Back to the Facebook days:

A $9000 contribution towards a group for a political ad got him fired from his role at Meta.

Meta Executives and Luckey have been fighting this battle over social media for the last few years over the justification of the firing, as Palmer became a mega success with Anduril.

Boz is the current CTO and head of the VR division. Carmack was the CTO at Oculus, and after the acquisition, continued to work with Meta and Team. He is considered one of the best, if not the best, engineers on the planet.

But this got diffused after the Tablet magazine published a profile on Luckey in which they got a quote from Zuckerberg.

And today’s announcement (almost a year later) :

What now?

This is not just a military partnership, which is, I think, super important for the US, but for Meta, most importantly, Palmer Luckey himself.

Disruption of Distribution

Thanks for reading. Do subscribe if you enjoyed it.

One of the consensus view among people today is how large established companies will continue to rise. The PE ratio column in table below represents that perception.

In a complex country like India, because of weak infrastructure, diverse geographies, and limited depth of consumption, establishing new offline distribution for any emerging brand has been difficult.

But this edge is soon going away:

In all the below industries, distribution advantage has already withered away from large players and a new player which controls the demand by directly aggregating customers on its platform has become the new king.

These trend will soon takeover other key consumer oriented industries which traditionally had large nationwide offline distribution through retail shops.

Two key pillars behind this trends are:

1. Buying decision for consumers moving from shop door to mobile screens. The convenience and cheaper prices are major driver of this trend. Customer facing platforms like Amazon are data oriented businesses and have greater insights on consumers, making it difficult for offline players to compete.

2. Small companies because of the software boom of last 20 years have access to same resources as large companies at a very cheap cost. A small business can use ERP like software using Freshworks & Zoho, set online store presence in 5 mins using Dukaan App , integrate payments through Razorpay App and can grab consumer attention by listing on any online marketplace creating a level playing field.

Most of the NC people are active stock market investors.

While the question of what is underlying is an important question for a listed company; where the company produces cashflows by selling goods and services. In crypto, it’s like hitting the nail on the hammer.

Investing for most of the NC people has been limited to stock markets. Using frameworks of equity investing to crypto equates to valuing economies like companies.

In crypto, there is a blockchain (economy) & a crypto coin (currency). So, when you think about crypto, think about a country that has a currency. The currency is used to buy goods and services in the crypto economy (not the real economy). Now there is a limited supply of this currency. While new businesses (exchanges, games, ownership transfer agencies) are getting built on this crypto country rapidly, to buy any goods from these businesses, you need the currency. So demand keeps rising. Hence, the price of crypto coins increases. (Disclaimer: Not all coins are built on economies, like Dogecoin)

Now, unlike normal economies, I don’t need to manage inflation & growth in this country. So, the currency supply has been fixed. Again, constrained supply leads to a price rise.

Networks and monetization

We spent the last 10 years building networks on the internet. Social media platforms like Instagram, Twitter, YouTube, and Discord are networks, which can be divided into billions of smaller networks, consisting of followers, friends, subscribers, backers, etc. These platforms gave many people an audience who didn’t previously have one.

We are now entering a new era of the internet, where we have the chance to upgrade these networks into economies.

How crypto enables the conversion of networks into an economy?

Let’s try and understand this through the music industry.

Today there are over 8 million musicians on streaming services, yet less than 15,000 musicians (less than 0.2%) make more than $50K/year. That’s because only a few people qualify to get a contract with a label company. A music label company (think T-series) looks after music distribution, marketing & production. It widely increases your reach. It has a lot of money to spend on marketing. But then only a few musicians bypass the funnel to reach a label company and sign a contract. Even the successful artists signed up by the label company get only 10-15% of total revenue earned by selling their music.

Hence songs like this end up having less than 10k views on Youtube.

How do we solve this problem?

Musicians already have a network of people following them on many social media platforms. Only if these fans had a structure to invest in musicians so that they can have enough money to produce and market their music to a wider audience. Fans get royalties in return after the musician becomes successful. Regulations don’t allow crowdfunding for anything other than charitable purposes. Guess what, crypto makes it possible.

Audius - A brand new crypto streaming platform for any musician to launch their career with no music label deal. A creator can raise money from its 1000 true fans and launch their career in Audius. When the creator becomes big, the 1000 true fans will receive royalties for being the first people to believe in the new creator.

And why just musicians. Isn't it possible to invest $3000 in the education of a worthy neighborhood child to support his ambition? And to get an upside in all this. Say a 2% equity stake in his company or 5% of his salary for the first 3 years if he earns a minimum of $20k per year and returns capped at $10k. Win-win for both parties.

Disintermediation & The problem of trust

Disintermediation refers to the removal of middlemen or intermediaries. Why do we need intermediaries? Because we can’t trust each other. I can’t trust a random businessman and lend him INR 5 lacs. So what do I do? I deposit the money at a bank that gives peanuts in savings interest. Banks then lend the money to the same businessman at much higher rates. Now, these banks have heavy legacy systems whose operational cost eats up 2% additional interest. And then they have to make profits. Both parties end up losing.

