200]  Order ID.
[09:34.260 --> 09:36.600]  21 equals 2 is a handling instruction.
[09:36.780 --> 09:37.760]  38 is the quantity.
[09:39.860 --> 09:42.480]  40 says that it's a market order.
[09:45.680 --> 09:47.360]  55 is the ticker.
[09:47.600 --> 09:47.620]  55 is the ticker.
[09:47.620 --> 09:49.160]  And then 10 is a checksum.
[09:52.380 --> 09:55.520]  So normally sent over your usual run-of-the-mill TCP.
[09:56.020 --> 10:03.880]  So your broker will have a TCP connection with an exchange gateway that they're, you know, taking all the client orders and sending them across.
[10:06.500 --> 10:08.620]  Generally, there's kind of a double acknowledgment.
[10:08.760 --> 10:10.280]  So there's the TCP act, right?
[10:10.420 --> 10:13.380]  But there's also an acknowledgment that comes back from every fixed message.
[10:14.540 --> 10:16.020]  So a lot of overhead.
[10:16.260 --> 10:18.640]  Generally, you've got a session two that's running all day.
[10:18.640 --> 10:23.820]  So you have, you know, session initiation message, normally around 3 or 4 in the morning.
[10:23.980 --> 10:29.960]  And then a session disconnect message, generally around 6, 7 p.m., depending.
[10:32.760 --> 10:34.540]  And this is wildly inefficient.
[10:34.920 --> 10:37.000]  This is literally ASCII, right?
[10:37.000 --> 10:40.100]  There's no clever encoding here, ASCII messages.
[10:40.360 --> 10:43.860]  It's great because if you're trying to debug something, you can see exactly what's happening.
[10:44.460 --> 10:47.980]  But consider the fact that a lot of this is implemented now in FPGAs.
[10:48.200 --> 10:57.880]  And so you have some poor, poor, very sad FPGA engineers who are writing Verilog and VHDL to create ASCII messages on the wire.
[10:57.880 --> 11:02.380]  They're generally also writing like their own Ethernet drivers.
[11:02.940 --> 11:04.020]  It's disgusting.
[11:04.640 --> 11:08.420]  There are almost certainly vulnerabilities in the fixed protocol.
[11:09.880 --> 11:15.280]  Surely, if you sent malformed messages to a lot of brokers, a lot of exchanges, you could do a lot of damage.
[11:16.800 --> 11:21.800]  Realistically, this continues to work because of logical controls, right?
[11:23.420 --> 11:25.960]  Sophisticated, trusted counterparties within a web of trust.
[11:27.440 --> 11:29.780]  And you'll immediately be booted off.
[11:29.900 --> 11:35.160]  And then there's like a whole other legal system, right, to correct your malfeasance.
[11:36.620 --> 11:39.420]  What I presented here is actually a gross simplification, right?
[11:39.480 --> 11:40.720]  Because there's not just one list.
[11:40.940 --> 11:46.600]  There's not just, you know, the New York Stock Exchange, the single server and the single data center that's matching up all the orders.
[11:47.140 --> 11:50.540]  They're actually distributed across the Northeast.
[11:51.780 --> 11:54.240]  And these are generally physically distant.
[11:55.520 --> 11:59.300]  You know, NYSE to NASDAQ, so Malwa to Carteret.
[12:00.140 --> 12:04.280]  At the theoretical limit, so speed of light, you're looking at about 182 mikes.
[12:05.640 --> 12:11.840]  Malwa to Secaucus, where NY4 is, where a lot of other exchanges are, you're looking at a theoretical minimum of 116 mikes.
[12:12.040 --> 12:15.360]  And then NY4 to Carteret, about 94 mikes.
[12:15.520 --> 12:17.000]  That's the theoretical limit.
[12:17.160 --> 12:19.140]  In reality, the latency is going to be a bit greater.
[12:20.820 --> 12:22.520]  Why does this actually matter?
[12:25.220 --> 12:27.180]  This is one of the things that caused me massive frustration.
[12:28.400 --> 12:30.820]  In around, I want to say, like it was 2014, right?
[12:31.060 --> 12:40.160]  There was this great fiction novel published, Flash Boys, where there was this, you know, idea presented of latency arbitrage.
[12:40.740 --> 12:42.620]  And it doesn't really happen anymore.
[12:42.620 --> 12:47.200]  But somehow, it still sticks in the, you know, this cultural understanding of the stock market works.
[12:47.540 --> 12:51.560]  The thinking is, if the price changes abruptly on NYSEE, right?
[12:51.880 --> 12:53.600]  Maybe there are some terrible job numbers.
[12:53.680 --> 12:55.380]  It turns out we're going to a huge recession.
[12:56.600 --> 12:58.060]  Everyone's losing their jobs.
[12:58.760 --> 13:00.140]  The economy's in freefall.
[13:00.340 --> 13:02.200]  Great news for the market, right?
[13:03.020 --> 13:05.080]  Because it means that interest rates are coming down.
[13:05.600 --> 13:09.680]  So, you know, we have a tick up on NYSEE.
[13:11.100 --> 13:17.460]  But because, you know, these systems are physically distant, it takes time for this to propagate to other exchanges.
[13:17.460 --> 13:28.220]  So, if, you know, you are offering to sell stock at 509 on, you know, one of the, let's say, at NASDAQ.
[13:28.420 --> 13:30.200]  Or at 510 at NASDAQ.
[13:31.980 --> 13:35.720]  You observe that change in price at NYSEE.
[13:35.900 --> 13:40.140]  You send, you know, a price correction message down to NASDAQ.
[13:40.240 --> 13:45.160]  And you're now racing with someone who's trying to buy from you at your still price at 510.
[13:49.100 --> 13:55.880]  In general, you know, if you're, you know, kind of a mid-tier broker, you're probably using fiber to do this.
[13:56.460 --> 14:02.240]  And the photon's bouncing around and it's going through, you know, some really, really high-performance switching equipment.
[14:02.420 --> 14:05.300]  But nevertheless, you're constrained by the speed of light and fiber.
[14:06.680 --> 14:13.720]  The real sophisticated high-frequency traders are generally using microwave at this point, right?
[14:13.720 --> 14:18.240]  Because it's straight line, photons not bouncing around, slightly higher speed of light in the air than in glass.
