Latest / Deep Values: Value Investing Deep Dives / MIT OpenCourse: The Mathematical Plumbing of Financial Markets
Transcript
- 0:00So picture this. It's 2013, and some of the
- 0:04absolute brightest mathematical minds at MIT are literally
- 0:08sitting on the floor of a crowded classer.
- 0:10Right, just packed in there. Yeah, totally packed.
- 0:12And they're trying to solve this billion dollar problem.
- 0:15I mean, these are students who can build flawless mathematical
- 0:18universes in their heads, right? They can probably map out
- 0:21complex differential equations in their sleep.
- 0:23Oh, absolutely. But they were facing this
- 0:26massive hurdle. They had absolutely no idea how
- 0:29the real world stock market actually functioned.
- 0:31Yeah, it's a completely different language.
- 0:32Exactly like they understood the sterile logic of numbers, but
- 0:36the chaotic, messy, you know, emotionally driven plumbing of
- 0:40global wealth that was just a complete mystery to them.
- 0:43A. Total black box.
- 0:44Right. So today we are tearing down the
- 0:47drywall to look at that exact plumbing.
- 0:49We're doing a deep dive into the actual mechanics, the
- 0:51institutional players and the complex mathematics that govern
- 0:56modern financial markets. It's going to.
- 0:58Be a fun one. It really is.
- 0:59Yeah, but before we get to the blueprints, I have a quick
- 1:01message I need to read for you to access more advanced
- 1:04researches, follow Guru's latest movement more and leverage our
- 1:08latest Deep Values agents. Welcome to download Deep Values
- 1:11APP from App Store to track stock market temperature.
- 1:15Welcome to download iOS app Buffett Indicator from Apple App
- 1:19Store or Bubble Watch from Google Play.
- 1:22Links in the description. Disclaimer All analysis is based
- 1:26on publicly available sources and may be subject to revisions
- 1:29or differing interpretations. All contents from this channel
- 1:32is for research and entertainment purposes only and
- 1:34does not constitute any investment or financial advice
- 1:36or service. Conduct your own due diligence
- 1:39before making investment financial decisions.
- 1:42OK, so getting back to it, our guide for this deep dive is an
- 1:45MIT Open Courseware lecture. Right, the introduction of
- 1:49financial markets, financial terms and concepts.
- 1:51Yeah, that's the one. And the architect of this
- 1:54course, Doctor Jay Shia, he's like the perfect lens for this
- 1:57because, well, he didn't actually start in.
- 1:58Finance. Oh, he didn't.
- 1:59He got his PhD from MI TS Course 6, which is electrical
- 2:02engineering and computer science.
- 2:04So he was an engineering researcher before making this
- 2:06huge leap to Wall Street as a quantitative analyst.
- 2:10Or, you know, a quant. And you really have to
- 2:13understand that specific breed of engineer to understand modern
- 2:17markets. Why is that?
- 2:18Well, because that transition going from pure engineering to
- 2:22high finance, that's the bedrock of modern algorithmic trading.
- 2:27I mean, when Shia joined Salomon Brothers, he was walking right
- 2:31into the epicenter of quantitative arbitrage.
- 2:33Oh wow, Salomon Brothers. Yeah, for context, they were The
- 2:36Pioneers of pulling pure mathematicians into the trading
- 2:40pits. They needed them to build these
- 2:42incredibly intricate arbitrage model.
- 2:45So they weren't looking at like traditional stock fundamentals.
- 2:48Not at all. They didn't care about the
- 2:50fundamental health of a company. They were hunting for hidden
- 2:53relative value. What does that mean exactly?
- 2:55It's like scanning the market for tiny, tiny mathematical
- 2:58inefficiencies. Say there's a minute pricing
- 3:01discrepancy between a 10 year treasury and a corporate bond.
- 3:04The quants would systematically buy one and short the other to
- 3:08lock in a risk free profit, and they do it before the rest of
- 3:11the market even noticed the gap. That is wild.
- 3:15And he clearly had a knack for it, right?
- 3:16Because he spent like over 17 years at Morgan Stanley running
- 3:20global trading desks. Yeah.
- 3:22And then eventually became the chief risk officer for the
- 3:25Harvard Management Company. Right, overseeing the
- 3:27university's massive endowment. But the origin story of this MIT
- 3:33class is what really gets me. So she and his colleague, Silly
- 3:36from the math department, they were running these standard 30
- 3:40minute recruiting interviews at MIT for Morgan Stanley.