Now, what if you can trust an unrelated business person such that it’s not possible for the person to default on his obligation. Again, crypto has the solution.

Aave: Aave is a completely automated, permissionless platform, where there are no bankers, no buildings, no toasters, no vaults, no cash, no holding your money, no forms to fill out, no credit ratings involved.

Everything is controlled by smart contracts. It’s fully automated. You don’t have to get approval from anyone and it takes minutes to take out a loan. Collateral gets automatically transferred to the lender in case of a default.

Foreign remittances take an average of 2-5 days costing 2-5% of the amount transferred on average. Crypto enables real-time transfers costing users only pennies ($1).

Industries that rely on intermediaries include (but are not limited to) the following:

1. Insurance – Property and casualty, auto, life

2. Financial Services – Consumer/retail banks, investment banks, brokerage, etc.

3. Transportation and Logistics – Services associated with supply chains, including exports, imports, logistics, and related services such as finance, fund transfers, contracts, and Forex.

4. Retail and real estate – Trade of durable and non-durable goods, which relies on huge margins and delays consumed by a system of intermediaries established centuries ago

Intermediates power everything in our world. Crypto makes these intermediates go away. Makes the process faster and cheaper.

This is due in large part to the costs, inefficiencies, and barriers imposed by the traditional financial system.

This idea of transferring ownership via the blockchain is extremely powerful. Today we trust centralized exchanges & intermediaries to transfer ownership of assets for us, paying hefty commissions & fees to ensure trades are executed & processed as ordered. If we can do the same thing using the blockchain, without giving up control of our assets, what long-term value do banks & other intermediaries really bring?

Credit Card Network

Suppose, I'm buying an iPhone 13 from Reliance Digital (RD). So it's obvious, I need to give money to RD, but the way that works is I'm buying it with an HDFC credit card. It's called an issuing bank. HDFC's a bank, they gave me a card, the card has a credit line attached to it. RD Store has a bank, it's also known as an acquiring bank. It's called an acquiring bank because they onboarded the merchant on their platform. Let's just say that RD uses Axis Bank merchant banking services.

And then there's a network that connects the acquiring bank to the issuing bank. So if I buy my $1000 iPhone13 at RD, what happens is, RD swipes that card, it goes to the Axis Bank. Axis Bank merchant services see, "Oh, this card starts with a four, that means it's a Visa, we got to send that over to Visa, if it starts with a five, we have to send to MasterCard. Visa sees, "Oh, it's 404276, That's the BIN, Bank Identification Number. That means Visa sends it to the issuing bank, HDFC. And then HDFC bank decides whether they will approve this transaction as HDFC bears the ultimate credit risk if the customer doesn’t pay back & hence collects most of the interchange fees from RD.

Reliance Digital generally pays a ~2.5% fee as interchange fees which is distributed among HDFC, Axis, VISA & customers (reward points). So, if you pay $100, RD will get only $97.5. You can watch the full explanatory video using this link

Visa ($500bn) & Mastercard ($350Bn) have a complete duopoly in card markets.

Source: Plaid, Joincolossus, A16z

The need for BNPL

Now the problem with these networks is that no one has any idea on customer purchasing behavior.

Nobody knows what SKUs (whether iPhone13 or Samsung S20) are being bought. That's crazy. Because there's literally technologically no way for Reliance Digital to send back to Axis Bank, to send back to Visa, to send back to HDFC, to put in my app, what items I bought at RD store. And a lot of this is just based on how old this technology is.

My credit card statement below will tell you what is the data banks can capture with no details as to what goods were bought:

This is what makes BNPL so interesting. It’s a **parallel** network, with item-wise information, that bypasses the issuing bank, card network, and merchant acquirer bank. It’s just the consumer, the merchant/shopkeeper, AND a new participant: the product manufacturer!

With individual item wise data, a product manufacturer (say Samsung) can offer multiple financing options for different products which was not possible earlier with credit cards.

How BNPL is solving the problem?

BNPL or buy-now-pay-later is a credit product that enables a user to purchase a product by paying for it in installments as opposed to paying 100% upfront.

1) For a Consumer, BNPL is amazing as it helps them buy the products they desire and don't want to pay entirely upfront. While the amount of information a user must supply varies by product, all BNPL providers provide a streamlined, low-friction experience — often requiring no additional fields beyond a merchant’s standard checkout flow.

2) For a Merchant as well, it is productive as BNPL increases conversion rates (visitors to purchasers) and increases average order value (users can afford higher value products). For the BNPL platform, it is great as it knows the use case of the loan, has access to customer data to underwrite, and can work with the platform to collect the loan. BNPL as a payment option can help lower shoppers’ hesitation and thus reduce cart abandonment.

The BNPL platform partners with the Merchant/Brand and takes a commission of 5-10% of the order value. The user doesn't pay any interest amount over and above the order value but pays one-third or one-fourth of the order value upfront and the rest in installments over 3 to 4 months.

