Why the Fed Can Look Hawkish. The Pieces of Financial Repression Are Quietly Falling Into Place.

Posted on by

America owes almost $40 trillion.

To put that in context, America spends more in defense than the next 6 countries combined. Just the interest to service America’s debt eclipsed the massive defense budget in 2024. Roughly a fifth of every tax dollar you send to Washington goes straight back out the door as interest.

What’s making this worse is that most of the debt was issued in 0% years, whereas the short term interest rates now are at 3.75%, which means the interest bill is further set to rise, aggressively.

Every government that has ever owed this much money relative to its economy has faced the same menu. There are three doors.

The Three Doors

Door one: Cut Spending. The math is brutal. Interest, Social Security, Medicare, Medicaid, and defense consume nearly the whole budget. No politician can propose meaningful cuts to any of those programs and still expect to have a promising career in Washington. That is why the deficit runs near 6% of GDP at full employment.

Door two: Raise Taxes. Same problem, opposite direction. Closing a gap this large would require increases so broad they’d hit everyone for almost every transaction. Possible option only for someone considering political suicide.

Door three: Inflate It Away. This is the quiet door. If wages and prices rise 4% a year while the government pays old lenders 2–3% on the bonds they hold, the debt shrinks in real terms. The debt gets diluted, rather than repaid. Economists have a name for the toolkit that makes this work: financial repression – engineering a world where the return savers receive on government paper stays below inflation. Wealth quietly transfers from the people who hold the debt to the government that owes it.

This isn’t a conspiracy theory. It’s documented American history. From 1942 to 1951, the Federal Reserve openly capped long-term government bond yields at 2.5%, while inflation averaged roughly 6% a year. Bondholders lost about a third of their purchasing power in a decade, and the World War II debt, proportionally even larger than today’s, melted away. It worked.

The Catch Today: It Can’t Be Announced

In 1942, the government’s lenders were patriotic households buying war bonds. Today they’re hedge funds, foreign central banks, and pension managers with a sell button for the moment any asset stops performing. The moment lenders see repression coming, they revolt. They demand higher yields to compensate, or stop showing up to auctions. Rates spike, the interest bill explodes, and the plan defeats itself. The United Kingdom got a taste in 2022, when a single poorly received budget sent bond yields vertical within days and toppled a prime minister in lesser time than it took for a head of lettuce to rot.

So door three has a strange requirement built in: it only works if it never looks like a decision.

There is no memo, no meeting, and more importantly, there doesn’t need to be one. A government carrying $39 trillion of debt has an enormous, structural incentive for three things to be true: for measured inflation to read low, for someone to reliably absorb its borrowing, and for its interest costs to stay below the economy’s growth rate. When an incentive that large persists for years, institutions drift toward it the way water drifts downhill, each agency making individually defensible choices that happen to lean the same direction.

What follows is an inventory of the drift.

Piece One: The Thermometer

The claim, in plain words

Imagine your heating bill is through the roof and the only thermometer in the building is owned by your landlord. Your bill depends on the thermometer’s reading. Your landlord might never once fake a reading, but you’d still want someone else checking the thermometer, especially if he owed money on the boiler.

The US government is in exactly this position with inflation. It owes $39 trillion, and how much that debt really costs depends on the inflation rate, which the government itself measures, through its own statistical agencies. If official inflation reads 3% while true inflation runs 4%, the government quietly wins twice: its inflation-linked obligations cost less, and the public believes its money is holding value better than it is. The claim of this section is not that anyone is faking the numbers. It’s that the agencies holding the thermometer are being weakened, defunded, and re-reviewed. This is happening all at once and all recently, by the one institution with a trillion-dollar-a-year interest in a lower reading.

The plumbing

All of this happened within roughly one year, on the public record:

  • Kevin Warsh launched five task forces, including one on the data sources the Fed trusts and one on its inflation framework, even while publicly promising the 2% target hasn’t changed.
  • The Bureau of Economic Analysis (BEA) is rewriting how the PCE price index (the Fed’s preferred inflation gauge) is calculated this fall, retroactively revising past years’ numbers as well. Goldman Sachs and JPMorgan estimate the new method makes recent core inflation read about 0.2% lower. UBS was blunter, saying the choice of revisions looks designed to suppress inflation and warning the new method is hard to verify externally.
  • The commissioner of the Bureau of Labor Statistics (BLS) (the agency that produces the Consumer Price Index) was fired hours after a weak jobs report. The Labor Department’s inspector general then opened a review of the agency.
  • The BLS now collects less price data than before, and the current budget proposal cuts its funding for inflation and jobs data.