- 3:43Just standard campus recruiting. Yeah, and they're sitting across
- 3:46in these absolute geniuses, but they hit this massive wall.
- 3:50The wall of practical application.
- 3:51Yeah, because the students would ask these incredibly high level
- 3:54theoretical questions. But you know, a 30 minute
- 3:57interview just isn't enough time to bridge the gap between
- 4:00academic mathematics and market reality.
- 4:02Wall Street desperately needed their raw computational power to
- 4:06build out systems, but you can't just hand over a billion dollar
- 4:09ortfolio to someone who doesn't understand market liquidity.
- 4:12Or credit risk. Yeah, it feels a lot like trying
- 4:15to hire a brilliant automotive engineer who can, I don't know,
- 4:20diagram the thermodynamics of a combustion engine from memory.
- 4:24Right? But they've literally never sat
- 4:26behind the steering wheel of a moving car.
- 4:28That is a perfect analogy. Like you wouldn't just hand them
- 4:31the keys to a Formula One car and say go win the race.
- 4:34Exactly. And that's exactly why Shia and
- 4:36Vasily started with just one single seminar.
- 4:39They had to actively build a bridge to pull that pure logic
- 4:43into finance. And it blew up.
- 4:45It became so popular it exploded into a full term class.
- 4:49They partnered with the math department.
- 4:50They did field trips to trading floors in New York.
- 4:53Oh, that's so cool. Yeah, they were basically taking
- 4:55these raw engineers and teaching them the mechanics of the
- 4:57machine. OK, so let's unpack those
- 4:59mechanics because when we look at what these Quan's are
- 5:02actually trading, the lecture maps out this super rapid
- 5:06evolution like it moves from the historical physical exchanges
- 5:11straight into modern electronic communication networks.
- 5:14Right, Ecns. Yeah, ECNS.
- 5:16And what's moving through those networks is this escalating
- 5:18ladder of complexity. You start with currencies, you
- 5:21know, going from metal to Fiat to Bitcoin, and then you hit
- 5:25equities. But they aren't dwelling on the
- 5:27basics of company ownership. They're looking at how you model
- 5:31explosive growth trajectories in the secondary market.
- 5:34Like the current AI boom. Exactly, using specific examples
- 5:37like NVIDIA stocks soaring off the back of AI and GPU demand
- 5:41and how you price index baskets like the S&P 500 are complex
- 5:46ETF. 'S But what's fascinating here
- 5:48is while equities get all the flashy headlines, the lecture
- 5:51makes it clear that the real heavy lifting happens in the
- 5:54fixed income market. The bond market.
- 5:55Yes, the bond market dictates the flow of global liquidity far
- 5:59more than the stock market. Oh yeah.
- 6:01When quants are modeling bonds, they're essentially modeling
- 6:04credit risk and time preference. A government or a corporation
- 6:08borrows capital and the mathematical model has to
- 6:11accurately price the probability of them defaulting over the life
- 6:15of that. Loan.
- 6:15OK, which brings up the 2008 financial crisis.
- 6:18Yeah, because the source actually points out that 2008
- 6:21wasn't an equity crisis, Not at all.
- 6:23It was fundamentally a credit crisis born out of the mortgage
- 6:27market. Exactly.
- 6:28It was a complete failure of the risk models.
- 6:31The models basically assumed housing prices wouldn't
- 6:34correlate downward across the entire country simultaneously.
- 6:37Big oops. Yeah, big oops.
- 6:38When the underlying credit risk actually materialized, all those
- 6:42securitized products built on top of the mortgages just
- 6:45collapsed. Wow.
- 6:46This is why watching the interest rate curve is so
- 6:48critical in these models. Like the lecture highlights the
- 6:51anomaly of an inverted yield curve.
- 6:53Right, where short term borrowing rates are actually
- 6:55higher than long term rates. Right.
- 6:58And from a mathematical perspective, that is a severe
- 7:01distortion. It implies that investors are
- 7:03pricing in higher immediate risk or they're anticipating the
- 7:07central bank stepping in to stimulate a slowing economy.
- 7:10I see. So when the Federal Reserve
- 7:12adjusts that short term rate, it forces these quants to
- 7:16recalibrate every single pricing model they run.