Here is an exhibit showing how Bajaj Finance makes money by offering BNPL products.

The key elements of BNPL are:

• typically smaller purchases

• no credit check/traditional underwriting. May use ‘soft check’ as part of identity verification

• BNPL services do not furnish data to credit bureaus (this debt is invisible to other creditors)

• no interest or fees to the user (if paid on time)

• typically, the user pays 25% at the time of purchase and three additional payments in two-week intervals

• user links BNPL service to a debit/credit card at time of purchase to automatically draw these payments

BNPL Landscape

Largest Players with valuation: Klarna ($46Bn), Afterpay ($29Bn) & Affirm ($32Bn).

Why BNPL can overtake credit cards?

1. Business model built on consumer welfare

   Most BNPL schemes charge a 4-6% commission on pay later sales from merchants, and this forms the majority source of revenues for them.

   Unlike credit cards, where the revenues for network processors rely on keeping consumers in debt and revolving their balances, the BNPL schemes encourage consumers to repay in time by setting up incentives such as freezing customer accounts on payment defaults. This is an important distinction.

   With the revenues reliant on sales commissions, the business model ensures that all efforts are aligned to reduce repayment defaults and to increase sales frequency and transaction sizes. Moreover, while the merchants are expected to pay a high share of sales as commission for Pay Later, the expected increase in conversions at checkout and repeat purchases make BNPL a desirable payment option for retailers.

2. Convenient On-boarding

   Another differentiator for BNPL schemes is the ease of signing up, where a customer only needs to set aside a few minutes and give out basic information to set up an online account. On the other hand, credit cards require you to fill an application form with multiple documents and to wait days to complete registration. This convenience is well-appreciated by Gen-Z and young millennials.

   To add, the target segment for BNPL schemes has particularly been the low-income and younger crowd, who lack a good or long credit history for traditional credit cards or personal loans. Buy now, pay later – although performs credit checks – allows for more flexibility in issuing credit, tempting consumers to avail themselves at the checkout. And contrary to belief, despite lower credit checks, the schemes have been successful in keeping the default rates down to less than 5%.

3. Transparency in Costs

   Young people’s distrust of established financial institutions is at the core of many fintech trends today including neobanks and digital lending. For pay later options, the transparency in costs and benefits upfront is pushed as a major incentive and benefit versus the credit card – which traditionally has been seen as highly opaque. Validating this narrative, young customers increasingly prefer not to hold credit cards or use revolving credit forms today, and see BNPL as a more convenient alternative to the long-term commitment or high interest rates that credit cards feature.

4. Awareness of End Use

   With the lack of a network processor to allow acceptance on millions of merchants for BNPL schemes, the merchant agreements are formed based on the preferences of target customers. While this limits the applicability of BNPL across retailers, the lenders such as Afterpay and Klarna assume more control and visibility of the end-use of the money by clearly specifying the merchants that offer BNPL service on their apps and websites. This is unlike the traditional loans and credit lines, where the lender has no control over where the fund is used.

   Combined, all these factors help distinguish BNPL from a traditional credit product. In addition, the overwhelming shift to mobile shopping for young millennials and Gen Z customers along with applications on big fashion, food, and entertainment merchants make BNPL an option more favorable for many than the traditional Credit Cards or Pay Later alternatives.

   

   References & Further reading:

   Alex Rampell (A16Z)

   Osborne Saldanha (Fintech Inside)

   Prince Jain (Unit Economics)

   BloombergQuint BNPL
   

Yes. Bitcoin. The Digital Gold.

But if you are like I was a year ago. And think that is the defining play here. Maybe there are some facts you should know.

Bitcoin is dumb. That’s not meant to be a criticism or negative at all. What it simply means is that Bitcoin is designed to do one simple thing, and that one simple thing is hold Bitcoin. OK that’s not strictly true. You can also transfer it to someone else. And you can also write a post about it on Linkedin. But there’s not much beyond that and that’s fine. It does its job extremely well.

Bitcoin is a specialized blockchain. It’s specializes in providing digital ownership of a coin having a scarce supply. And that is useful if you want to be the digital version of shiny yellow metal.

Enters Ethereum

Ethereum is a general-purpose blockchain. Besides what you can do with the Bitcoin, you can also develop applications on top of it.

You may think, ok so some people are developing applications. So what? I don’t see this technology anywhere.

Let me draw a comparison to explain the adoption

The reason we enjoy our iPhones is because of apps that developers build. Instagram, Tiktok, Zerodha, Spotify, Uber and others. Well..

iOS, the operating system for iPhone had 200,000 developers making apps for iPhone in 2012; 5 years after its launch in 2007. Ethereum has crossed the same number in the same time.