Each item has an innocent explanation: budget pressure, methodological housekeeping, a new chairman’s review. Statistical agencies genuinely do revise methods routinely. However, a measurement system doesn’t need to be corrupt to drift friendly. It just needs less capacity to catch what it’s missing, and every “update” gets chosen under new management, by the debtor.

Piece Two: The Outsourced Printing Press

The claim, in plain words

During the crisis years, the Fed printed money and bought government bonds itself. Everyone could see it happening. The Fed’s balance sheet ballooned from under $1 trillion to $9 trillion, and “money printer goes brrr” became a household phrase. It was effective, but embarrassingly visible.

Today the Fed says it has stepped back and “the market” buys the government’s debt. Technically true, but trace where the market’s money comes from, and the trail leads back to the Fed. It’s like a manufacturer that closes its own factory, then quietly supplies every contractor in town with the machines, materials on credit, and a guarantee against losses. Production doesn’t stop, it just stops appearing on headquarters’ books.

The plumbing

  • The parked money went first. During QE, about $2.5 trillion accumulated in a holding pen at the Fed (the reverse repo facility) where money market funds parked cash overnight. As deficits exploded after 2022, those funds drained it almost to zero, moving the cash into the government’s flood of short-term bills. Since it’s empty, every further dollar would land on reserves directly.
  • Then the Fed reversed. By late 2025 the pen was too low to absorb anything, and funding markets strained. In December 2025, the Fed ended quantitative tightening, removed the aggregate limit on its Standing Repo Facility (the backstop that finances bond dealers is now uncapped) and resumed buying Treasury bills. The same sequence played out in September 2019. Twice the system was asked to carry the debt without Fed support and both times it couldn’t. The support returned within weeks.
  • The buyers who remain stand on Fed scaffolding. Who replaced the Fed, per Treasury’s own advisory committee: money funds, investment funds, banks, broker-dealers. It’s a relay, and the central bank sits somewhere in every runner’s funding chain. Banks hold about $1.8 trillion, some 6% of the publicly held debt, with no one group quietly buying everything. The money funds spent the parked cash. The dealers finance inventory in repo markets the Fed now backstops without limit. Hedge funds running the “basis trade” borrow from those dealers, at a record $830 billion by September 2025, or 3.5% of privately held Treasuries (a concentration risk sized by the Fed’s own researchers and flagged by the Bank of England).
  • Then the scaffolding was widened by rule. In April 2026 the enhanced Supplementary Leverage Ratio (eSLR), a capital rule for large US banks was rewritten, opening room (by one congressional estimate) for up to $2 trillion more in Treasury holdings.

No secret buyer, no slush fund. Just a chain of private buyers funded by the Fed’s old money, financed through the Fed’s facilities, and governed by rules freshly loosened in their favor. The Fed’s balance sheet no longer measures the support. The support moved into the plumbing, where no headline goes.

Now that’s all well and good, but isn’t this just financial jugglery? Doesn’t there also need to be some new, real demand for the bonds, someone that will buy regardless? Well, I’m glad you asked.

Piece Three: The Deputized Buyers

The claim, in plain words

Imagine Congress passed a law saying every gift card sold in America, every Starbucks card, every Amazon balance, had to be backed, dollar for dollar, by government IOUs. Then imagine the country’s biggest retailers formed a club to sell a shared gift card, with a twist: the government interest earned on the backing would be split among the retailers, so every store now profits in proportion to how many customer dollars it moves onto the card. Nobody is forced to buy one. But the entire retail industry is suddenly a commissioned sales force for government debt.

That is nearly exactly what has happened, except the “gift card” is a new kind of digital dollar called a stablecoin, and the retailers are the most powerful payment companies on earth. The claim of this section is that the government wrote the law that makes stablecoins a captive funder of the Treasury, and corporate America then built the distribution machine, not out of patriotism, but because the law made it profitable.

The plumbing

  • The law came first. In 2025, Congress passed the GENIUS Act: every stablecoin must be backed, dollar for dollar, by cash and short-term US government debt. The existing stablecoin giants already hold roughly $250 billion of Treasuries, more than most countries.
  • Then came the sales force. In June 2026, more than 140 companies (Visa, Mastercard, American Express, Stripe, BlackRock, Coinbase, major banks) announced their own stablecoin, Open USD, launching later this year, with a twist: the interest earned on its Treasury reserves is shared among the participating companies. Every checkout page in the consortium now profits from moving customer money into government-debt-backed tokens.
  • An honest caveat: when an American moves money into a stablecoin, the net new demand for Treasuries can be small as their bank or money fund may have held Treasuries already (a substitution point researchers at the Kansas City Fed have made). Domestically, the law works less as a new buyer than as a lock-in: money that enters can, by law, fund only one borrower.
  • The real prize is abroad. For hundreds of millions of people overseas, a stablecoin is the first dollar account they’ve ever had, and their money was never in the US system to begin with. Every foreign dollar that flows in is genuinely new funding for the US government, recruited by app downloads instead of mandates, and filed under “payments innovation.”