- 7:19OK. So we've got equities and we've
- 7:20got the fixed income credit markets.
- 7:22Then you layer on commodities like gold, copper, lithium for
- 7:26EV batteries, right? And the lecture even brings up
- 7:29that surreal moment when crude oil futures literally dipped
- 7:34into negative territory. Going below $0.00?
- 7:37Crazy. Wild time.
- 7:39But built on top of all of these underlying assets are the
- 7:42derivatives, you know, options, forward swaps.
- 7:45And I have to push back here a little bit because for an
- 7:48outside observer, derivatives often just look like massive
- 7:50side bets. Like if you don't actually own
- 7:53the underlying stock or the physical barrel of oil, aren't
- 7:56you essentially just operating a sophisticated casino, just
- 7:59gambling on price action? That is a super common
- 8:02perception, and to be fair they are heavily used for
- 8:04speculation. Right, but if we connect this to
- 8:07the bigger picture and look at the underlying mechanism, a
- 8:11derivative is fundamentally an instrument of risk transfer.
- 8:15Risk transfer. Yeah, think of it like buying an
- 8:18insurance policy on a burning building.
- 8:20OK, Yes, speculators might buy that policy just to cash in on
- 8:24the fire, which totally looks like gambling.
- 8:27But the primary structural purpose of the market is for the
- 8:30actual owner of the building to transfer the risk of it burning
- 8:34down to someone else. I see.
- 8:36If you are a global airline, your entire business model
- 8:40collapses if jet fuel prices double overnight.
- 8:44True. So you use derivatives like
- 8:47forward contracts or swaps to lock in your fuel costs.
- 8:50You are basically paying a premium to dump your volatility
- 8:53risk onto the market. Here's where it gets really
- 8:55interesting though, because if an airline is dumping their
- 8:58risk, someone else has to be on the other side of that trade,
- 9:01catching it. Products don't trade in a
- 9:02vacuum. And that brings us to the
- 9:05players in the ecosystem. The lecture actually provides
- 9:09some crucial historical context here regarding the institutions
- 9:12themselves, specifically the divide between commercial banks
- 9:17and investment banks. Right.
- 9:19And that distinction is vital for understanding how capital
- 9:21flows. Yeah, because commercial banks,
- 9:23they manage everyday deposits and loans, while investment
- 9:27banks handle the capital markets, equities, fixed income,
- 9:30mergers. But didn't those lines get
- 9:32blurred? They did.
- 9:34The lecture notes how JP Morgan was traditionally a commercial
- 9:37bank, but after the Glass Steagall Act was repealed in
- 9:401999, it reverted back to an investment bank model right,
- 9:43which essentially allowed these massive institutions to combine
- 9:47ordinary consumer deposits with really aggressive capital market
- 9:50activities. OK.
- 9:51So that's the banks, but then you also have this critical
- 9:54dynamic between LP's and GPS, right, Limited partners and
- 9:56general partners. Let's drill into that dynamic
- 9:59for a second because that feels like the bedrock of private
- 10:02equity and hedge funds. It's not just a title, it's a
- 10:04very specific flow of capital and risk.
- 10:07Exactly. The limited partners, the LP's,
- 10:11they are the pensions, the endowments, the sovereign wealth
- 10:15funds. Ones with the deep pockets.
- 10:16Exactly. They have massive pools of
- 10:19capital, but they lack the specialized mathematical models
- 10:23to generate excess returns. OK, so they need the GPS.
- 10:26Right. The general partners are the
- 10:29fund managers, the quants and the traders.
- 10:31So the LP provides the capital and the GP provides the
- 10:34strategy, taking a management fee and a percentage of the
- 10:37profits. So within this institutional
- 10:40framework, the course identifies 3 distinct types of traders
- 10:43operating on the floor, right? First, you have the hedgers.
- 10:46These are the corporate players, like our airline example, or,
- 10:49you know, an exporter trying to reduce currency risk so a sudden
- 10:52shift in the euro doesn't just wipe out their profit margin.
- 10:55Yes. Second, you have the market
- 10:56makers. These are the major bank trading
- 10:59desks, and I've always thought of market makers as like a toll
- 11:02booth operator on a busy highway.
- 11:05That's a really accurate way to look at it.
- 11:06Because they aren't taking a directional bet, right?
- 11:08Yeah, They're just constantly offering a buying price and the
- 11:11selling price, the bid in the ask.