Ethereum is a decentralized blockchain like Bitcoin, but is far more programmable, meaning other applications and currencies can be built on top of it. Its native cryptocurrency is Ether, which today stands as the second most valuable cryptocurrency after Bitcoin. Perhaps even more interesting, transaction fees on the Ethereum network have recently approached $15 million per day, meaning there's a real and growing demand to use the Ethereum network.

Amazon web services provides infrastructure like servers and platform for startups to build. At the core level. Ethereum is same and much more. It's just a cloud computing platform. Like AWS, you build and deploy apps to it. Also it’s open source and uses blockchain infrastructure.

Some key financial metrics for Ethereum for the stock market nerds:

Total transaction fees, also known as network revenue, increased 200x to $1.7 billion in Q1 2021, compared with $8 million in Q1 2020. For the month of April, Ethereum generated annualized revenue run rate of $8.6 billion—comparable to AWS in 2015.

Total transaction volume on Ethereum increased 20x to $713 billion in Q1 2021, compared with $33 billion in Q1 2020.

Daily active addresses, a proxy for daily active users, increased 71% to 607k in Q1 2021, compared with 364k in Q1 2020

Twitter: James Wang

The Banking Killer

Decentralized finance(DEFI) applications built over blockchains (making deposits, lending them etc) has already crossed $80bn in AUM in just the last 6 months more than HDFC Bank’s loan book after almost 3 decades of existence.

The ability to seamlessly transfer assets from one person to another around the world (in mere seconds) at a low cost is a feat that traditional financial systems cannot efficiently accomplish with current infrastructure. Even simple money transfers can be typically delayed up to 5-7 business days if a transfer requires multiple leaps on various international transaction networks, with no promise of success.

The biggest underlying issue is the “cost of trust” that we currently pay. The entire finance sector is built on it. Finance sector is like a pipe that transfer money and risk throughout the world. To use the pipe, you pay a fees. The way that trust is mediated today is largely through companies, corporations, institutions and, to large extent, mediated through legal contracts. So merchant pay fees on card transaction to Visa and banks, business pay higher lending rates, investors pay brokerage.

Ethereum is a coordination platform for people on the Internet, who don't necessarily trust each other or know each other. It's kind of extending the superpowers of the Internet. The Internet, in its first stage, gives us incredible communication powers with real-time video and whatnot. Now, we're extending this to the ability to be this trust platform for people who don't trust each other. The whole thrust of blockchain is to remove intermediaries and remove the cost of trust we pay while trusting on the blockchain to act as trustless network.

Decentralized protocols provide a monetary system as a service that anyone around the world can tap into without significant barriers to entry. Products like credit, investments and even insurance are available with internet access as the only gatekeeping factor. 

We don’t realise but over 2 billion people in the world still don’t have access to basic financial services:

• Some people aren't granted access to set up a bank account or use financial services.

• Lack of access to financial services can prevent people from being employable.

• A hidden charge of financial services is your personal data.

• Money transfers can take days due to internal human processes.

• There's a premium to financial services because intermediary institutions need their cut.

• Governments and centralized institutions can close down markets at will.

• Trading hours often limited to business hours of specific time zone.

More complex transactions are almost unthinkable. Influencers and creators should be able to monetize their likes, retweets and views, with micropayments streamed from followers, without any platform taking a cut. People should get paid if they opt in to view ads or consent to sharing their information. Transferring ownership of valuable assets, from art to real estate, shouldn’t take several intermediaries and tons of paperwork.

The second big problem with the internet today is that we access it through a handful of companies with “walled gardens.” Like Apple, Facebook, Google. Not anyone is free to build on these lands. Developers have to stick to specific guidelines and go through an approval process that can take days, knowing that rules can change without warning, and that their app submission can be rejected.

This is a barrier to innovation and it’s going to change. A developer who gets a new app idea on weekday should be able to launch it on Weekends.

Ethereum has power to permit new protocols for company organization, media distribution & funding, innovative marketplaces, decentralized finance and much more.

“We cannot take for granted that the future will be better, and that means we need to work to create it today.” - Peter Thiel

References:

1. James Wang

2. Justin Drake

3. Ethereum Foundation

4. Bloomberg

Life can only be understood backwards, but it must be lived forwards.

Let’s dive in.

What is Future?

Something which is going to happen, in foreseeable time, and has not happened yet.

Why should we know about the future? Can we not live without it?

You can, but you lose out on some benefits. To give you a perspective let us presume you would be well positioned to find a prospective employer if you knew of a job opening before everyone else did and acted accordingly.

Now suppose you become aware of a vacancy in Mckinsey. You invest a lot of time to identify connections in Mckinsey i.e. you took actions about a probable future. It however turned out to be a rumour. So, it means not only knowing about future helps, but knowing the right things about future helps.

Suppose now there is genuinely a vacancy at Mckinsey, and one of your friends sent it to a WhatsApp group of Chartered Accountants. Now everyone knows about it. Do you think you stand the same chance of getting the job now? Probably not.

So, we conclude, not only do we need to know right things, but we must know them before everyone.

In other words. We must find a secret.

