Latest / Deep Values: Value Investing Deep Dives / Why Spreadsheets Are Ruining Business Valuation
Transcript
- 0:00Welcome back to another deep dive.
- 0:02You know, usually when we think about evaluating a business, we
- 0:05picture this absolute precision. Right, like an exact science.
- 0:11Exactly. It's like looking at the
- 0:13blueprints of a house. You measure the walls, you
- 0:16calculate the load bearing beams, and you just know exactly
- 0:19what the structure is worth. It feels very solid.
- 0:21Well, it feels like engineering. The assumption has always been
- 0:25that these really complex, totally chaotic markets can be
- 0:29tamed by, you know, a neat little mathematical formula.
- 0:32You just plug in the numbers and the ultimate truth just pops out
- 0:35the other side. But then you step into the
- 0:37actual world of modern finance and suddenly you realize
- 0:40everyone is staring so closely at their laser measuring tools
- 0:43that they don't even notice they're building the entire
- 0:45house on quicksand. Yeah, that is a great way to put
- 0:48it. Today we are exploring A
- 0:50fascinating and honestly slightly terrifying perspective
- 0:53from a 40 year veteran of NYU Stern.
- 0:57He is widely known as the Dean Evaluation Professor Aswath de
- 1:00Medoran. Our mission today is to figure
- 1:03out why, despite having more data and AI tools than ever
- 1:06before in human history, our ability to actually value
- 1:09companies has gotten worse, not better.
- 1:11It's the ultimate paradox of the information age, really.
- 1:14I mean, we are drowning in data but somehow completely starving
- 1:16for common sense, yeah. Absolutely.
- 1:19Now before we dive into how to fix this broken system, I have a
- 1:21quick update tailored to your journey into market analysis.
- 1:25To access more advanced researches, follow Guru's latest
- 1:28movement and leverage our latest Deep Values agents.
- 1:31Welcome to download Deep Values APP from App Store to track
- 1:35stock market temperature. Welcome to download iOS app
- 1:37Buffett Indicator from Apple App Store or Bubble Watch from
- 1:41Google Play. Links are in the description.
- 1:43Disclaimer All analysis is based on publicly available sources
- 1:47and may be subject to revisions or differing interpretations.
- 1:50All contents from this channel is for research and
- 1:53entertainment purposes only and does not constitute any
- 1:55investment or financial advice or service.
- 1:58Conduct your own due diligence before making investment or
- 2:00financial decisions. Good to keep in mind.
- 2:02Right. OK, let's unpack this, because
- 2:04Professor de Madaran is arguing that the art of valuation has
- 2:07basically been hijacked by spreadsheets, and if we don't
- 2:09rediscover the basic logic of business, we are going to lose
- 2:13our grip on financial reality altogether.
- 2:15To really understand how we got here, we actually have to look
- 2:18the physical evolution of how valuations used to be done.
- 2:22The actual literal process of it.
- 2:24Exactly. We go back to the early 1980s.
- 2:27Valuing a business was a remarkably physical, friction
- 2:31heavy process. Yeah, the author talks about
- 2:33valuing his first company back in 1981, and it's almost comical
- 2:37by today's standards. He actually had to write a
- 2:39physical letter with a fountain pen to the investor relations
- 2:43office of a company just to request their annual report.
- 2:46A fountain pen, and then he'd wait weeks for it to arrive in
- 2:49the mail just. Sitting around waiting for the
- 2:51data, right? And once he finally had it, the
- 2:53actual modeling was done on a blank Ledger sheet with redrawn
- 2:56lines. He used a rimitive calculator
- 2:59that didn't even have present value buttons on it, and an
- 3:02eraser. Wow.
- 3:03He noted that he learned very quickly never to use a pen
- 3:06because just one single mistake meant redoing the entire column
- 3:10by hand. It sounds like torture,
- 3:12honestly, but there is a hidden advantage there, isn't there?
- 3:15Because the sheer physical effort required to put a number
- 3:18on a page, it forced a level of restraint.
- 3:21Yeah, that friction served a vital cognitive purpose.
- 3:25Because every single cell had to be hand computed, your model was
- 3:29forced to be what statisticians call parsimonious.
- 3:32Which means keeping it simple. Right.
- 3:35You used as few line items as possible.