How the Fed Can Look Hawkish

Now to address the paradox of this post’s title. The Fed’s hawkishness is real, and it should be stated at full strength: the chairman holds rates high against open political pressure to cut, publicly reaffirms the 2% inflation target, several of his colleagues favor a hike, and the 30-year bond still yields over 5%, comfortably above inflation. None of that is fake. So how can financial repression be assembling at the same time?

That’s because the hawkishness and the assembly happen on different control panels. The chairman’s hawkishness lives on the one panel everyone watches: the policy interest rate. The pieces in this post live on the panels almost nobody watches (bank capital rules, reserve plumbing, Treasury issuance choices, stablecoin statutes, statistical methodology) controlled variously by regulators, the Treasury, Congress, and the Fed’s own back office. Not one of the three pieces above requires a low policy rate. The funding chain behind the auctions, the stablecoin buyer, and the measurement rebuild all run identically whether the funds rate is 1% or 5%.

In fact, the hawkishness positioning is a requirement for credibility of the system. Repression fails the moment bondholders see it coming. Nothing keeps bondholders calm like a chairman who sounds like Paul Volcker. Whether intended as cover or perfectly sincere makes no difference to how it functions. The loud fight on the visible panel buys time and credibility, while the quiet panels are rewired. When the two have collided, we know which one wins, in 2019 and again in 2025, rate discipline ran into the plumbing, and the plumbing won within weeks.

An important point to note: The repression is an architecture, not yet an outcome. The hawkish policy rate isn’t evidence against the architecture. It’s the front of the house.

Quiet Where It Collects, Loud Where It’s Priced

If the pieces fall into place this quietly, would anyone ever see it happen? Yes, but not where most people look. The tax is collected silently, from savers, in purchasing power, statement by statement. But the evidence is priced loudly, in the two markets no government controls. The 1940s original managed to be quiet in both places. This one can only be quiet in the first.

Here’s why. The 1946–1951 version ran silently for a decade because every exit was sealed: capital controls everywhere, currencies pegged to a dollar pegged to $35 gold, and creditors who were mostly America’s own citizens with nowhere else to go. Repression in a locked room is quiet and total.

Today’s room has doors, and the best-informed creditors on earth are already using them. Since 2022, the world’s central banks have bought about 1,000 tonnes of gold a year, three years running (China for twenty straight months, Poland among the decade’s biggest buyers, India and Central Asia alongside to name a few). The World Gold Council (WGC) reported that in reserve manager surveys, 95% expect official gold holdings to keep rising while 74% expect the dollar’s share of reserves to fall. Central banks now collectively hold more gold (27% of reserves) than US Treasuries (22%), the first time in the modern era. You cannot financially repress a foreign central bank.

So expect a hybrid: 1940s at home, 1971 abroad. Inside the dollar system, the quiet melt proceeds, and stays quiet. Outside it, the free actors force the adjustment into the open – a dollar that gradually buys less of the world’s neutral assets, a long-bond yield that stays high as the exit toll, and a gold price that does what it couldn’t do in 1946: move. The saver’s statement will never announce the tax. The two numbers that announce it anyway are the long bond (the toll charged by those who can leave) and gold.

One side effect worth expecting: a melt-up. When the measuring stick shrinks, everything measured with it reads higher, so a repression era is rarely a falling stock market. More often, it’s record nominal highs that quietly buy less. Which assets melt up depends on the long bond. In the 1970s, with long rates free to rise, stocks went sideways for sixteen years in nominal terms (down 70% in real terms) while gold rose 24-fold. In the 1940s, with long rates capped, stocks inflated along with everything else. Today’s configuration is the 1970s one, until the final piece (the temporary-that-gets-extended long-bond purchase program) clicks in. Either way, rising prices won’t mean the plan is failing. They’ll mean the unit is shrinking, and the way to check is to price the market in the one money no one can print. Measured in gold, the melt-up disappears.

Could this all be wrong? Yes, and here’s what wrong would look like: real yields staying positive across the whole curve including bills, the debt’s maturity lengthening, the Fed restarting its shrinkage and holding its nerve when rates strain (the test it failed in 2019 and 2025), stablecoins diversifying out of Treasuries, central banks choosing Treasuries over gold again. Every one is observable.

Until then, the claim stands on math alone: $39 trillion, three doors, two locked, and every institution that would make the third door expensive has gotten measurably weaker, quieter, or friendlier within the same few years. Maybe that’s drift. Maybe it’s design. From inside a savings account, the tax collects the same either way, no matter how hawkish the speeches sound.