What should you find a secret for? Money seems like a good incentive for a secret. Now let us see what kind of secret brings money. Knowing a secret about a business brings money. So, then you can invest or build the same business.

There are two kinds of secrets that may interest us: secrets about nature and secrets about people.

Natural secrets exist all around us; to find them, one must study some undiscovered aspect of the physical world. Secrets about people are different: they are things that people do not know about themselves or things they hide because they do not want others to know. So, when thinking about what kind of company to build or invest, there are two distinct questions one shall ask: What secrets is nature not telling you? What secrets are people not telling you?

Now let us talk about 2 gentleman - who have identified a lot of secrets in their lives and made a lot of money out of it too.

Peter Thiel and Warren Buffett.

We will try to analyze their Top Fundamental Ideas to identify secrets. What makes them different and finally, whether there are any common grounds.

Buffett invests in businesses which will look the same 10 - 20 years from now - think Coca Cola, Dairy Queen, Kraft Heinz. He tends to invest in industries which will not see any change (AKA innovation) for decades.

Peter Thiel, on the other hand, invests in companies that want to innovate and change an entire industry or create a new one. This has worked out for him with many of his investments, like Facebook, SpaceX, Spotify, Airbnb, and Lyft.

Important point: Both made money, Buffett even more.

Secrets in nature involves understanding physics. Turns out Buffet uses more physics to understand natural secrets than a Physicist would use in his life.

The Idea he uses the most is Inertia.

Newton’s second law of Motion states Force equals mass times acceleration. (F = M*A)

Let us understand how it affects products and consumption. Suppose a new product attracts us with force say 12(=4*3), when you continue to consume the product over time, your mass (M) i.e. tendency to continue consumption for existing product increases (say M goes from 4 to 6 over time).

Now if a new product is introduced with same force of 12, your acceleration (attraction) towards new product is only 2(=12/6) rather than 3(=12/4) earlier. Hence unless the product come with a Higher force (higher attraction i.e. better product), you will not move towards the new product.

It is always easier to change your name at a young age, than old. That is Inertia.

The longer we have been buying something the more ingrained the product becomes in our lives. (Bournvita, Maggi).

Buffett bets against change. The less the world changes, the more Buffett succeeds.BNSF Railway, where Buffett invested $44 Billion operates a freight railroad network. Around 40 percent of railroads involve the transport of coal, so Buffett’s investment will do especially well if the travel and energy consumption patterns of the 21st century look like the past.

The history of invention is a story of the dismissal of new ideas. From radio and the telephone, to cars, airplanes, and laptop computers, many life-changing inventions were dismissed initially as irrelevant or useless. People were so ingrained to earlier beliefs that they rejected switchingearly. (Think Bitcoin as currency of the world for everything – see how you reject it)

Enters Peter Thiel.

Thiel believes best companies create their own sectors and that Technology is the fundamental driver of growth in Industrialized world.

It does not matter how often you are right or wrong. What matters is how right or how wrong you are. People over-weigh the rightness of an outcome and under-weigh the magnitude. It is okay to be wrong often if the cost is low. You only must be right once if the gain is big.

Principle 1: Better to risk Boldness over Triviality.

Thiel believes plenty of money and investors already pursue companies with more moderate ambitions (simple businesses) and a lower (perceived) degree of risk. This tends to push up demand and valuations for those companies and correspondingly depresses returns (Think buyers purchasing at higher price). Also, less ambitious firms, almost by definition, do not change the world.

Another, paradoxical reason, a company with a readily obtainable goal (say Food for pets online) lacks a technological barrier to entry (Think SpaceX – Sending reusable rockets) because, of course, the original problem was easy. And their end markets are typically quite small, meaning that they may not achieve the scale necessary for providing large exit to an investor. (Pet food online has smaller market than possibilities in space)

The biggest secret in venture capital is that the best investment in a successful fund equals or outperforms the entire rest of the fund combined.

Principle 2: Founding teams are important.

No company knows how their company’s business model will evolve over time. When investing in a start-up, you invest in people who have the vision and the flexibility to create a success (PayPal – Thiel’s founding company changed their Business model 5 times).

On Contrary, Buffet does not like to invest in companies whose success is based on the smartness of its people. He thought of a company's processesas the place where its value resides.

While Buffett waits for amoat(competitive advantage) to be established before he invests, Peter Thiel invests based on the potential of a future moat.

Thiel, by contrast, is the un-Buffet. PayPal and Facebook are pretty much the antithesis of Buffet-style investing. Both companies had no(perceived) value to offer when they started.

Investment in Facebook and others was driven by a founding principle of Thiel. Mimesis.

Mimetic Theory rests on the assumption that all our cultural behaviors, beginning with the acquisition of language by children are imitative. Thiel sees the world as a theatre of envy, where, like mimes, we imitate other people’s desires. Hence Facebook. Remember – why you signed up for Facebook. Because all your friends were already there. This is Mimesis.