[14:19.480 --> 14:22.220]  And so, you know, there sometimes is this racing effect.
[14:23.600 --> 14:33.860]  But, you know, what is discussed in that fiction novel is this idea that you can, like, buy on one and immediately sell on the other.
[14:33.860 --> 14:36.760]  And like that, I want to make it clear that that doesn't happen.
[14:41.300 --> 14:57.040]  But the microwave thing also gives rise to this really fascinating thing where rain and, you know, atmospheric moisture will absorb and attenuate your microwave signal.
[14:57.040 --> 14:59.040]  And so you have this fascinating result.
[14:59.220 --> 15:06.740]  There was a great paper by Andre Schilko, a prof out of Laurier, Every Cloud is a Silver Lining.
[15:06.840 --> 15:13.160]  And he actually found that when it is raining over Ohio, it becomes cheaper to trade on equities markets.
[15:13.160 --> 15:27.300]  Because the people who are trying to run the low-latency arbitrages off, you know, one of the Chicago futures exchanges in Aurora, Illinois, their microwave goes down, they fail over to fiber, it's a level playing field for everyone, and you actually have an improvement in market quality.
[15:28.220 --> 15:36.440]  And so you can imagine some, you know, attacks there where, you know, maybe you put up a big, like, aluminum sheet next to one of the antennas.
[15:38.520 --> 15:52.440]  There was a case several years ago where there was, you know, a very large lawsuit over who could own a plot of land between the Aurora data center in the path of Malwa.
[15:55.180 --> 16:00.120]  And, you know, so it, like, it's coming down to where the real estate is.
[16:07.020 --> 16:15.380]  Another fascinating feature of this is that, and there was another great talk about this, presented at the Security Traders Association of Chicago.
[16:16.940 --> 16:20.520]  We've got this in the Northeast, but naturally there are other exchanges around the world.
[16:21.920 --> 16:27.520]  And you can't microwave across the ocean because you'd need, you know, probably ships floating all the way along.
[16:27.880 --> 16:31.260]  And, you know, fiber, as we've already established, way too slow, right?
[16:31.380 --> 16:32.880]  That's like 1990 stuff.
[16:35.360 --> 16:44.800]  So, there have been many experimental licenses filed with the FCC for experimental use of HF Spectra.
[16:45.160 --> 16:48.360]  And many of these have been filed by some of these trading firms.
[16:51.600 --> 17:04.220]  And so, if any of you are in the amateur radio community, you might recall about a year and a half ago, there was a petition to the FCC to allow industrial use of HF Spectra.
[17:05.180 --> 17:14.500]  It was, you know, and this is my personal opinion, it was largely because a lot of these firms want to stop coming up to the limits on what they can do with an experimental license.
[17:15.060 --> 17:16.540]  But, yeah, that's all connected.
[17:17.480 --> 17:19.440]  I want to stress, though, that this is all fine.
[17:19.580 --> 17:22.560]  Everything I've talked about so far, like, them's the rules, right?
[17:24.260 --> 17:31.500]  And before I get into, like, the actual meat and potatoes of market manipulation, I want to go back to our limit order book.
[17:31.700 --> 17:35.300]  Suppose you... and I'm going to stick with the Vaporwave theme.
[17:35.680 --> 17:37.260]  You're Marvin the Manipulator.
[17:38.120 --> 17:39.660]  You're that bust there.
[17:39.840 --> 17:42.860]  You want to buy 100 shares of, you know, stock XYZ.
[17:42.940 --> 17:43.920]  You really have two options.
[17:43.920 --> 17:50.760]  One, you can join the bid at $5, join the queue at the back of the FIFO queue, or you can cross the spread.
[17:50.920 --> 18:00.720]  You can send a market order and buy immediately at $5.10 from Ian the investor, the Vaporwave dolphin there, who's selling.
[18:04.700 --> 18:07.360]  In, you know, the market order case, you have immediacy.
[18:07.660 --> 18:11.080]  In the limit order case, you don't necessarily have immediacy.
[18:11.080 --> 18:13.780]  But you're probably getting a price in equilibrium.
[18:14.840 --> 18:23.100]  So throughout, you know, the rest of the explanation, we're going to have Ian the investor, who's the guy you're trading against, and then Marvin the Manipulator, who's you.
[18:23.540 --> 18:26.860]  There are others in this market, but we can kind of abstract those away.
[18:28.960 --> 18:30.920]  You actually have a multitude of options, right?
[18:30.940 --> 18:33.880]  You don't need to join the bid at $5.00, you can join at $5.01.
[18:34.080 --> 18:39.740]  In which case, if Marvin decides, hey, actually, you know, I want to sell right now, he'll trade with you first.
[18:41.880 --> 18:48.120]  And, you know, this is what actually happens when you're on your, you know, discount brokerage app and you send a market order.
[18:50.660 --> 18:55.600]  Once it gets through, like, the broker systems, it will go out and, you know, buy on the offer.
[18:57.040 --> 19:04.820]  Likewise, if you send out a limit order, because of a thing called the order exposure rule, your passive limit order will actually go on to the exchange.
[19:05.160 --> 19:06.000]  That's a requirement.
[19:06.000 --> 19:09.540]  So, like, off your discount brokerage app, you'll actually go on to this board.
[19:11.820 --> 19:14.100]  But, you know, there's kind of a danger here.
[19:14.620 --> 19:22.840]  And you say, well, you know, if I can buy at $5.01, and I know eventually Marvin's going to want to sell to me, why would I pay those extra $0.09?
[19:25.560 --> 19:30.620]  And, you know, to explain that, it comes down to a lesson in economics.
[19:30.840 --> 19:32.460]  This idea of adverse selection.
[19:33.360 --> 19:35.460]  So, you're looking to buy a used car.
[19:36.280 --> 19:39.140]  And there are two Danny DeVitos from Matilda there.
[19:39.340 --> 19:42.660]  And they're both selling a car, and they both say the car is $25,000.
[19:43.620 --> 19:47.400]  And you can choose, you know, one or both of them to buy your used car from.
[19:47.400 --> 19:51.440]  But one of these, you know, Frank Reynolds always lies, and the other always tells the truth.