- 11:13Yep, just capturing the tiny spread between those two
- 11:15numbers. Exactly.
- 11:17They don't care if the market crashes or moons, they don't
- 11:19care who wins the race. They just want maximum volume
- 11:23driving through their toll booth.
- 11:25They're actively trying to remain neutral and avoid risk.
- 11:28Which leaves the third group, the risk seekers, right?
- 11:33These are the proprietary trading desks and the aggressive
- 11:36hedge funds. They are actively using capital
- 11:39to initiate risk. They take a directional stance,
- 11:41basically betting that their mathematical models have found
- 11:44an inefficiency that the rest of the market has missed.
- 11:47OK, so this raises A fundamental question about the whole
- 11:49ecosystem. What's that?
- 11:50Beneath all the complexity of the derivatives and the
- 11:53algorithms, what is the core tension like?
- 11:56The lecture boils it down to this constant tug of war between
- 12:01lenders who are demanding a return on their capital and
- 12:03borrowers or managers who need that capital to execute their
- 12:07strategies. Right.
- 12:08Every single participant is running this complex calculus to
- 12:12maximize their gain while minimizing their exposure, all
- 12:16in an environment that is strictly defined by uncertainty.
- 12:20And if the defining characteristic of the
- 12:21environment is uncertainty, how do you actually?
- 12:24Win. You weaponize mathematics.
- 12:26This is where we finally see what those MIT engineers are
- 12:29actually building. The lecture outlines 3 primary
- 12:33applications of math in high finance.
- 12:35OK, let's go through. Them The first is pricing
- 12:36models. Because pricing a complex
- 12:39derivative isn't intuitive, right?
- 12:41It requires heavy computational lifting.
- 12:45The source points to the Black schools Merton model from the
- 12:471990s and John Hull's work on differential equations.
- 12:51And the black schools model is just a perfect example of
- 12:54mathematical revolution in finance.
- 12:56Really. How so?
- 12:57Well, before it, pricing an option was largely just
- 12:59guesswork based on historical volatility.
- 13:02But Black Skulls and Merton figured out a differential
- 13:05equation that allowed traders to dynamically hedge an option
- 13:09using the underlying stock. Yeah, it effectively created a
- 13:12risk free portfolio. It removes move the guesswork
- 13:15entirely and replace it with a mathematical certainty of
- 13:18pricing. I mean, it was so revolutionary,
- 13:20it won a Nobel Prize in economics.
- 13:22That's insane. But pricing is only the first
- 13:25pillar. The second use they talk about
- 13:27is risk management, which arguably dictates survival even
- 13:31more than pricing. Absolutely.
- 13:32It's all about position sizing. Like if your model finds an
- 13:36edge, how much of your capital do you actually commit?
- 13:38But it's also about overriding human biology.
- 13:41Oh yeah, the psychological part. Right, because if I buy a stock
- 13:45and it drops 15% my instinct is pure panic.
- 13:48Or if it goes up 50%, my greed tells me to hold on forever
- 13:51because I'm obviously a genius. And that biological reaction is
- 13:54exactly what destroys portfolios.
- 13:57A quantitative risk management process strips the emotion out
- 14:00of the execution. It dictates mathematically
- 14:02exactly when to take profit and exactly when to cut losses.
- 14:06It forces you to respect the volatility of the asset rather
- 14:09than your own adrenaline. So true.
- 14:12And the third use of math is trading strategies themselves,
- 14:15defining your actual edge. Like, are you a systematic
- 14:18trader where the algorithm executes 100% of the trades
- 14:21without any human intervention, Or are you discretionary, using
- 14:25the models to inform your human judgement?
- 14:28Are you running a value strategy?
- 14:30Hunting for fundamentally underpriced assets or growth
- 14:33strategy? Projecting future cash flows for
- 14:35tech giants? Or do your models follow the
- 14:37momentum of the trend or bet on mean reversion, assuming an
- 14:41overextended asset will inevitably snap back to its
- 14:44historical average? Exactly.
- 14:46But all those strategies lead to this massive philosophical
- 14:50collision in finance. The lecture highlights the
- 14:53debate between efficient market theory and behavioral finance.
- 14:56The classic debate. Yeah, so efficient market theory
- 14:59argues that all available information is instantly priced
- 15:02into an asset. The classic analogy is the $100
- 15:05bill on the sidewalk. Right.