This post argues an interpretation, not an accusation. This is NOT investment advice, and the author is in no way a financial or investment advisor. The factual claims (the debt and interest figures, the reverse-repo facility’s drawdown from $2.5 trillion to near zero, the December 2025 end of quantitative tightening and the uncapping of the Standing Repo Facility, banks’ ~$1.8 trillion of Treasury holdings, the leverage-ratio rewrite effective April 2026, the BLS firing, budget cuts and inspector-general review, the BEA’s retroactive PCE revision and bank estimates of its direction, the Fed’s five task forces, the GENIUS Act’s reserve requirements, the Open USD consortium’s announced yield-sharing structure, the 30-year’s 27 days above 5%, the central-bank gold purchase data, and the ECB’s reserve-composition estimates) are drawn from public records, official statistics, and mainstream reporting. The interpretation (that these amount to financial repression assembling itself) is contestable, and the section above lists exactly what evidence would refute it.

Guide to Angel Investing

Posted on by

I’ve been dabbling in startup investing for a few years now. Nothing extraordinary, and I’ve definitely not made it a career, but I have developed a guide of sorts that I follow when making the final decision of whether or not to invest.

This is written from the viewpoint of a single angel investor. Here it is:

  1. Risk: Understand the high risk in angel investing. Mentally, assume you’ll write off all investments as soon as you make them. Invest only what you can afford to lose.
  1. Avoid life-support investing: If a company comes for funding, where the founder says/implies something to the extent of: “fund us or we’re going out of business” – usually gravitate to not investing. Life support investing is a bad idea (not to mention stressful!).

Lean towards companies that will survive regardless of whether or not you invest in them.

  1. Three companies come in for funding:

A – Entrepreneur has the best idea ever! They need funds to build the product.

B – Entrepreneur has the best product ever! They need funds to market.

C – Entrepreneur has a product which shows great traction. They need funds to                    scale.

Everything else equal, investing in C has the most promise.

  1. Follow-up investor: For the new investor, prefer being a follow up investor rather than leading rounds.
  1. Investment hierarchy: First invest in companies with great traction. If that does not exist in the options, then invest in those with great people, followed by those with a great product.
  1. Bet on people rather than product: Early stages, funding is a bet on people rather than product. Products can change more easily than people.
  1. Never invest in something you don’t understand: Doesn’t matter what it is, who is involved, who has invested or how well the market is going.
  1. Entrepreneurial grit matters: Invest in entrepreneurs who show they’ll simply never give up on the company – no matter what. Perseverance is crucial. When things get hard (and in startups they will!), you want someone who has the grit to stick through it to figure a way out.
  1. Scaling: Always think about how the company can scale. If scaling is expensive or hard, lean towards not investing.
  1. Avoid niches: Niche markets are starting points – but that’s it. The ultimate market cannot remain in a niche space. Successful companies need to scale.
  1. Viable business or science project?: Is the company capable of being a business or will it remain a technology/project? Not all technologies can become businesses.
  1. The Fence: If you’re on the fence about a company for a long time – defer the investment.
  1. Entrepreneur’s knowledge: The entrepreneur needs to know more about the space than you. If you seem to know more about his own space – that’s a red flag.
  1. First versions will suck: Products require several iterations of trials and failure to become great. The key is if the entrepreneur/team is unfazed by failure to keep at it despite the obstacles.
  1. What if Google/FB does it?: Assume the incumbent is always going to copy the entrepreneur if he does well. What is his strategy?
  1. Does the product create delight?: Do you feel delighted when you use the entrepreneur’s product or service? If not, at least do the customers absolutely love the product? Defer the investment until that is true.
  1. Simple enough for your mother: Can your mother use the entrepreneur’s product or service? Simplicity is key for widespread adoption.
  1. Failure is a learning process: Past failures > No failures.
  1. Return time period: Expect returns no less than 5-7 years.
  1. Differentiate between the decision and outcome: Decisions must be taken using all information available to you at that time. You can’t predict the future and hindsight is always 20/20. A bad outcome does not necessarily mean a bad decision and vice versa. Don’t beat yourself up too much over a bad outcome, and then again don’t let a good outcome go to your head.

Good luck! And as always, if you do discover the next Google or Facebook, do let me know! 😉

With that said, here is the obligatory legal disclaimer that goes with any investment advice – Please invest at your own risk! 

“I could have built that in 2 weeks!”

Posted on by

Yes but did you? That’s the point. You’ll often hear an over zealous programmer or engineer exclaim that they could have built [insert hot startup] in no time so what was the fuss about?