[19:52.860 --> 19:55.800]  So, but you know that one of them's lying.
[19:56.640 --> 19:59.000]  One of these cars isn't worth $25,000.
[19:59.460 --> 20:02.580]  And probably the average value of a used car is $15,000.
[20:02.920 --> 20:04.040]  Obviously, it's not, right?
[20:04.120 --> 20:06.560]  Used cars post-COVID, it's insane.
[20:08.160 --> 20:11.560]  But you make an offer to buy the car for $15,000.
[20:12.660 --> 20:15.580]  And Danny DeVito on the left there say, you know, that's a great deal.
[20:15.740 --> 20:17.380]  I'll throw in an egg for this drawing time.
[20:18.280 --> 20:20.460]  Danny DeVito on the right, you know, you're the trash man.
[20:20.540 --> 20:21.140]  You eat garbage.
[20:21.520 --> 20:22.240]  Get out of here.
[20:22.520 --> 20:24.540]  So, you make the deal with the Danny DeVito on the left.
[20:25.280 --> 20:28.900]  You put sawdust in the transmission, the car is garbage, and you got ripped off.
[20:29.500 --> 20:33.660]  And the one who actually had the great car walks away from the deal and you missed out on a good car.
[20:38.890 --> 20:42.330]  Because I was, you know, I initially had this idea of using Danny DeVito on this side.
[20:42.410 --> 20:43.810]  This is kind of a side point.
[20:43.910 --> 20:45.130]  I recently bought a used car.
[20:45.370 --> 20:49.710]  If you search trustworthy used car salesman on Google Images, you get that picture.
[20:52.530 --> 20:55.350]  Which I thought was pretty funny, but is a total digression.
[20:55.450 --> 20:57.610]  Why does this actually matter in the context of trade?
[21:00.790 --> 21:02.290]  You've got your bid at 501.
[21:02.910 --> 21:04.070]  You're standing out there.
[21:04.070 --> 21:05.790]  You're willing to trade at 501.
[21:05.950 --> 21:07.070]  You're confirmed there.
[21:08.650 --> 21:13.210]  If Ian the investor decides to sell to you, you can't back away from that.
[21:13.350 --> 21:13.590]  Right?
[21:13.690 --> 21:15.650]  You're committed to buying at 501.
[21:18.290 --> 21:23.630]  But Ian's decision of whether or not to sell to you at 501, they aren't constrained.
[21:23.770 --> 21:24.710]  They can choose to do that.
[21:24.790 --> 21:25.690]  They can choose not to.
[21:26.350 --> 21:28.510]  But Ian the investor is smart, right?
[21:28.590 --> 21:29.370]  Dolphins are smart.
[21:29.910 --> 21:34.010]  And he'll only choose to sell to you if he knows that the price is about to go down.
[21:34.010 --> 21:37.810]  If he believes, if he estimates, if he forecasts, the price is about to go down.
[21:38.110 --> 21:48.370]  And so there's a corollary there where you're only going to end up buying from Ian if it's a losing trade for you.
[21:49.610 --> 21:52.390]  So that's why you don't always do that.
[21:54.150 --> 21:59.030]  But, you know, you still don't want to pay those extra $0.09.
[21:59.790 --> 22:05.210]  So you start to think, how can I, you know, induce Ian to sell to me?
[22:06.330 --> 22:07.670]  And say you want to buy a lot more.
[22:07.790 --> 22:08.950]  You want to buy 5,000 shares.
[22:09.970 --> 22:13.190]  So you join the bid for 5,000 shares at $5.
[22:13.310 --> 22:16.150]  That's a large order in this market, right?
[22:17.550 --> 22:22.470]  And Ian looks at that and he says, geez, like someone wants to buy 5,000 shares of the stock.
[22:23.070 --> 22:25.530]  They must know that the price is about to go up.
[22:25.710 --> 22:30.810]  They might have like a really strong belief that the price is about to rise.
[22:31.770 --> 22:36.250]  And if they have that strong belief, probably that means that I, as Ian the investor, am wrong.
[22:37.790 --> 22:42.630]  So in all likelihood, Ian's going to cancel his offers at 5,000.
[22:43.930 --> 22:45.630]  Place a new one maybe up at 5,000.
[22:46.110 --> 22:50.050]  Maybe someone else in the market, one of those other participants, sees your big bid.
[22:50.470 --> 22:51.890]  Says the price is about to go up.
[22:51.970 --> 22:52.870]  I'd better buy now.
[22:53.390 --> 22:56.610]  And all of a sudden, by placing your order there, you've driven the price up.
[22:56.670 --> 22:57.570]  You've scared Ian away.
[23:02.750 --> 23:05.650]  So, you know, let's try something else there.
[23:06.430 --> 23:14.350]  And this is where we get into the idea of spoofing, which is, you know, the one canonical example of market manipulation that I'll really go through and harp on today.
[23:15.150 --> 23:17.310]  You put it in your 5,000 share buy order.
[23:17.950 --> 23:22.730]  But you also put in a 10,000 share sell order alongside Ian at 510.
[23:24.030 --> 23:26.150]  And now Ian's confused, right?
[23:26.310 --> 23:27.910]  He says, well, you know what?
[23:27.910 --> 23:31.110]  It looks like there's a really motivated seller here.
[23:31.410 --> 23:35.830]  I bet they're going to, you know, cross the spread and sell at $5 soon.
[23:36.450 --> 23:38.050]  I want to get ahead of that.
[23:39.910 --> 23:49.870]  And so, you know, instead then, either Ian or someone else takes that 10,000 share signal and sells to you.
[23:50.070 --> 23:53.670]  And you've just got, you know, you've just bought your stock without impacting the price.
[23:55.490 --> 23:57.370]  Obviously, you don't actually want to sell that stock.
[23:57.470 --> 23:58.970]  So afterwards, you cancel your order.
[23:59.410 --> 24:01.870]  It was just there to induce that selling.
[24:02.810 --> 24:10.350]  But the danger is, you know, the danger there is, what if there actually is a buyer there and they buy from you at 10,000, right?
[24:10.590 --> 24:12.770]  While your order is out, because again, it's firm.
[24:14.750 --> 24:22.970]  So another interesting special case of it called layering is where instead of putting it, you know, right on the first offer, you put a little bit higher up.