While there are lot of differences in how they invest, there are quite a lot of similarities underlying their approach, which makes them great investors.

Both do unpopular things. Remember, we discussed, nobody should know about your secret.

Both Buffett and Thiel are IndependentThinkers. It is not without its risks though, because there is no cover from the crowd and because it frequently leads to conclusions with which no one else agrees.

Secondly, both work from First principles, i.e. thinking from scratch. Both Buffett and Thiel evaluate business from a fresh perspective.

Boiling things down to fundamental truths (e.g. Physics), and then reason up from there. They do not rely on reasoning by analogy. Both understand the fundamental data, assumptions, and reasoning by themselves.

Let me explain.

So, if you compare market crisis of 2020 from 2008 and say that market will eventually rise in 2-3 years. You are assuming by drawing an analogy from past.

The right way to think goes something like, what is a virus? How much value a business loses in a lockdown? How much demand for Future is affected? How much time can people survive with their savings? These and other fundamental question give you facts, on which you should finally base your judgement.

Conclusion

There are only two ways of predicting the future better than the probabilities of chance. You can follow the patterns of history that do not change in time. Or you can invent it through agency.

References:

1. David Perell’s Blog

2. Zero to One Book

I think about Intel a lot. Its early creation can be traced back to world war 2. The company’s founders invented the modern integrated circuit which is responsible for all data processing in the modern world.

Intel was a leader in memory business until 1984. Think RAM. 80% market share. Then Japanese entered.

Japanese companies were able to build enormous factories to build out these memories. Their economies of scale on building and their access to capital at the time was just enormous.

Here’s a conversation between Japanese manufacturers:

“Win with the 10% rule, find AMD and Intel sockets.” “Quote 10% below their price. If they requote, go 10% again. Don't quit until you win.”

It was so crazy that Intel usually had the best technology and the best chips. They were 6–12 months ahead of the competition. But because these are physical chips that have to be created and Intel didn't have enough production resources to meet demand, they had to source them from other manufacturers, most of them were Japanese who then rapidly gained competitive edge.

Now Intel has a big problem. Memory business is incurring huge losses. It has become a commodity. They need to solve the problem.

They have been internally working on something called as microprocessors other than the dominant memory business.

But they have no production capacity. The PC where microprocessors were used was not a thing yet. It was like Virtual reality headsets today. Application was limited to few enterprise consumers. They took the bet. And completely made the transition. The CEO was Andy Grove.

Intel vs AMD

Intel started designing microprocessors now. They had an existing contract with AMD for production of chips to cover the production shortage earlier. But they realised it will cost them margins.

Intel stopped supplying designs to AMD. AMD went to court. Intel won.

As a result , all the PC manufacturers used Intel 386 processors. It took until 1991 before AMD got a competitive product out in market. During the whole first six years of the PC wave, Intel is the only player providing CPUs—the most important component of the computer.

As soon as AMD came, Intel went to all the PC manufacturers—Compaq, Dell, Gateway, IBM—and said "You know all the marketing costs that you guys are spending to market your own PCs competing against one another? We will pay 50% of your marketing cost if you include in your marketing collateral that you have Intel inside." Woww. And this was beginning of the biggest revolution in computing industry.

Beginning of Intel’s problems

June 6, 2005, was the most triumphant moment for Intel. The chipmaker was already dominating the market for processors that powered Windows-based PCs. Then Steve Jobs took the stage at Apple's World Wide Developers Conference to announce that he was switching the main Windows alternative, Macintosh computers, to Intel chips as well. The announcement cemented Intel's status as the leading company of the PC era.

There was just one problem: The PC era was about to end. Apple was already working on the iPhone, which would usher in the modern smartphone era. Intel turned down an opportunity to provide the processor for the iPhone, believing that Apple was unlikely to sell enough of them to justify the development costs.

As I wrote in my previous post, Intel invented a chip standard called x86 that was chosen for the IBM PC in 1981 and became the standard for Windows-based PCs generally. As the PC market soared in the 1980s and 1990s, Intel grew with it.

The key to success in the PC business was performance. Chips with more computing power could run more complex applications, complete tasks more quickly, and run more applications at the same time. During the 1990s, Intel and its rivals raced to increase their chips' megahertz ratings — a measure of how many steps the chips could perform in a second.

One thing these early chipmakers didn't care about was power consumption. Higher-performance chips often consumed more energy, but this didn't matter because most PCs were desktop models plugged into the wall. Even laptops had large batteries and could be plugged in most of the time.

But this became a problem in the late 2000s, when the market began to shift to smartphones and tablets. These devices had smaller batteries (to keep the weight down), and users wanted to use them all day on a single charge. Existing x86 chips were a poor fit for these new applications.

“Early PCs based on Intel chips were referred to as microcomputers, and companies like DEC dismissed them as toys. DEC used to manufacture computers which was not exactly room size but far bigger than PCs. They dismissed microcomputers for exactly the same reasons Intel dismissed the mobile market — selling a $2,000 PC was a lot less profitable than selling a $50,000 minicomputer, and DEC didn't expect PCs to be a big enough market to be worth the effort.” Intel suffered a similar fate itself when mobile phones came.