- 15:06If you think you see $100 bill, don't bother bending over
- 15:09because if it were real, someone or an algorithm would have
- 15:11already picked it up. Exactly.
- 15:13In a perfectly efficient market, generating consistent excess
- 15:16returns, what they call alpha is theoretically impossible.
- 15:20But the quants wouldn't have jobs if the market were
- 15:23perfectly efficient. Precisely.
- 15:25And that brings in behavioral finance, which argues that the
- 15:28market is made of humans, and humans are deeply, predictably
- 15:31irrational. We panic sell at the bottom and
- 15:34we revenge trade at the top, and that systemic irrationality
- 15:38creates temporary mispricings. It leaves real $100 bills on the
- 15:43sidewalk for the algorithms to sweep up.
- 15:45And the lecture notes that when a quant firm finds one of these
- 15:48statistical edges, they protect it aggressively.
- 15:51Oh yeah, the engineers have to sign these lifetime non
- 15:54competes. They are completely locked in.
- 15:56You can't just take the firm's proprietary code to a rival
- 15:59hedge fund. Though I love that one student
- 16:01brilliantly pointed out during the lecture that the firms don't
- 16:03actually pay you your massive salary for the rest of your life
- 16:06after you leave despite the non compete.
- 16:08Yeah, it's totally A1 sided lock in.
- 16:10But I I have to stop and ask a foundational question here.
- 16:13If the math is so advanced, and if the Black Skulls models are
- 16:17Nobel worthy, and if these quants are literal MIT geniuses,
- 16:22why do massive hedge funds still blow up?
- 16:25That's a great question. Right.
- 16:26Why do they lose billions of dollars?
- 16:28Doesn't the whole idea of a mathematical money printing
- 16:30machine breakdown In reality? It seems like treating finance
- 16:34like physics is a fatal flaw. It is a fatal flaw, and the
- 16:38lecture makes this exact distinction crystal clear to the
- 16:41engineering students. Finance is not physics.
- 16:44How so? Well, in physics, if you have a
- 16:46deterministic equation, A+B will equal C every single time,
- 16:51regardless of how you feel about it.
- 16:52But in finance, you're operating in a stochastic environment.
- 16:56A+B gives you a probability of C There is no absolute certainty.
- 17:00Your model might give you a 54% chance of being right.
- 17:04You aren't predicting the future, you are exploiting A
- 17:07statistical edge over a distribution of thousands of
- 17:10trades. So when things go wrong.
- 17:12When hedge funds blow up, it's usually because they mistook a
- 17:15high probability for a physical certainty and they over leverage
- 17:18their position right as a rare tail risk event occurred.
- 17:21Wow. OK, so to prove that strategy
- 17:23requires absolute discipline over probability, the course
- 17:28gives these brilliant students a pretty brutal initiation.
- 17:31Assignment 0. Yes, assignment zero.
- 17:34They each get a hypothetical $10,000.
- 17:36They had to select exactly 1 stock or one ETF.
- 17:40And they aren't restricted to going long.
- 17:42They can buy an inverse ETF if they want to short the market.
- 17:46They execute the buy at the market close on September 13th
- 17:51and they are forced to hold it until an automatic sell on
- 17:53November 15th. So roughly 2 months.
- 17:56But the genius of the assignment is the psychological trapdoor
- 18:00built into the timeline. Oh, this is the best part.
- 18:02On October 11th, exactly halfway through, they're given a one
- 18:06time option. Yeah, they can liquidate their
- 18:08position, take whatever balance is left and roll it into a
- 18:11completely new asset for the second month.
- 18:14Or they can grit their teeth and do absolutely nothing.
- 18:17It's incredible because it forces the student to confront
- 18:20their own behavioral finance. I want to put you, the listener,
- 18:23in that hot seat for a second. Oh yeah, think about this.
- 18:25If you had to deploy 10 grand into a single asset today,
- 18:29knowing you couldn't touch it for 30 days, what's your play?
- 18:32And imagine the absolute psychological torture of that
- 18:36October 11th option. It would be agonizing.
- 18:38Let's say your model was wrong and you're down 25%.
- 18:41Do you admit defeat, lock in the loss, and try to make it back on
- 18:44a new trade? Or do you trust your initial
- 18:47math and hold strong, hoping for mean reversion?
- 18:50It strips away all the theory and just exposes human panic.