[24:23.150 --> 24:23.990]  Say at $5.12.
[24:24.250 --> 24:27.310]  So then, you know, to buy from you, they've got to go through multiple levels.
[24:27.430 --> 24:28.750]  In all likelihood, you won't get him.
[24:32.530 --> 24:33.410]  And that's illegal.
[24:39.270 --> 24:41.590]  You know, there are many examples of it.
[24:41.830 --> 24:45.550]  Sort of the canonical example that comes to mind is a case that the U.S.
[24:45.650 --> 24:46.750]  Department of Justice brought.
[24:46.750 --> 24:56.250]  I want to say it was in 2020 against an entire trading desk at a large American investment bank who were trading metals.
[24:56.250 --> 24:59.150]  So think about things like primarily gold.
[24:59.750 --> 25:01.250]  You know, maybe they won't buy gold.
[25:01.350 --> 25:06.990]  But they actually would sometimes put in fake sell orders to induce sellers to their levels.
[25:08.730 --> 25:10.930]  Again, this case took place 2006...
[25:10.930 --> 25:13.950]  The actual events were around 2006 to 2015.
[25:14.610 --> 25:21.030]  The SEC and CFTC and Department of Justice prosecuted it in 2020 because they had all the data.
[25:21.070 --> 25:23.470]  And I think in this case they also had a whistleblower.
[25:24.770 --> 25:34.070]  Side note, whistleblower, great retirement plan because part of the settlement goes into a whistleblower fund and it goes to the whistleblower.
[25:34.810 --> 25:38.870]  And some of these awards can be quite large in the order of millions of dollars.
[25:40.150 --> 25:48.050]  In this case though, for the folks doing the market manipulation, they end up spending some time in prison.
[25:48.890 --> 25:50.570]  I think it was about six months.
[25:50.910 --> 25:54.030]  Which, you know, I don't think is a great deal.
[25:58.590 --> 26:02.270]  Okay, so we can look at another interesting case.
[26:03.870 --> 26:05.750]  There's a lot of talk in the media about dark pools.
[26:05.950 --> 26:08.730]  And I think they get a really bad rap because they kind of sound evil.
[26:08.870 --> 26:13.870]  Dark pools, you picture like, you know, a spooky swimming pool with like a shark swimming around in it.
[26:16.710 --> 26:24.170]  Dark pools in the simplest terms are marketplaces where there is no pre-trade transparency, but there's complete post-trade transparency.
[26:24.590 --> 26:30.330]  So, you want to buy 5,000 shares again, but you don't want to scare away any of the sellers.
[26:30.330 --> 26:36.530]  So, you can place it in something called a dark pool where you're resting at the midpoint of the bid-ask spread.
[26:36.770 --> 26:42.290]  You're willing to pay half that spread to avoid being, you know, seen out in the market openly beforehand.
[26:43.890 --> 26:53.250]  And then, you know, if Ian, the investor, decides that they want to sell, and they're willing to sell at $5, the trade will actually come and trade with you first at $5.05.
[26:53.550 --> 26:58.090]  So, you know, Ian gets a slightly higher price, that extra five cents.
[26:58.610 --> 27:02.190]  You don't have your market impact, but you pay up a little bit for that.
[27:04.710 --> 27:07.670]  After the trade happens, it gets published to the entire market.
[27:07.670 --> 27:10.290]  So, everyone knows that this trade happened after the fact.
[27:11.810 --> 27:14.710]  No one knows, in theory, who initiated that trade.
[27:14.710 --> 27:18.210]  So, who was there first, and who was in dark, except Ian.
[27:18.490 --> 27:21.770]  Obviously, Ian knows that he was selling, and he got traded with at $5.05.
[27:21.770 --> 27:23.710]  So, there's probably another buyer in there.
[27:23.850 --> 27:28.870]  So, Ian can get some signaling off of that, which is some of the, you know, fun strategy that I really enjoy.
[27:29.190 --> 27:31.630]  But really, there's nothing nefarious about dark pools there.
[27:32.990 --> 27:39.950]  Unless, you know, unless you bring in, you know, some of your market manipulation skills in there.
[27:44.230 --> 27:49.250]  The way that many dark pools work is that your order is always going to be pegged to the midpoint.
[27:49.410 --> 27:56.310]  So, if the bid and the offer both move up, if the market moves up, your order will automatically adjust upwards to stay at the midpoint.
[27:56.310 --> 28:03.310]  And the reason it's at the midpoint, again, is if, you know, you were at $5 willing to buy.
[28:03.310 --> 28:06.150]  And suppose that was dark, and Ian came to sell to you at $5.
[28:06.190 --> 28:13.190]  The fact that the trade happened at $5 would obviously mean that, like, the person who was there first was the one at $5.
[28:13.190 --> 28:17.110]  So, it stays at midpoint to try and, like, hide that trade directionality.
[28:20.270 --> 28:26.950]  But, you know, in theory, if you know it's always going to be at midpoint, maybe you put a fake offer in at $5.06.
[28:27.030 --> 28:27.990]  It doesn't have to be large.
[28:27.990 --> 28:35.470]  It just has to be enough to bring that best offer price down to the point where the midpoint is now, say, $5.03 in this case.
[28:36.430 --> 28:36.830]  Right?
[28:37.170 --> 28:38.750]  Because it's a sliding scale.
[28:38.750 --> 28:43.470]  So now, when Ian, the investor, comes to sell, he'll sell to you at $5.03.
[28:43.470 --> 28:46.670]  Instead of at $5.05, you get another $0.02 per share there.
[28:47.670 --> 28:48.450]  Also illegal.
[28:49.310 --> 28:53.970]  There's an example in Canada where some participant was doing that exactly.
[28:54.530 --> 28:56.050]  They'd have an order in dark.
[28:56.170 --> 29:03.790]  They'd put a fake offer in, improving the best offer, to slide down their midpoint order.
[29:04.830 --> 29:08.410]  Then when there's a trade against them, they'd immediately cancel that offer.
[29:09.850 --> 29:17.890]  In a lot of these cases, like, the real giveaway is that there are either trade orders that weren't really intended to be executed.
[29:18.370 --> 29:18.850]  Right?