Intel’s problems were manifold:

• Intel has lost Apple’s Mac business thanks in part to the outstanding performance of the latter’s M1 chip.

• In a similar vein, AMD chips are now faster than Intel on the desktop, and extremely competitive in the data center.

• Large cloud providers are increasingly investing in their own chip designs; Amazon, for example, is on the second iteration of their Graviton ARM-based processor.

Intel’s possible reemergence

Look at the picture below:

The picture above depicts concentration of semiconductor manufacturing in Asia. Intel owns the majority capacity outside Asia.

As I mentioned in previous post, chips matter for many use cases outside of PCs and servers which is to say that TSMC(Taiwan based) who manufactures majority of it matters. Nearly every piece of equipment these days, military or otherwise, has a processor inside.

This is a big problem if you are a U.S. military planner. And in that planning the fact that TSMC’s foundries — and Samsung’s — are within easy reach of Chinese missiles is a major issue.

New CEO Pat Gelsinger for the first time in a decade accepted Intel’s fading role in semiconductor market. He announced:

• $20bn investment in Foundry capacity in United States. US govt is estimated to give many incentives as semiconductors are very strategic geopolitically.

• Outsourcing production to TSMC. (This solves a major problem as Intel’s manufacturing was unable to keep pace with the new semi technology and other designers like AMD moved ahead with TSMC as partner)

• Offering their foundries for competitors like Apple, Qualcomm who are only in designing. Now the foundries business would have a separate P&L. (Till now, Intel was manufacturing for itself)

This is significant as he is ready to leave Intel’s existing image of an integrated manufacturer behind to move forward.

Thank you for reading!

References:

• Stratechery

• Acquired

• Vox

People say data is the new oil. I disagree.

The only thing which can closely match the characteristics of oil is semiconductors. Every device which takes any type of input data to give you output data needs a semiconductor or a processor to process it.

If semiconductors are the new oil. Taiwan is the new Middle east.

Modern semiconductor manufacturing is at least as important to the economy as oil was in the 1970s. But in the case of oil, at least it was available all over the world albeit at higher prices than in the Middle East. Imagine a world where oil only came from one country, and how important that country would have been for the last hundred years. Taiwan could become by far the most geopolitically important country in the history of the world.

You might think why Taiwan. Why can’t someone replicate these manufacturing facilities at Taiwan. It takes around $1bn to as much as $7bn depending on complexity of semiconductor getting manufactured to set up a plant along with other auxiliary industries and high tech expertise.

China plans to invest $1.4 trillion only in technology investments by 2025. (India’s total GDP ~$2.5 trn).

In the past, the semiconductor industry has gone through many negative demand shocks. The biggest being cloud computing. Suppose you are a video editor. Earlier you use to store data. Process data. All in your personal computer. Cloud which is a remote storage and processor allows you to rent a piece of storage and processing power equal to what you require. Not 10 gm more not 10 gm less. Exactly what you need on cloud. Now you don’t have to maintain 10 Terabytes of storage capacity or processing power locally. You just call your server provider and ask him to increase your storage. So now nobody keeps large storage capacity individually but everything is centrally located on cloud and hence demand for semiconductor reduces.

Amazon web services(AWS) is the cloud provider for Netflix. It provides millions of people video on demand without maintaining huge storage and processing power. Whatever is the demand, AWS can fulfill it.

Strangely enough, the iPhone and smartphones, in general, were another negative shock. That’s because they slowed down the growth of the PC market, and PCs are a lot more semiconductor intensive than smartphones.

Now, we have a positive demand shock in the form of artificial intelligence. Because data quantity is so important for quality for any artificial intelligence system. Google’s engineers started to deploy voice recognition based on AI to Android smartphones. They realized that if each Android phone used Google’s voice search for just three minutes a day, Google would need to build twice as many data centers(servers).

Intel Inside going outside

The semiconductor industry has been segmented into three major entities — Integrated Device Manufacturers (e.g. Intel, AMD (pre-2009) and Samsung) - both manufacturing and design of chips, fabless companies (e.g. Apple, Qualcomm, Broadcom, Altera, NVIDIA) - only designing chips and foundries which make chips for the fabless companies (e.g. TSMC, UMC, Samsung, GlobalFoundries) - only manufacturing.

Historically, Intel(Mkt cap - $261bn) and AMD focused on making CPU-based chips while NVIDIA(Mkt cap - $328bn) focused on making standalone graphics chips (GPU) for the PC and server markets.

CPU is designed to handle a wide-range of tasks quickly (as measured by CPU clock speed), but are limited in the concurrency of tasks that can be running. A GPU is designed to quickly render high-resolution images and video concurrently.

Now Intel use to dominate the CPU market and through it the desktop and laptop market. Because it use to produce high processing power chips. Better than anyone else.