- 3:37You just couldn't afford to be distracted by noise, so you had
- 3:40to think critically about the actual drivers of the business.
- 3:43You ask yourself, does this specific metric actually matter?
- 3:46And if the answer was no, it just didn't go on the Ledger
- 3:48exactly. So contrast that with a modern
- 3:51valuation. He gives the example of valuing
- 3:54MercadoLibre. You know, the massive Latin
- 3:56American e-commerce company in 2026.
- 4:00Today you log into a platform like Capital IQ and in under 5
- 4:04minutes you download every single financial filing
- 4:07MercadoLibre has ever made. Plus the historical data for
- 4:10every single competing retail company in Latin America.
- 4:13Right, straight into an Excel spreadsheet.
- 4:15In 5 minutes you pull more data than a 1980s analyst could have
- 4:19processed in a lifetime. It's wild.
- 4:21And then you feed it into a generic, incredibly powerful
- 4:24financial model and the software just spits out evaluation.
- 4:28All the mechanics, all that friction we talked about is
- 4:30completely gone. It's like it's like giving
- 4:33someone a 500 piece tool kit when all they actually need is a
- 4:36hammer. They get so distracted by the
- 4:38sheer variety of all these shiny tools that they end up building
- 4:42a convoluted, fragile contraption instead of just
- 4:44hitting the nail. You know, they just get lost in
- 4:46the spreadsheets. That is the core issue.
- 4:49Deema Darren argues that removing the friction actually
- 4:52took away the necessity to think Today, analysts don't really
- 4:56value companies anymore, they just do financial modeling.
- 4:59Which is not the same thing at. All not at all.
- 5:01They extrapolate endless columns of historical data into the
- 5:04future, and they falsely equate extreme detail with actual
- 5:08precision. Like more rows equals more
- 5:10truth. Right there is this dangerous
- 5:12assumption that just because a spreadsheet has 300 rows instead
- 5:16of 30, it inherently must be more accurate.
- 5:19But in the process, the actual narrative of the business just
- 5:22vanishes. The real story of how the
- 5:24company actually makes money gets buried under an avalanche
- 5:27of cells. Because when you value
- 5:29something, the numbers you input are telling a story about that
- 5:33company's future. But if you let the tools do the
- 5:36driving, that story gets totally lost in the noise.
- 5:40I think about your own work for a second, right?
- 5:42How often do you get overwhelmed by the sheer volume of data?
- 5:45Oh, all the time. How?
- 5:47Often do you lose the core narrative of a project just
- 5:49because it's so incredibly easy to keep adding more and more
- 5:52information to the file. It's a universal problem.
- 5:55So if we are lost in these massive spreadsheets churning
- 5:59out endless rows of data, what exactly are we putting into them
- 6:03that is steering us so far off course?
- 6:05Because we aren't just making small mathematical errors here,
- 6:08we are fixating on entirely the wrong metrics.
- 6:11Yeah, we really are. Which brings up the specific
- 6:13numbers the finance industry absolutely obsesses over,
- 6:16starting with discount rates and the infamous CAPM.
- 6:19Right the capital asset pricing model.
- 6:21This is where Damodaran really takes aim at academia.
- 6:24For decades now, academic finance has obsessed over asset
- 6:28pricing models to perfectly calibrate discount rates.
- 6:31And to put it simply, for everyone, a discount rate is
- 6:34just the return investors demand for taking on the risk of a
- 6:38specific stock, right? Exactly.
- 6:40And the academic theory is that if you can perfectly calculate a
- 6:42company's required rate of return, you can accurately
- 6:45discount all its future cash flows back to today's value.
- 6:48But the author calls these models a complete dead end.
- 6:52He says 99% of his time is spent on things that actually matter,
- 6:55and discount rates don't even make his top ten list of
- 6:58concerns. It's so true, and here is the
- 7:00statistic that totally proves his point.
- 7:03Roughly 80% of all globally publicly traded companies have a
- 7:06cause of equity. That falls somewhere between 7%
- 7:09and 12%. Wait, really?
- 7:11The vast majority of global businesses are all just clumped
- 7:13together in that tiny 5% window. Yes, with the median hovering
- 7:18right around 8.5%. And there is a logical reason
- 7:22for this clustering. You know, as markets mature and
- 7:25become globally interconnected, the underlying risks that
- 7:28companies face start to homogenize.