The key that they’re missing out on is that it’s simple to clone, but extremely difficult to innovate.

It’s a decision tree. Every node represents a set of decisions available to you with all the possible permutations as below.

decision tree

What is visible to everyone is the path travelled, not the paths forgone. Once you know your destination, tracing the tree back to the root is relatively simple. However, starting from the root without knowledge of your destination is where true innovation happens. That is a process filled with trial, error, failure, and course-correction over and over again till you reach your final destination.

Conceptually speaking there are several startups that are in the same space – most fail but some do better than others, with one becoming a market leader. Path, Instagram, Oink, etc are all relatively similar, but you’ve heard of some and not the others. Why? They’ve all taken different paths in the tree which has enabled some greater success than the others.

First mover advantage is real. Reaching first on the scene with a product that scales gives you a distinct advantage to becoming the market leader no matter who you’re competing against. Take the case for Facebook, a company with near infinite resources and the top destination for the web. They were late to the ephemeral photo space behind Snapchat. Facebook released a Snapchat clone called Slingshot in an attempt to beat them. Despite Facebook’s resources behind Slingshot, which one do you have on your phone? They also released a Flipboard clone with Facebook News (yes they had/have a separate news app), an Instagram clone with Facebook Camera (before acquiring the former), and a Foursquare clone with Facebook places (now defunct) amongst others. They failed in every one of those cases for the exact same reason that they succeeded in out-competing Google+: they were there first.

Switching costs in networks ensure that network effects are always in play. If a user is on a hot app with his friends, he’s not going to switch to a clone that does the same thing and also convince his friends to switch, even if the second one is slightly better. The only way to convince the user to overcome the switching cost is to offer something that is 10x or an  order of magnitude better than what he is using.

Success always looks easy from a distance. That’s because it’s only the path travelled that is visible, not the entirety of the tree. The next time someone says “I could have built that in 2 weeks” simply ask “then why didn’t you get there first?”.

 

 

 

Solving a user problem vs a technical problem: the difference between creating value and wasting your time

Posted on by

What is the kind of problem you are working tirelessly to solve? Are you actually working to create value or simply wasting your time? The difference is in the end goal of what solving the problem accomplishes. Some very big user problems have fairly simple technical solutions. They are easy to use, solve an actual need and fairly simple to interact with. On the other hand, some very big technical problems really don’t end up doing much for the end user. The former gives you something that people want, while the latter gives you an extremely beautiful piece of crap.

Engineers are particularly susceptible to this. It is natural to fall into the trap of believing that just because something is challenging from a technical perspective, it must be a valuable problem to solve. From a user’s perspective, all that’s important is that the product does what it is meant to. “How” really doesn’t matter. If you are selling mousetraps, does it get rid of the mice better than everyone else? If yes, that’s great. Users aren’t going to care much of what happens within the product as long as it works. You can spend all your time trying to create a very advanced, state of the art, nuclear powered mousetrap that creates mini fission explosions to obliterate the mice, but at the end if it doesn’t get rid of the mice, is unwieldy or uneconomical, no one’s buying it. It’ll be an extremely cool project no doubt, but at the end of the day, a fairly useless one!

Every problem that you solve should tie back to the user and/or to enhance their end experience. Be careful of not getting stuck in your own technical world that is insulated from the user completely. Though it does vary by industry, for consumer facing products more specifically – it is the end user who is king. It is easy to get lost in day to day technical, legal, product, financial, investment problems, but if what you’re working on doesn’t help the end user – you’re not moving forward.

Expressed in slightly broader terms, whenever you look at a hard problem, you need to assess its value independently to determine whether it is truly worth your effort. Becoming the top restaurant in the city: Is it hard? Of course! But how valuable is it given that the next 10 best restaurants are almost always going to be head to head with you?

Passionate engineers, by nature, get excited by challenging technical problems. However, take care that you never lose sight of the user so you don’t end up spending all your effort on creating something which doesn’t end up doing anything.

On the flipside, thinking from a user’s perspective brings focus, where the complicated tasks that you were dreading, often no longer seem to be relevant or worth pursuing. It simplifies things! Not everything is wrong with the world… 🙂

Overthrowing the 600lb Gorilla: Create a Platform Shift

Posted on by

As the little guy, albeit with some pretty large aspirations, how do you exactly go about taking on the 600 lb gorilla that has billions of dollars in resources and tens of thousands of employees? The gorilla has been in the industry for ever, knows all the other chimps in the space and is the undisputed leader. So, how do you, with limited resources, almost no name and a tiny team expect to, well, overthrow this big guy? You did have large aspirations, remember?