[29:19.010 --> 29:20.190]  That spoofed order.
[29:20.770 --> 29:23.150]  Or doing a trade that's uneconomical.
[29:23.450 --> 29:23.570]  Right?
[29:24.030 --> 29:28.010]  Buying too high, selling too low to try and, like, move the price momentum.
[29:30.230 --> 29:35.990]  There are some other less interesting schemes, I think, because they rely less on, like, the mechanics and strategy.
[29:37.210 --> 29:50.050]  Probably the worst in a recent memory is the price fixing of LIBOR, the London Interbank Offered Rate, which is probably the most important number that you've never heard of.
[29:51.450 --> 29:59.690]  Nearly all derivative products around the time of the global financial crisis were priced tied to this London Interbank Offered Rate.
[29:59.930 --> 30:11.050]  And so, by extension, you know, the price that, you know, you might pay on, like, short-term financing, your line of credit, whatever, indirectly, it's also tied to this LIBOR rate.
[30:11.690 --> 30:28.810]  The way that LIBOR was calculated was that the benchmark provider at the time was the British Bankers Association would call up, I want to say it was 16 banks, and just ask them, hey, what are you offering, you know, what would you offer in overnight lending right now?
[30:29.270 --> 30:41.270]  The banks would, you know, give their answers, the British Bankers Association would take off the highest number and the lowest number, the average of the rest, and publish that out as LIBOR, and then everyone's derivatives get repriced.
[30:44.310 --> 30:47.410]  But there was a group chat, as there always is.
[30:49.090 --> 30:53.750]  If you're not in the group, or, you know, if there isn't a group chat, then there is one, and you're just not in it.
[30:53.750 --> 30:56.490]  I believe they were called the Wessex Boys.
[30:57.250 --> 31:03.930]  They agreed, amongst themselves, what answers they should give to the British Bankers Association when they came calling.
[31:05.250 --> 31:09.830]  They would agree beforehand, and then when they get called up, they would just give that number.
[31:09.830 --> 31:13.370]  Remember, this is just indicative, it's indicative, it's a poll.
[31:13.610 --> 31:17.390]  They're not actually committed to lending up these numbers, so they would just make them up.
[31:18.710 --> 31:20.650]  They went on for several years.
[31:20.930 --> 31:23.930]  It was so bad that LIBOR doesn't exist anymore.
[31:24.850 --> 31:34.150]  And it got a lot of business analysts, a lot of jobs, working on, like, transitions to move away from LIBOR to other interest rates.
[31:35.130 --> 31:36.830]  There's also cornering the market.
[31:36.830 --> 31:52.690]  The canonical example, you know, might be the Hunt Brothers, who cornered the silver market, or Enron, who in some cases cornered the electricity market in California by encouraging producers to switch off their generators to make sure the price would go up.
[31:54.430 --> 31:58.230]  Or was it trading spaces where they cornered the orange juice market?
[32:01.550 --> 32:03.730]  Yeah, trading places, that's right.
[32:05.430 --> 32:07.210]  Yeah, concentrated orange juice.
[32:08.390 --> 32:20.930]  And then marking the close, because a lot of the prices of, like, you know, the portfolio statement that you get in the mail is tied to the price that's the last one at the end of the day.
[32:20.930 --> 32:25.630]  So, you know, if you were to manipulate the market, the incentive is to do it close to the end of the day.
[32:25.730 --> 32:30.290]  So marking the close is sometimes, like, a special case of manipulative trading.
[32:32.910 --> 32:34.870]  That's largely my prepared talk.
[32:37.250 --> 32:40.670]  I try to keep things, like, generally, you know, non-incriminating.
[32:40.830 --> 32:48.410]  But since there's time for questions that I haven't had a chance to prepare for, I'll answer things a lot more freely than I might otherwise.
[32:51.680 --> 32:52.220]  Yeah?
[32:53.020 --> 32:53.700]  All right.
[32:53.840 --> 33:05.400]  So if the SEC is only prosecuting stuff that seems like the statute of limitations has already passed on, how does that work exactly, first of all?
[33:05.480 --> 33:09.800]  And second of all, what's the incentive not to just do all the illegal things?
[33:11.080 --> 33:12.540]  I'm not a securities lawyer.
[33:13.420 --> 33:16.960]  But there's always going to be a mix of, like, criminal actions and civil actions.
[33:17.240 --> 33:18.680]  Can you repeat the question with the people?
[33:18.760 --> 33:19.380]  Oh, yeah.
[33:19.680 --> 33:35.460]  So if, you know, these cases take a long time to bring before a judge, and there's, you know, there may be a statute of limitations, you know, what's the incentive to not just doing a bunch of illegal shit and then hoping that you can run down the clock?
[33:36.520 --> 33:47.580]  You know, the answer is that I'm not a lawyer, and I hope to never be in a situation where I'm, you know, exploring the prosecutorial matters in great detail.
[33:49.820 --> 33:51.760]  Because probably then I've done something really wrong.
[33:54.000 --> 33:59.760]  But, you know, maybe the glib answer is that there's always going to be a mix of, like, criminal and civil actions.
[34:00.360 --> 34:03.080]  And if they can't get you in one, then they can get you in the other.
[34:05.860 --> 34:09.880]  But, you know, there are a lot of instruments they have under, like, the U.S. code.
[34:10.700 --> 34:13.840]  And then even if, you know, you don't, they don't get you there.
[34:14.040 --> 34:19.500]  Like, they'll seek to recoup your ill-gotten gains, as they call it.
[34:19.560 --> 34:22.360]  And the bill on that can generally be quite large.
[34:23.760 --> 34:26.520]  So, you know, you probably have to, like, sell the Maserati.
[34:29.340 --> 34:29.900]  Yeah?
[34:30.260 --> 34:33.560]  I've not seen trading places, but what is cornering the market?
[34:33.840 --> 34:34.920]  Like, how do you buy orange juice?
[34:35.060 --> 34:35.600]  Oh, yeah.
[34:35.800 --> 34:45.040]  So, in trading places, they're trading commodity futures.
[34:47.160 --> 34:55.360]  So, contracts in which we commit to, you know, I commit to, say, buying from you orange juice in the future.
[34:55.360 --> 34:56.680]  You're an orange juice producer.