- 7:29So a mature consumer goods company in Europe faces very
- 7:33similar macroeconomic risks as one in North America.
- 7:36Exactly. Yet the finance industry spends
- 7:38countless hours and millions of dollars in consulting fees
- 7:42arguing over whether a specific company's discount rate should
- 7:45be 8.63% or 8.83%. OK I have to push back here
- 7:49though because doesn't calculating a highly specific
- 7:53discount rate like exactly 8.83% derived from some complex multi
- 7:57factor model, doesn't that give evaluation scientific
- 8:00credibility? I mean, if we just brush past it
- 8:02and throw a blanket 9% on there, aren't we essentially just
- 8:05guessing? Well, this raises an important
- 8:06question, right, About what credibility actually means in
- 8:09finance. Because Dumb Darren would argue
- 8:10that this is just the illusion of scientific rigor.
- 8:13It's false precision. False precision.
- 8:14OK. Finessing a fraction of a
- 8:16percent on a discount rate is literally a rounding error
- 8:19compared to the heavy lifting of the valuation.
- 8:22What actually moves the needle is getting the company's future
- 8:25cash flows, its growth rates and its profit margins right.
- 8:29Because valuation is an exercise in futuring, right?
- 8:34Like predicting the actual real world operations of the
- 8:37business. Exactly, if you just blindly
- 8:39follow textbook rules for discount rates without
- 8:41understanding the business operations, a Claude bot or a
- 8:45ChatGPT can easily replace you. I mean, AI has already memorized
- 8:49every single asset pricing formula ever written.
- 8:52That makes a lot of sense. The real value is in predicting
- 8:55human behavior and market adoption, not tweaking a decimal
- 8:58point on a generic risk formula. Right.
- 9:00And this blind faith in formulas leads directly into another
- 9:03massive trap, the terminal value.
- 9:05Oh yeah, Which is often the single biggest number in a
- 9:08discounted cash flow model. Because you can't just project a
- 9:11company's cash flows for five or ten years and then a assume the
- 9:14business vanishes into thin air, you have to calculate its value
- 9:18for the rest of time. And textbooks teach this
- 9:21perpetual growth model, typically called the R -, G
- 9:24formula. Which takes your cash flow and
- 9:26divides it by your discount rate minus your expected growth rate.
- 9:30But the problem isn't the math itself, is it?
- 9:33The problem is the fundamental assumption baked into that
- 9:36formula. The R -, g equation assumes the
- 9:39company will continue to grow forever into Infinity.
- 9:43Which defies all common sense and historical reality.
- 9:47I mean, how many companies actually last forever?
- 9:49There are a few. Even if we cap it at 100 years,
- 9:51almost none make it. And if you are valuing a modern
- 9:54technology company, you're usually looking at a 15 to 20
- 9:57year horizon before the technology cycle completely
- 10:00turns over and makes their core product obsolete.
- 10:02So you cannot plug a modern tech startup into an equation
- 10:06originally built for a 19th century railroad that expected
- 10:09to operate a physical monopoly until the end of time.
- 10:11No, you can't, and there is a related failure here regarding
- 10:15mean reversion, which is just the basic idea that extreme
- 10:19things eventually drift back to the average.
- 10:21Let me guess, spreadsheets are terrible at handling this kind
- 10:24of natural evolution? Oh, horrible at it.
- 10:27Let's say you're valuing a really risky new startup today,
- 10:30so you assign it a high cost of equity, like 12%.
- 10:34Over the next 10 years. In your spreadsheet, you project
- 10:37that this startup will grow rapidly, conquer its market, and
- 10:41become a massive, stable, mature enterprise generating billions
- 10:46in predictable cash. OK.
- 10:47So the business matures. Right, but analysts will leave
- 10:50that 12% discount rate frozen for the entire lifetime of the
- 10:53valuation. Wow.
- 10:55If the company becomes mature and stable on your spreadsheet,
- 10:58it's risk profile has fundamentally changed.