Assume the incumbent is in the industry of creating horse-carts. He knows everything there is to know about horse carts – which breed  works best for what purpose, what  food the horses should eat, how comfortable the cart needs to be and so on. You can’t compete with this guy on horse-carts. He has tens of thousands of horses for every imaginable type of cart and he is killing it.

So, how do you beat him at his own game? Well, you change the game! If the entire industry is making horse carts, you do the unthinkable and get rid of the horse! Bring in a motor car.

Initially you will be scoffed at – “A man derives his worth from his horse”, “It’s so unnatural”, “Instead of a horse, you are creating a bunch of explosions 2 feet away in your fancy-schmancy car? Did you hit your head somewhere?”

You are crazy! Why would anyone want to copy a crazy person? So, while you are being scoffed at, you get ample time to solidify your base and perfect your product. You debut your invention and start selling to the early adopters. People are coming around to the idea – “The engine did not explode on anyone yet, right?” You start taking off – slowly, but surely.

Meanwhile, the 600lb gorilla is still selling billions of dollars’ worth of horse carts, although growth is taking a hit. Maybe it’s a one-time thing? Maybe they need to innovate more. The CEO calls for increased investments in R&D. The department gets to work and finds a substantial 5% increase in the horse’s energy output if they feed it a new combination of grain!

You release version 2 of your motor car – which is even better! People are warming up more to your invention – “You don’t need to take care of such a large animal anymore”, “My friend’s neighbor has one of those motor-things”

At Gorilla, Inc. there is a crisis. Sales are 80% of what they were. The CEO thinks they need to start looking into this motor car thing after all. He pulls up a team from across his company of the best horse-trainers and the best cart-makers to go and figure out this motor thing.

The Gorilla debuts its own motor car. Unfortunately, it doesn’t run too well. Turns out, the engine doesn’t eat a liquid diet of horse grain.

One extremely forward thinking employee in Gorilla, Inc. goes up to the CEO. “Horse-carts are old news. We need to stop making them and put everything we have into perfecting our motor car.” The CEO isn’t too happy – “You’re telling me that we should abandon a product line that accounts for 70% of our revenues over something that could just be a fad? Have you completely lost your mind? Who hired you?” The CEO fires the employee – “We can’t have people making such ridiculous comments in time of crisis here.”

Even though you’re doing well, the entire process has taken its toll on you. Either you’ve greyed all your hair or maybe lost most of it but there is a lot more ground to be covered. Now, there are a lot more companies now trying to create their own motor cars, though you clearly have the most advanced one – for now.

You’ve put the 600lb Gorilla on a diet. Play your cards right and you may end up taking it down after all.

—

If you study the rate of development in different industries since 1980, you’ll see that it is only the technology industry that has made significant progress. The rate of innovation in medicine and pharmaceutical industries is the same; commercial aviation has actually regressed and average air-speeds have slowed down since the Concorde got decommissioned; finance has become massive, though it works on the same principles as before; there hasn’t been a significant platform shift in manufacturing; and classroom instruction is pretty much the same since the 1800’s.

Most of these industries use newer technology as tools to do the same things that they keep doing, slightly better. It is the equivalent of finding a 5% increase in output by feeding the horse better grain. Industries need to be re-thought from the ground up for you to discover a motor car in there.

Now, technology is progressing to the “internet of things” – you know, when your front door pings you on your phone asking if it should lock itself since you forgot to do it. Mobile internet and cloud storage are already in place, while robotics and artificial intelligence are making a lot of progress. Technology will soon permeate the very essence of what you interact with – covering pretty much every industry.

So, while the incumbents use technology to make their horses slightly better, when will you debut your motor car?  

For the US to remain competitive, it needs to outsource *more* jobs to China & India

Posted on by

Globalization has made the world a smaller place. A consequence has been the shift of jobs, especially in manufacturing, to China and India. Contrary to popular belief, the fall of US manufacturing is to the benefit of the US by making the industry more competitive than before.

Lets take a bottoms-up approach. With respect to American manufacturing, the way it works is:

I am a US corporation. My costs are ridiculously high and profits are very low.
I discover I can reduce my costs by 40-70% if I move manufacturing to China. So, I do that (lets hold the fact that I need to layoff a bunch of people in US for now and come back to that).
As a company, that increases my profits which enables me to re-invest it and grow. Being a US corporation, US government sees more cash by taxing my higher profits, which it can use to build society, healthcare, infrastructure etc. and do other public good.

If all American companies do this, then US industry in general grows, becoming more competitive – creating more jobs everywhere – both in the US and China. To support growth, I will need more skilled employees in the US and more unskilled employees in China because that’s what their strengths are at. It is basically trading on comparative advantage making everyone better off. This helps everyone by creating more jobs overall. I would argue this means that more jobs get created by more financially stable companies than before, compensating for the original layoffs multi-fold.