[34:56.680 --> 35:01.120]  You've got your grove of oranges down in, you know, Florida or Georgia.
[35:03.340 --> 35:08.360]  And you want to be certain that you're going to be able to sell those oranges when your harvest comes.
[35:09.880 --> 35:12.200]  So, you and I enter into a contract.
[35:13.500 --> 35:18.700]  I commit to you that I will buy your oranges in the future at the harvest time.
[35:18.700 --> 35:24.440]  And we agree on a price today that we're going to do the deal out in the future.
[35:26.600 --> 35:30.120]  We now have, like, a contract, and that contract's actually tradable.
[35:30.240 --> 35:44.140]  And so, there are exchanges, primarily like the Chicago Mercantile Exchange and the Intercontinental Exchange, where I can go, or you can go, you can go and sell that contract to provide orange juice.
[35:44.840 --> 35:49.660]  And if you sell that contract, now you need to, you know, go find a new buyer for your orange juice.
[35:49.960 --> 35:58.820]  But you've also, you know, if the future price of oranges has gone up, you've now recognized capital gain.
[36:00.100 --> 36:05.140]  But I can also go out, and I can also sell my obligation to buy orange juice in the future.
[36:07.100 --> 36:24.480]  And so, if I were to go out, and I were to buy all of the outstanding contracts to buy orange juice, and I am now the only person who will be able to take delivery of orange juice in the future.
[36:27.940 --> 36:41.000]  If, you know, like, the grocery store, whatever, wants orange juice to be able to sell, they now only have one seller who's able to provide, you know, that guarantee of delivery of orange juice.
[36:41.000 --> 36:49.380]  And so, by buying up all the supply, I'm able to corner the market and prevent you from being able to get your sweet breakfast treat.
[36:53.440 --> 36:53.800]  Yeah.
[36:53.800 --> 36:56.500]  Massive amount, and generally also a huge amount of leverage.
[36:56.920 --> 37:02.380]  So, the Hump brothers, for example, I want to say was the 70s or the 80s, they decided to try and corner the silver market.
[37:03.200 --> 37:05.880]  Two brothers out of, I don't know, probably Chicago.
[37:07.900 --> 37:09.700]  Maybe they were in New York, but I won't say that here.
[37:12.740 --> 37:14.380]  But buy up all the silver futures.
[37:14.540 --> 37:17.260]  And they were able to do that by doing it largely on borrowed money.
[37:18.860 --> 37:27.960]  And the futures exchange then imposed limits on the amount of money that you can borrow and imposed limits on the positions that you can hold, and so they were forced to sell them.
[37:28.720 --> 37:41.460]  And so, you know, the exchange, which at the time was like kind of a cooperative run by all the members, the investors agreed to change the rules so that their cornering of the market wouldn't be successful.
[37:41.680 --> 37:47.980]  In a sense, you can also kind of think about it in the same way as, you know, buying GameStop shares to try and squeeze the shorts.
[37:49.800 --> 38:02.780]  You know, in principle, if, you know, you're a short seller, you're borrowing the shares from someone else in order to sell them.
[38:02.900 --> 38:08.240]  In the hope that the price will go down so that you can buy back the shares later and return the stock to the person you borrowed it from.
[38:12.040 --> 38:19.160]  But if someone, you know, the person you borrowed that stock from comes along and demands it back, you have to give it to them right then and there.
[38:20.360 --> 38:21.160]  You've got a day.
[38:21.260 --> 38:21.920]  You've got two days.
[38:21.920 --> 38:25.680]  And you can probably arrange to borrow it from somewhere else and get it to work out.
[38:25.680 --> 38:28.900]  But for the sake of argument, you've got to deliver it right then and there.
[38:29.360 --> 38:40.540]  So the thinking is, you know, if you've got a thin market and you've got a lot of, you know, dispersed buyers who all agree to buy up GameStop stock.
[38:44.660 --> 38:54.480]  One, as the price goes up, the short seller will have to put up more capital to be able to maintain that position with the broker because the broker starts to get a little bit scared that they won't be able to buy back the position.
[38:57.200 --> 39:03.300]  And if they can't put up that capital, then they'll sell the position then and there and stop, you know, shorting the stock.
[39:06.500 --> 39:19.440]  The other result, too, is that if you're, you know, buying the stock from actual natural investors who might be loaning out their shares, you buy that stock, the person who they're loaning that stock to now has to deliver it so that they can make settlement on the sale.
[39:20.540 --> 39:28.340]  So, you know, the thinking going that that's sort of one way to kind of a short squeeze being a type of cornering the market.
[39:29.480 --> 39:37.640]  You know, I wouldn't say the jury's out on whether or not that constitutes market manipulation and largely because it hasn't been brought to a jury.
[39:39.620 --> 39:46.820]  Like, it's the point where Congress, in their investigation, wasn't able to decide whether or not that's right or wrong.
[39:48.500 --> 40:01.240]  Like, whether or not a bunch of dispersed people who have, like, no, you know, common intent to do something can, like, just by virtue, can the emergency system have, like, some moral good or moral bad?
[40:01.420 --> 40:01.840]  I don't know.
[40:01.960 --> 40:02.800]  Congress doesn't know.
[40:03.920 --> 40:04.440]  Yeah.
[40:05.200 --> 40:10.560]  There's a certain fraction of bad actors out there who are doing stuff they shouldn't be doing against the law.
[40:10.780 --> 40:15.540]  And you had access to the public data transactions, and you were able to analyze that data.
[40:15.540 --> 40:24.380]  Could you find ways... would it be productive to find ways of detecting the bad actions so you can tip off the SEC and bug bounty kind of thing?
[40:24.600 --> 40:24.920]  But...
[40:24.920 --> 40:29.460]  And then, you know, just a bunch of these things, and then pick up the rewards?
[40:31.040 --> 40:31.480]  The...
[40:31.480 --> 40:32.160]  Oh, that's interesting.
[40:32.360 --> 40:33.300]  It's a clever idea.
[40:34.200 --> 40:39.780]  I think, like, realistically, if you were that good at doing it, you would just probably pick up a salaried job.
[40:43.120 --> 40:45.320]  Unfortunately, the SEC, I don't think, pays all that well.