- 11:00It should naturally drift down toward the average, closer to 8%
- 11:03or 9%. So it's like assuming a wild,
- 11:06rebellious teenager is eventually going to grow up and
- 11:08turn into a predictable, boring adult, but still charging them
- 11:12the outrageous car insurance rates of a teenager for the rest
- 11:15of their life. That is the perfect way to put
- 11:17it. By leaving the rate frozen, you
- 11:19are building an internally inconsistent fantasy world.
- 11:23The company is simultaneously a mature giant in its cash flows
- 11:27but a risky startup in its discount rate.
- 11:29If these mathematical models are so deeply rigid, and they assume
- 11:33companies live forever while freezing risk and time, how are
- 11:36we actually supposed to view market movements?
- 11:39Because the stock market clearly doesn't follow these textbook
- 11:42rules. No, it doesn't.
- 11:44And that disconnect requires us to understand the most critical
- 11:47distinction in all of finance, the difference between valuing
- 11:52an asset and pricing an asset. There are two entirely different
- 11:56games with entirely different rulebooks.
- 11:58Value is driven by fundamentals, right?
- 12:00The actual cash flows, the tangible growth and the risk.
- 12:03But price, on the other hand, is driven by the market demand,
- 12:06supply, the mood of investors and momentum.
- 12:09Exactly. Behavioral finance doesn't
- 12:11explain value at all. It explains why price can
- 12:14deviate from value, sometimes for incredibly long periods.
- 12:17A major mistake traditional appraisers make is attempting to
- 12:20mix the two or using historical data to dictate what the market
- 12:24should be doing today. And the historical risk premium
- 12:27is a classic example of this. Analysts look at the last 80
- 12:30years of stock market returns to figure out the premium investors
- 12:34demand for owning stocks over safe government bonds.
- 12:37But. Relying on historical data
- 12:38assumes the world is inevitably going to revert to the past.
- 12:42Take the small cap premium, for example, the widely taught idea
- 12:46that small companies inherently return more than large
- 12:49companies. Appraisers still proudly add
- 12:52this premium into their models today.
- 12:54Even though empirically the small cap premium hasn't
- 12:57actually existed in the market since 1981, they are essentially
- 13:00assuming the market is going to magically revert to 1986 just
- 13:04because an old textbook says so. Exactly.
- 13:07Damodarin champions a totally different approach, the implied
- 13:10equity risk premium. Instead of looking backward, you
- 13:13look at the market today, you take the current prices of
- 13:15stocks, you look at the expected future cash flows, and you
- 13:18mathematically back into the risk premium.
- 13:20So you let the market dictate it.
- 13:22Yes, it is exactly like solving for X in an algebra equation.
- 13:27You let today's market tell you what it is currently charging
- 13:30for risk, rather than telling the market what it should be
- 13:33charging based on 1950s data. You have to respect the market's
- 13:37current reality. The 1987 Black Monday crash
- 13:41illustrates this perfectly. Actually, on October 19th, 1987,
- 13:45the S&P 500 dropped 22% in a single day.
- 13:49Huge drop. But there was no massive news
- 13:51event, no war breaking out, no catastrophic earnings reports.
- 13:55The fundamentals of the companies didn't change at.
- 13:57All the businesses were identical, but the price of risk
- 14:00completely transformed. David Aaron takes this a step
- 14:03further with a brilliant thought experiment.
- 14:05Imagine you valued Procter and Gamble the week before the
- 14:07crash. You look at their cash flows,
- 14:09you look at the market's price of risk, and you conclude the
- 14:11company's worth $90.00 a share. Now value Procter and Gamble the
- 14:15week after the crash. The company is exactly the same.
- 14:18They were selling the exact same amount of toothpaste and soap.
- 14:21Does their value stay at $90.00? A stubborn textbook appraiser
- 14:24would say yes. They'd argue the cash flows
- 14:26didn't change so the intrinsic value didn't change.
- 14:28The market is just acting crazy and will eventually realize the
- 14:32appraiser is right. And they would be entirely
- 14:34wrong. Your job is to value the company
- 14:36given where the market is today. The market, which represents
- 14:40millions of investors, woke up and collectively decided they
- 14:44require a higher return to take on risk.