So, as the US Govt., I would not focus on manufacturing, since that needs more unskilled labor and China can do it much cheaper comparatively. So, I may as well send all my manufacturing there and focus my resources on maintaining and building the skilled labor advantage – whether that means increasing access to education, lowering costs to education, attracting the world’s smartest people to work in America or encouraging entrepreneurship in general. This way, I can export my services and innovation that comes out of that. So, I can import my kid’s toy train from China, but export new drugs or even, say Microsoft Office out (stuff that can only be built using US’s core advantage of skilled labor). On the other hand, since US is and will continue to be a hub for skilled labor, most of the world’s innovative and profitable companies (like Apple) will be of US origin.

However,a caveat is in the manufacturing of complex machines. Complex machines require more skilled labor for quality and if people are willing to pay more for a better quality product, then that is a clear market opportunity. So, that is a case for the resurgence of US manufacturing, where skilled labor is needed for quality goods. But, unskilled labor nonetheless should be exported out.

It is really hard to be great at everything. The human race was able to reach today by specialization in fields and then trading. The same argument here. US specializes in skilled labor and leave the unskilled part to someone else.

The reality behind “We got 70k users in 2 days of launch”

Posted on by
For any startup, gaining traction depends on two things: 1) How many people are you bringing and 2) How many users getting engaged. The first one has a lot to do with marketing, while the second one has everything to do with the product.

How many people are you bringing?
Companies launch not over one day, but over a process that can last almost an entire month or maybe more. Phrases like “We got 70k users in 2 days of launch” often have a back-story behind them.

1) Instagram – For instagram, they had Jack Dorsey (founder of Twitter and about 2m followers) back them. So, when they launched, it was fairly easy for them to get their initial users since they needed to get people like Dorsey reach out to their network and the press picked it up pretty quickly.

2) LinkedIn – I was speaking to one of the co-founders of LinkedIn, who was VP of Marketing. He said that initially most of their users were PR based because in the early days, the product was not too viral. LinkedIn took its time to grow compared to other social networks.

3) SkyFire (funding: $41m to date) – Speaking to one of the founders, they launched when the iPhone had just launched and so were a very press friendly story. They also had closed their Series A which enabled them to invest aggressively in PR. So, a lot of their users were PR driven and then word of mouth referrals.

This is pretty interesting since mostly when you hear stories like “we had 70,000 users in 2 days of launch” – a lot of that is paid for or has significant help. It is very rarely viral just from the product.

The effectiveness of PR campaigns is often questioned since a lot of it depends on content, relevance and visibility. However, besides PR, there are other avenues like SEM, SEO, Social etc. which you can tap into for your initial influx of users.

How many people are getting engaged?
A key point is that after the initial influx of users is established, it is up to you to create a wonderful user experience with your product – create something that is useful and usable to have them stay and give referrals. Instagram grew virally after its initial influx because its users simply loved the product. Similarly, users gained significant value from having a LinkedIn profile, which enabled the network to grow further.

For this reason, before you invest in marketing, you need to keep gathering data, testing and failing in front of a smaller network, probably your own Facebook network till you figure out your product actually works. The product almost will never go viral if you only invite all your Facebook friends, though you will get some very solid data from them using it. Doing this will ensure that when you finally pay for the initial influx of users to come and “launch to the world”, you are better prepared.

Good Luck!

Unreasonable Obsession for the Startup Grind

Posted on by

When getting a co-founder, employee, investor or any other stakeholder for your startup, one of the most important metrics you want to scan for is belief in the startup over and above any monetary compensation. Simon Sinek says it beautifully – “If you hire people just because they can do a job, they’ll work for your money. But if you hire people who believe what you believe, they’ll work for you with blood and sweat and tears.”

In any startup – things will go wrong, stuff will break, investors will pass, your product will suck, employees will quit, hiring will be a problem, deadlines will be pushed, users will not be engaged, partners will bail – but you and your team still need to keep pushing through.

You will face rejection over and over and over again – from investors, users, and pretty much everyone you talk to. You will hear people give several reasons why the startup will not work – but you and your team still need to keep pushing through.

Ben Silbermann, CEO of Pinterest said, “There are lots of ways for investors to say no to you, and I’m pretty sure I’ve heard every single one”

Elon Musk compared running a startup as “eating glass” and “staring into an abyss of death.“

The TechCrunch stories of “we got 100k users in 2 days” – all have a long grind where they failed over and over again before their overnight successes.

Instagram slogged through Burbn, Twitter slogged through Odeo, Foursquare slogged through a failed acquisition by Google, Rovio slogged through 51 games prior to Angry Birds, Starcraft was almost abandoned, SpaceX blew through $120m almost bankrupting Elon Musk with 3 failed rockets – but they all kept pushing through.