[40:48.200 --> 40:51.860]  The Whistleblower Awards are generally designed so that...
[40:51.860 --> 40:59.640]  The understanding is that if you blow the whistle on something, it could be a bit of a career-limiting move.
[40:59.820 --> 41:10.980]  And so they need to, you know, ensure you against the understanding that you might not be able to find another job, and they need to, you know, make it economically worthwhile for you to be willing to forego future earnings.
[41:11.120 --> 41:13.460]  So there is, like, a little bit of that skin in the game.
[41:15.440 --> 41:18.460]  There are some weird cases where, like, that doesn't necessarily apply.
[41:20.880 --> 41:23.120]  But, yeah, largely, you've got to have a skin in the game.
[41:23.220 --> 41:26.580]  There is no, as far as I know, no market manipulation bug bounty program.
[41:27.340 --> 41:28.060]  Yeah.
[41:29.520 --> 41:33.820]  In your presentation, you're talking mostly about, you know, you have here.
[41:34.380 --> 41:34.980]  Yeah.
[41:34.980 --> 41:38.720]  But really, what you're mostly going to be training against is a program.
[41:39.280 --> 41:39.740]  Yes.
[41:39.740 --> 41:46.060]  And so, even though, you know, you're watching this, a human probably was going to spot manipulation.
[41:47.080 --> 41:51.240]  But the computer ain't because it's just some JavaScript running somewhere.
[41:51.620 --> 41:57.480]  I think, like, in a lot of cases, and I'm generalizing here, and like, I can't speak to anyone's specific processes.
[41:58.500 --> 42:09.630]  But, you know, if Ian is someone who's reasonably sophisticated, who, like, in the long run, you're probably going to run up against a relatively sophisticated counterparty.
[42:11.090 --> 42:16.670]  You know, let's look at an extreme example where, like, you have inside information about a merger that's about to happen.
[42:16.810 --> 42:20.430]  And you go out and you buy a bunch of out-of-the-money call options on that merger.
[42:21.610 --> 42:26.070]  Sure, you might buy some of those out-of-the-money call options from, you know, another individual investor.
[42:26.110 --> 42:30.270]  But in all likelihood, on the other side of that trade, at some point, is going to be someone pretty sophisticated.
[42:31.590 --> 42:40.130]  And when you make a bunch of money on that trade, equivalently, that sophisticated counterparty is going to lose a bunch of money in that trade.
[42:40.170 --> 42:45.490]  And they're really diligent about looking at where they lost money and where they made a lot of money.
[42:45.550 --> 42:51.770]  Because sometimes where you make a lot of money on something, right, maybe you were doing something wrong and you were taking a risk that you didn't know about.
[42:52.770 --> 42:56.610]  But, you know, where there's that loss, they're going to start digging into it.
[42:56.690 --> 43:00.170]  And then go back and look at the market data, you know, step through the tape.
[43:00.650 --> 43:06.190]  And then maybe see that you had that flickering, you know, sell offer deep in the book.
[43:07.410 --> 43:10.110]  And then, you know, maybe tip off the regulator on that.
[43:12.110 --> 43:15.950]  So, eventually, a run-up with, like, against a human looking at it.
[43:16.970 --> 43:28.410]  It, like, remains to be seen what the bullshit generator large language model, like, types of products will do for this.
[43:28.570 --> 43:35.470]  But, like, in all likelihood, like, some linear regressions will get you pretty far.
[43:37.410 --> 43:37.850]  Yeah?
[43:38.130 --> 43:40.610]  You just mentioned out-of-the-money call options.
[43:40.690 --> 43:41.590]  Can you define that, please?
[43:41.870 --> 43:42.230]  Oh, yeah.
[43:42.230 --> 43:59.090]  So, call options are a contract in which the holder has the right but not the obligation to purchase or sell a security in the future.
[44:06.290 --> 44:15.650]  Say I buy a option to buy stock in Tesla six months from now at a certain strike price.
[44:16.190 --> 44:22.110]  I pay a small premium to you today for the privilege of entering into that contract with you.
[44:22.110 --> 44:25.990]  If the price of Tesla goes up above, say, my strike...
[44:25.990 --> 44:28.890]  I don't know what Tesla's at now, but for the sake of argument, say it's at 420.
[44:31.490 --> 44:44.490]  If I enter into that contract with you to buy Tesla shares at 430, and the price of Tesla drops, you know, goes to $10, all I lose is that premium that I've given up to you.
[44:45.410 --> 44:46.710]  This is a simplification, right?
[44:46.710 --> 44:47.990]  This is not investment advice.
[44:50.070 --> 45:00.010]  If the price goes above 430, say to 440, I settle the contract.
[45:00.350 --> 45:04.030]  I assign you on the contract as the technical terminology.
[45:04.050 --> 45:08.510]  I buy the stock from you at 430, and I immediately sell it at 440.
[45:08.510 --> 45:13.030]  So I get that $10 difference less whatever premium I pay to you.
[45:13.650 --> 45:16.910]  You can see, you know, say the premium is $1 in that case.
[45:17.090 --> 45:39.030]  So if I only put up $1 and I get that kind of leverage, it allows me to buy a very large number of these call options and develop a very large position where, you know, if it's a $1 premium, instead of buying one share of Tesla at 420, it could buy 400 shares of exposure at $1.
[45:40.030 --> 45:58.330]  And so, you know, when you look at really egregious insider trading cases, like the most obvious and stupidest ones that tend to get, like, the most laughed about headlines are the ones where people are using these types of very leveraged products.
[45:59.850 --> 46:01.350]  What does out of the money mean?
[46:01.350 --> 46:08.310]  Oh, out of the money being kind of a term of art.
[46:10.350 --> 46:23.370]  If Tesla's at 420, an at-the-money call option would be considered one to buy at 420, an in-the-money call option would be one to buy at 400, an out-of-the-money one would be to buy one at 430.
[46:23.550 --> 46:26.230]  Deep out of the money would be, like, $500 or whatever.
[46:26.790 --> 46:34.190]  So an idea of, you know, technically it's called the delta, how far you are from the strike.
[46:34.470 --> 46:34.910]  Thank you.
[46:35.130 --> 46:35.370]  Yeah.