But when mobile phones were introduced, since they were designed to be small and hence could not hold much battery arising need for better CPU which consumes less power.

Enters System on Chips(SoC)

So basically what a SoC does is, it focuses on a combined hardware solution, such that hardware for specific task are inside the chip and software in your phones do less work and consumes less power and save your mobile battery which is already small. Long battery life.

The smartphone offered the first significant platform for SoC technology to demonstrate its potential and put the SoC on a collision course with the standalone CPU.

The Average selling price for an SoC chip (NVIDIA Tegra or Qualcomm Snapdragon) is in the range of $10-20 while for a leading edge Intel CPU chip (Skylake) is in the $150 range. The CPU chip will also need to be supported by other chips (e.g. power management) to provide all the functionality that is provided by a single SoC. SoC has everything combined in itself. As I mentioned hardware does the job and software is used less. Saving power.

Hence Intel was disrupted by Qualcomm and Nvidia in mobile phones. Also in its primary markets for servers(cloud) where it has to focus on core CPU, it went behind the competitor AMD.

AMD interestingly divested its manufacturing business in 2009 and shifted focused on design and moved ahead of Intel by bringing better designs. It outsourced manufacturing to TSMC. TSMC left Intel behind in manufacturing because TSMC was only focused on manufacturing. Intel from pioneer went suddenly behind in every thing.

The story of Intel has many lessons. More later.

Thank you for reading!

References:

1. Gavin Baker

2. Pushkar Ranade

Advertising is the most important variable to any business. Without it, no one knows you. No one understands what problem you solve. No one cares.

Advertising for small businesses and startups used to be so difficult. You could only do mass ads which will cost you your house. You had no knowledge of reaching out to a target audience.No knowledge of what they like or want.

Enters Facebook.

And brings with it the ability to send targeted ads to potential customers by studying their behavior on its platform based on the activity of users. Ads were boring and useless until Facebook filled them with utility.

Let’s zoom in on Facebook.

Attracting Talent

Peter Thiel and Marc Andreessen are torchbearers of global innovation. Facebook is not a current portfolio company for any of their fund. Both sit on the board of directors. Get me one similar company in the world. Top executives have hardly left despite so many controversies surrounding it. Insiders know more than outsiders.

Superior execution

Mark Zuckerberg does execution like nobody else in this world. You might see him as copying stories feature from Snapchat, short videos from TikTok, hashtags from Twitter. No one ever talks about why the others haven’t built a platform like Facebook which can simply copy the best ideas. Wouldn’t it be the best idea to work on in hindsight?

Augmented reality and virtual reality

Platforms for computing have continuously evolved from mainframe to personal computers to laptops to mobile. What we right now have in form of mobile is almost the most matured version. Nothing fundamentally revolutionary can happen with mobile phones now. They will continue to be more or less the same.

Augmented reality just makes things look more real. It’s a game-changer for education, retail, designers, etc.

Similar is a foray into virtual reality. While we continue to embrace working from remote locations, not having a shared space is still what we miss. Facebook is pioneering the VR revolution. Quest 2 virtual reality headsets have sold over 5 mn units since launch in the last 6 months.

There is one more interesting thing. Which I love about Facebook.

The end of every activity is to sell a product to consumers. That’s the end-use. Companies sell products and earn revenue. Then they give salaries, pay for advertisement and raw materials. On the opposite side are many companies who provide these backward services of advt, supplying raw material to final good suppliers in order to make the sale possible.

Now you see 2 companies - Facebook and Amazon.

Facebook does advertising. A backward service to enable the sale of the final product. Amazon sells the final goods. Facebook has opened its marketplace and is trying to sell goods directly. Because obviously, they have data on what you want. It made $1bn in revenues and growing at over 100% Y-O-Y. Still early days.

Amazon sells goods. It is getting into advertising. Because it knows what you want to buy. Since you search it on the platform. So they sell that data to advertisers who can book the top spot in your search results.

I think going to Amazon is a waste of time more so because there is no direct utility that you must go to buy the good. What if what you want is available on Insta at the cheapest price with next-day delivery. You will simply switch. Instagram has better data. It can serve you better. Plus you are already using Instagram every day.

The only reason you don’t go is it is not yet made possible due to the lack of basic infrastructure. Logistics, payments, sorting, warehousing, tagging all that which goes into a good reaching your house from the producer. It’s not integrated and seamless yet.

But once it’s done. It just becomes plug-and-play for Facebook.

Think like this 20 years back if you wanted to purchase something from someone - you will go to their shop, withdraw cash from your bank acc and then give cash, the person deposits cash in his bank acc. 15-day process. Now a person opens his digital store using the “Dukaan App” in 5 mins and integrates UPI in another 5 mins. And he is ready to sell and receive money in his account.

I agree it’s difficult for Facebook to become an e-commerce giant yet, but not more than what Amazon would need to become a social media company.

Question is what's more sticky.