When things go bad, and they always do, you need people who work with you with “blood, sweat and tears.” Startups are never easy and only unreasonable obsession can power you through the grind. Make sure you have it.

What Makes Silicon Valley Work?

Posted on by

Silicon Valley or the Bay Area has a completely different “vibe” for entrepreneurship. It is the epicenter of the world’s technology innovation, where 40% of all US venture capital is invested and where 90% of the highest venture returns occur. Why?

Why Silicon Valley and not New York which is filled with Ivy alumni; or Boston, home of Harvard & MIT; or even somewhere in UK or Europe or Asia? All of them have amazing engineering and business schools filled with very smart people, so why is the Bay Area the epicenter of innovation?

It comes down to the culture. The culture in Silicon Valley is completely different – where it encourages entrepreneurship to another level.  So, what is so unique about Silicon Valley culture?

1) It is OK to Fail – Yeah, you read that right. Embodying the culture, at Stanford, all our entrepreneurship professors have continually drilled into our heads – It is OK to fail! Failure is a learning process. If you start a company and it fails, then start another one. Fail fast and fail often. Go out and take risks. If you do not fail, you are not taking enough risk. Don’t be afraid to fail. Why?

Innovation by definition has a high failure rate. So, if you are afraid of the downside, you will also miss out on the tremendous possible upside. Even if you do fail, the amount you learn would be worth the cost of failure since you will not make the same mistakes again. In Silicon Valley, starting a failed company counts as experience! 

This gives Silicon Valley a very risk-seeking culture.

In India or China, if you start a company and it fails, you are considered an idiot and probably need to move to another city!

2) Trust the Young – The community trusts the young  with their ideas and does not dismiss them because of their age. The VCs on Sand Hill Road follow suit. This has given rise to entrepreneurs who have raised millions of dollars, while still in their late teens or early twenties.

This is in contrast to industries with entry level jobs, where you start at the bottom and need to be “trained” to gain experience.

The thought process is – when a new gadget comes out, who is more likely to play around with it and figure it out? A teen or his dad? The current generation of teens have been playing around with technology almost since they were learning how to walk. So, when it comes to companies revolving around technology, the young often know what they are doing and should not be dismissed.

3) Career path out of college –  Coming from undergrad at Northwestern, success there (similar to other schools in the East Coast) was more defined as landing a job in consulting or banking. Even though there was a large entrepreneurial push, the concept of becoming an entrepreneur right out of college was relatively rare.

At Stanford, entrepreneurship is more mainstream, where working on your startup after graduating is a common path apart from traditional consulting or banking jobs.  Though most of  those start-ups don’t succeed, that kind of entrepreneurial environment encourages the next Google or Sun or Cisco or Tesla to get built.

—

This mindset of Silicon Valley will also seem pretty ridiculous to anyone outside the Bay Area. Imagine talking to a banker on Wall Street about it being okay to fail and trusting the young. If he does believe you, he will get himself fired.

This is also why, to learn more about why the Valley works, you will need to visit and see the culture for yourself.

Early Valuations are Bets on People

Posted on by

Early stage valuations of start-ups are just bets on people – the underlying assumption being that a good team can figure out a product-market fit more easily.

In the early days, when the product is not concrete,  the initial idea can pivot, morph or get completely overhauled within a span of as little as 3-4 months. The initial team is the most crucial element of this, since the composition of the initial team will determine what direction the product takes. Swap the initial founder out with someone else and you will have a completely different product.

When it comes to markets – they should be large/have the potential to be large or you are not in the right space. If they are large, expect them to be crowded with people still trying to “crack it.” Within this large market, your product needs to “fit” into a specific area. An example is how Twitter and Facebook “fit” into the market. Facebook is more for a personal network, while Twitter is more of an information broadcasting service. Though both are in the social networking space, they are still fundamentally very different in how they “fit” within the same market.

Figuring out how your product fits into a market is called achieving product-market fit (… duh!) – arguably the only thing that early stage start-ups need to focus on. This means that initial teams need to iterate, learn, unlearn and relearn, where the initial product basically embodies all the skills, knowledge and experience of the founders to successfully fit into the market.

In the early stages for investors, when both the product and the market are not clear, it just boils down to the founders. Have they done this before? How do they think? Can they build it? Will it get traction? Will users keep coming back? Will it make money? Is it a sustainable business? and countless other issues/risks.

So, this means that if you are a first-time entrepreneur, you often need to go further before you can raise any/much capital. You will probably need to show a working product with some traction before a venture round, whereas someone who’s done it successfully several times before probably just needs to make a phone call with an idea.