[46:37.710 --> 46:44.370]  So how do you feel that the relationship between the financial industry and the regulators has evolved since 08?
[46:44.510 --> 46:51.010]  Because I just remember, like, Wolf of Wall Street portrayal, which is like, you know, there's a perception that they're probably too close.
[46:51.990 --> 46:57.510]  And then in addition, are there any, like, resources you recommend for people who are beginners to this stuff but, like, interested?
[46:58.650 --> 47:06.750]  So, like, it's unfortunate that there really aren't any good resources on, like, I call this realm market microstructure.
[47:07.790 --> 47:14.470]  There are some great academic papers, but there's not really a good bridge from academia to industry.
[47:15.710 --> 47:19.930]  There's a good seminar series that runs Tuesdays at, like, around 11 a.m.
[47:20.150 --> 47:23.190]  Eastern called the Microstructure Exchange, where academics present their papers.
[47:23.830 --> 47:25.170]  It's really quite good.
[47:26.130 --> 47:34.050]  There are a few canonical textbooks, empirical market microstructure, market liquidity, and empirical market microstructure.
[47:34.670 --> 47:40.430]  But, like, there really is nothing that bridges theory into practice, unfortunately.
[47:43.610 --> 47:51.130]  Yeah, like, I was a software engineer, really, before, like, I sort of did an apprenticeship, if you will, into this space.
[47:51.970 --> 47:56.530]  Which is also partly, like, why I'm here at what's, like, otherwise a hacker convention.
[47:58.410 --> 48:01.670]  And then I kind of skirted around your other question there.
[48:01.670 --> 48:06.730]  But I'll say that, like, in The Wolf of Wall Street, my personal favorite character was the SEC agent.
[48:09.790 --> 48:16.890]  And, like, in the end, he did kind of do his part of the job, you know?
[48:19.750 --> 48:26.430]  You know, maybe there's leniency, but ultimately the SEC did get him.
[48:28.270 --> 48:29.410]  Was it quick enough?
[48:29.590 --> 48:30.130]  Probably not.
[48:30.350 --> 48:35.330]  But, like, you know, eventually they got him.
[48:38.310 --> 48:38.830]  Yeah.
[48:39.430 --> 48:40.230]  Either of you.
[48:49.670 --> 48:50.190]  Yeah.
[48:51.610 --> 48:57.870]  Are there, like, canonical examples of market manipulations that regularly get litigated against markets that don't involve an actor?
[48:58.130 --> 48:58.590]  So why?
[48:58.750 --> 49:02.950]  Because they're kind of, like, showing the behavior of the fact that it impacts the future.
[49:06.270 --> 49:12.030]  As you get more, like, complicated, I can't think of good examples.
[49:12.770 --> 49:27.650]  Like, the classic trade-driven one would be, like, a cross-asset price manipulation, where you have, say, like, a really illiquid stock with, like, a very liquid derivative instrument tied to it.
[49:27.730 --> 49:31.410]  But the price of that derivative instrument is based on the price of the stock.
[49:31.690 --> 49:45.890]  And so you establish, you know, a really big position in that other instrument, which you also kind of see in the crypto world, too, where, like, you've got a token whose value is based on a value generated by an oracle.
[49:45.890 --> 49:51.990]  And so instead of, like, trying to manipulate the price of the token, you instead manipulate the input into the oracle.
[49:52.310 --> 50:04.490]  And if that input into the oracle is another security, then, you know, buying that up might not take very much capital, but has an outsized impact on, you know, the token that's tied to the oracle.
[50:06.330 --> 50:13.150]  Which, you know, you may see similar things happen in traditional finance.
[50:14.110 --> 50:22.170]  But, like, again, it's probably a lot easier to catch than if you're doing it in, like, crypto world where no one's really watching.
[50:24.790 --> 50:27.990]  I would call it cross-asset manipulation.
[50:28.990 --> 50:29.870]  Yeah?
[50:30.810 --> 50:47.350]  In the protocol you described at the beginning for detailing the buy and sell orders, is there any necessity or hurry in increasing resolution on the timestamp from millisecond to microsecond?
[50:47.350 --> 50:47.890]  Yeah.
[50:48.150 --> 50:53.070]  Generally, the world runs at nanos.
[50:55.290 --> 50:55.890]  Yeah.
[50:57.710 --> 51:00.170]  Generally, we're looking at nanos these days.
[51:01.470 --> 51:02.310]  Which is quick.
[51:02.470 --> 51:02.650]  Yes?
[51:03.150 --> 51:05.630]  Is it a little bit about pump and dumps?
[51:05.850 --> 51:06.230]  Yeah.
[51:06.490 --> 51:11.710]  Like, social media has made that very, very interesting.
[51:12.890 --> 51:13.330]  Yeah.
[51:13.710 --> 51:15.650]  Like, the pump and dumps...
[51:15.650 --> 51:19.650]  The question is, you know, has social media affected pump and dump schemes?
[51:23.110 --> 51:25.230]  They've sort of always existed in some form.
[51:25.390 --> 51:28.650]  Like, in the olden days, there would be, like, a newsletter that goes out.
[51:36.130 --> 51:46.350]  It might just be that, like, it's become a lot stupider with, like, really dumb Discord channels that makes it a lot more, like, exciting to read about.
[51:47.770 --> 51:49.370]  I think it's always happened.
[51:49.750 --> 51:54.810]  And, like, it's tough to enforce ex ante, right?
[51:54.810 --> 52:05.070]  Like, it's tough to find these channels and, like, shut them down and do enforcement unless you, like, first see the price movement and then you try and, like, tie it to something.
[52:05.370 --> 52:11.970]  Generally, the regulator, they can see, like, what accounts are actually trading the stock and then, you know, they'll go to the broker and they'll try and tie stuff together.
[52:12.130 --> 52:12.950]  But, yeah.
[52:13.330 --> 52:15.170]  Anyway, I think that's probably about all the time.
[52:15.170 --> 52:15.850]  I'll be around.
[52:16.210 --> 52:20.150]  You can come find me in the Lock Bypass Village and we can talk more.
[52:21.390 --> 52:28.170]  But, yeah, thank you all and thank you for indulging me in this, you know, cybersecurity adjacent thing that I find interesting.