- 18:53And how they are graded is just as important as the
- 18:56psychological pressure, right? Because the lecture doesn't just
- 18:59rank them by who has the most cash on November 15th.
- 19:01Doctor Shia grades the quality of their strategy using a very
- 19:05specific formula that penalizes reckless volatility.
- 19:09Because if I just, you know, throw my money into a wildly
- 19:12volatile meme stock and happen to get lucky on the day we sell,
- 19:15that doesn't mean I'm a good, good quantitative engineer.
- 19:16Exactly, it just means you bought a winning lottery ticket.
- 19:19So how does the formula actually weed out the gamblers from the
- 19:22actual risk managers? By tracking the daily profit and
- 19:25loss over the two months, you sum up every single daily gain.
- 19:29Let's call it G. OK, G for gain.
- 19:31Right, then you sum up every single daily loss.
- 19:34Call that L The final quality metric is g -, L / g + L.
- 19:39Let's breakdown the mechanics of that denominator g + L.
- 19:42Why does adding the total gains and losses together act as a
- 19:46penalty? Because it measures the friction
- 19:48of your volatility, so G -, L is just your simple net profit.
- 19:51Right, makes sense. But let's say student A makes a
- 19:53slow, steady $20 a day with almost 0 down days.
- 19:57OK, their G + L denominator is very small, so their overall
- 20:01ratio stays incredibly high, close to 1.
- 20:03But what about a volatile student?
- 20:05Right, Student B is riding a roller coaster.
- 20:07They make $2000 one day, lose 1800 dollars the next day, make
- 20:10$1000, lose $900. They might end up with the exact
- 20:14same net profit as student A, but their G + L denominator is
- 20:18massive because of all those wild swings.
- 20:21So when you divide their net profit by that massive
- 20:23denominator, their score completely plummets.
- 20:26Wow. It mathematically proves to the
- 20:28students that a smooth, consistent strategy with
- 20:31ironclad risk management is objectively superior to a
- 20:35volatile strategy that just happens to survive.
- 20:38Exactly. It teaches them that minimizing
- 20:40the drawdown is just as critical as maximizing the upside.
- 20:44It forces an engineer's brain to respect the fact that in the
- 20:47real market, volatility acts as a drag on compounding capital.
- 20:51It is such a phenomenal way to close the gap between theory and
- 20:54reality. I mean, we started by looking at
- 20:56how raw computational power was recruited from the halls of MIT
- 21:00to build a bridge into the chaos of the trading floor.
- 21:03We've unpacked how quants model everything from inverted yield
- 21:06curves to derivatives that act as insurance policies, and how
- 21:10they navigate this whole ecosystem of hedgers, tollbooth
- 21:13market makers, and risk seekers. It's a complex ecosystem.
- 21:16It really is. And ultimately, we've seen that
- 21:18at the core of all this structural math success relies
- 21:21on acknowledging the irrationality of human behavior.
- 21:24It truly does. And you know the sizing.
- 21:25All of this leaves me with one final, broader implication to
- 21:28consider. Well.
- 21:30We spent a lot of time discussing how behavioral
- 21:32finance relies on human irrationality, the panic, the
- 21:36greed, the fear to create mispricings, right?
- 21:39It's the human element that leaves those $100 bills on the
- 21:42sidewalk for the quants to find. But we are rapidly moving into
- 21:46an era where algorithmic execution and artificial
- 21:49intelligence are taking over an overwhelming majority of Market
- 21:53Volume. Oh wow, I see where you're
- 21:55going. Yeah, as human decision making
- 21:57gets further and further abstracted away from the actual
- 22:00trades, what happens if AI eventually strips the emotion
- 22:04out of the market entirely, making it perfectly rational and
- 22:07perfectly efficient? Do the $100 bills disappear
- 22:10forever? That is a terrifying thought for
- 22:12a quant. Right.
- 22:13If the market achieves perfect mathematical efficiency, does it
- 22:16effectively kill the quantitative edge altogether?
- 22:19Man, now that is a structural paradox to chew on the next time
- 22:23you see those red and green arrows flashing across the
- 22:26screen. The skyscraper might look sleek
- 22:28and intentional from the outside, but the plumbing
- 22:30keeping it upright is constantly, constantly evolving.
- 22:33Thank you so much for joining us on this deep dive.
- 22:35Stay curious, keep exploring, and we'll catch you next time.