- 14:47The price of risk fundamentally went up, therefore your discount
- 14:50rate goes up and your valuation of Procter and Gamble must go
- 14:53down. Here's where it gets really
- 14:55interesting. Because if you are an appraiser
- 14:57who stubbornly holds markets in contempt, if you look at a 22%
- 15:02drop and say, well, the market is wrong, my spreadsheet is
- 15:04right, you are living in a fantasy.
- 15:06You have to live in the world you are actually in, not the
- 15:09world you wish you were in. And.
- 15:10Navigating that convergence between price and value is only
- 15:13getting harder. You know, 15 years ago, the
- 15:16immigrant hoped that behavioral economics would give us tools to
- 15:19predict how quickly a mispriced stock would correct back to its
- 15:22true value. But now.
- 15:24Today, he admits it's getting significantly worse.
- 15:27Because of the sheer volume of noise, we used to think that
- 15:30pumping more information to markets would make them
- 15:32perfectly efficient. But today, information is just a
- 15:36CEO's unhinged late night tweet or a viral rumor on a forum.
- 15:41Yeah, the data is messier. It's noisier.
- 15:43So massive gaps between price and value can open up and last
- 15:46far longer today than they did in the past.
- 15:49Absolutely. And if market prices are
- 15:51entirely driven by future expectations, mood and momentum,
- 15:56but appraisers are stuck looking backward, they inevitably crash
- 16:00into the biggest backward looking system of all modern
- 16:02accounting. Modern accounting Over the last
- 16:0540 years, accountants have attempted to make the balance
- 16:07sheet act as a proxy for the actual market value of a
- 16:10company. They call this initiative Fair
- 16:12Value Accounting. And Damodarin bluntly calls that
- 16:15an oxymoron. You can either be an accountant
- 16:18rigorously recording historical costs, or you can be an
- 16:21appraiser trying to value the uncertain future.
- 16:23You cannot be both. Right.
- 16:25And the attempt to merge the two has resulted in metrics that are
- 16:29entirely decoupled from economic reality.
- 16:32Take the price to book ratio for example.
- 16:34Historically, this was legendary for finding undervalued stocks.
- 16:38You just compare the market price of the company to the book
- 16:41value of its physical assets sitting on the balance sheet.
- 16:43Sounds logical, right? But for the S&P 500, the price
- 16:47to book ratio has been sitting above its historical 75th
- 16:50percentile for most of the last 15 years.
- 16:53Wow. If you use that metric to time
- 16:56the market, if you told yourself stocks are trading at six times
- 17:00their book value, they're way too expensive, I'm staying out.
- 17:02You would have missed out on an incredibly lucrative decade and
- 17:06a half of. Returns because the metric is
- 17:07broken. Modern accounting is just
- 17:09fundamentally flawed in how it treats modern businesses.
- 17:13Let's look at the core issue here, capital expenditures
- 17:16versus operating expenses. If I am a 20th century
- 17:19manufacturing firm and I spend a billion dollars to build a new
- 17:22factory, accountants treat that as a capital expenditure.
- 17:25It's an asset. It goes on my balance sheet and
- 17:27my book value goes up. But what if you are a modern
- 17:30technology company? Your equivalent of a factory
- 17:32isn't a brick and mortar building.
- 17:34It is research and development. It is the complex code your
- 17:37software engineers are writing every single day.
- 17:40And accountants don't treat R&D as an asset, they treat it as an
- 17:43immediate expense. Which is wild.
- 17:46And there's a historical reason for this.
- 17:48Based on the accounting principle of conservatism, if a
- 17:51manufacturing firm goes bankrupt, they have a physical
- 17:54factory they can sell off to pay creditors.
- 17:56Right, there's a hard asset. But if a software company writes
- 17:59terrible code and goes bankrupt, they have nothing to sell, so
- 18:03accountants conservatively expense the R&D immediately.
- 18:06I mean, that might sound logical for a liquidation scenario, but
- 18:09it completely hollows out the book value of driving technology
- 18:13companies. It leads to these absurd
- 18:15situations where a company like Apple, literally one of the most
- 18:18valuable cash generating machines in human history, can
- 18:22have a wildly distorted and briefly even negative book value
- 18:26of equity. Exactly because they
- 18:28aggressively buy back their own stock and their biggest actual
- 18:31asset, their R&D, simply doesn't exist in the eyes of the
- 18:34accountants. Judging a modern tech giant by
- 18:37its book value is like judging the quality of a modern
- 18:40smartphone by weighing it on a kitchen scale.
- 18:43I mean, you're measuring something, sure, but the
- 18:45physical weight has absolutely nothing to do with where the
- 18:48actual value of the phone lives. You're completely missing the
- 18:52software, the ecosystem, the network.
- 18:54What's fascinating here is that the biggest intangible asset of
- 18:57all isn't even the R&D they've already done.
- 19:01It is future growth. Take an upcoming AI company like
- 19:04Anthropic, for instance. OK, when Anthropic eventually
- 19:07goes public, no matter how clever the accountants try to
- 19:10be, the book value they calculate will only reflect what
- 19:13Anthropic has already done. Which is just a tiny fraction of
- 19:16its actual worth. The market price is anthropic,
- 19:19based on what it is expected to conquer in the future.
- 19:21Exactly. Historical accounting cannot
- 19:23capture the future, so trying to bend accounting rules to reflect
- 19:27market value is a massive, expensive exercise in futility.
- 19:31It generates huge fees for accounting firms doing goodwill
- 19:34impairment tests. Which is just a fancy way of
- 19:36saying they take counts millions of dollars to formally declare
- 19:39that an acquisition they made years ago is now worth less,
- 19:42right? Like the stock market already
- 19:44figured out and priced in months prior.
- 19:46It is all just reverse engineering the past to satisfy
- 19:49regulatory rules. It really is.
- 19:51Pulling this all together, if our spreadsheets have replaced
- 19:54actual critical thought, if our mathematical models are rigid
- 19:58and broken, and our core accounting metrics are
- 20:00fundamentally useless, what happens to us in the long run?
- 20:04Like where is this entire financial system heading?
- 20:07Well, Damma Darin looks 100 years into the future, and his
- 20:11vision is pretty surpering. He envisions a world where
- 20:15valuations are entirely outsourced to machines.
- 20:18You won't hire an analyst, you'll just ask your local bot.
- 20:21What is Microsoft worth? And that bot will have instantly
- 20:24read every financial paper, every book, and every historical
- 20:27spreadsheet ever created. It will generate a perfect
- 20:30looking, legally bulletproof appraisal using all the right
- 20:34academic buzzwords to satisfy the courts and the accountants.
- 20:38Which poses an existential threat to business schools as we
- 20:40know them. David Dern questions if they
- 20:43will even exist in a few decades.
- 20:45For a century, our educational system has rewarded people for
- 20:48memorizing content and becoming highly specialized in applying
- 20:52formulas. But teaching content or
- 20:53specialized formulas is totally useless when an AI can recall
- 20:57them instantly and perfectly. Exactly the real skill moving
- 21:01forward, and Damodaran argues this needs to start in second or
- 21:05third grade, not in an MBA program, is the capacity to
- 21:08reason, to think critically, and to apply basic common sense to
- 21:13complex problems. So what does this all mean for
- 21:15you, the listener? Navigating your own career, your
- 21:18own investments, or just trying to understand the world?
- 21:21It means you have to actively fight the urge to hide behind
- 21:24complexity. It means stepping away from the
- 21:26noise of hyper complex models and thousand row spreadsheets.
- 21:30Your edge isn't in processing more data than the machine your
- 21:33edges and focusing on the logical human story behind an
- 21:36investment or business decision. Because if we lose that, the
- 21:40consequences are profound. The ultimate fear isn't just
- 21:43that the boss will eventually take over the math evaluation.
- 21:46I mean, the math is already gone.
- 21:47Right, it's gone. The real fear is that if we
- 21:50outsource all our critical thinking to these tools, we will
- 21:54lose the fundamental human capacity to ask common sense
- 21:57business questions. If that happens and something in
- 22:00the AIC Jones fundamental logic breaks, our great grandchildren
- 22:04won't even realize the system is broken.
- 22:06They won't know how to ask the right questions to fix it.
- 22:09Exactly. If market prices eventually
- 22:11become entirely dictated by AI algorithms validating other AI
- 22:14algorithms, will the very definition of a good business
- 22:18lose its connection to human utility altogether?
- 22:20It is just a massive global house built on quicksand and no
- 22:24one remembers how to check the foundation.
- 22:26The terrifying thought to leave you with.