How Tokenization Will Rewrite Monetary Economics
Every time long-term yields make new highs for the first time in X years, the “it’s going to end badly” crowd comes crawling out of their bear caves to begin another round of I told you so. Honestly, after years of watching this, it has become comedy to me. I know, I know: the economics textbooks, the debt math, the historical precedents, the inevitability of the endgame, the Fourth Turning, blah, blah, blah. This scary story has been wrong so many times that the bears have to use that forbidden financial phrase for debt, but this time I swear it is different. Using a sitcom reference from my youth, Maybe this is the Fred Sanford moment and this time it really is “the big one!”
In August 2011, the U.S. suffered its first-ever sovereign credit downgrade. A year later, Europe was supposedly approaching its inevitable endgame until Mario Draghi stood up in July 2012 and said the ECB would do “whatever it takes.” Before that it was Japan. In the last decade we had China’s mathematically impossible debt situation. More recently, the bears briefly emerged for COVID, the violent Fed tightening cycle of 2022, Silicon Valley Bank and then tariffs, when suddenly everyone dusted off the Smoot-Hawley and the Great Depression books. Now it is the debt supporting AI infrastructure. Somewhere in between all of these endgame events, the global sovereign debt binge somehow coexisted in the 2010s with trillions of dollars of negative-yielding debt, another outcome that was once supposed to be impossible. I was told in my youth at Morgan Stanley that the bond guys were supposed to be the smart ones.
The problems were real every time, and the problems are real today. Debt matters, deficits matter and higher interest expense matters. They matter and they need a solution to reset it but since Japan began this, the solution has been building slowly in the form of the digital economy. The recurring mistake comes from assuming that arithmetic that looks unsustainable using today’s variables, today’s technology and today’s financial architecture makes the outcome predetermined while fighting exponential innovation. The bears come out of their caves, announce that this time the math has finally won, explain why policymakers and markets have run out of options and wait for the inevitable endgame. The system then adapts, variables change, technology creates capabilities that were absent from the original model and capital finds new ways to move through the system. Eventually, the bears retreat into their caves until the next fight gives them another opportunity to announce that this time really is different. For those looking to possibly adapt away from the inevitable endgame trap, I highly recommend reading a book by Daniel Pink which came out just before the Great Financial Crisis and I reference often for its impact on me on this situation, A Whole New Mind: Why Right-Brainers Will Rule the Future.
That is why I can look at today’s debt reality and be far more comfortable than the endgame crowd. I went to Silicon Valley in 2013 because I wanted to understand the other side of the equation: what happens when the denominator changes? I would rather sit on my exponential-innovation island than assume tomorrow has to look like yesterday when we are solving math equations from before Einstein, listing companies planning to go to Mars and supply the world with humanoids. That matters more today than at any point since that trip. Exponential innovation is taking us toward ubiquitous intelligence, while tokenization and crypto rails are beginning to connect that intelligence to an enormous existing stock of global wealth. Technology is preparing to dramatically reduce the friction between owning wealth and accessing liquidity, while AI agents can continuously manage that wealth at machine speed. The biggest macro story ahead may therefore involve something very different from governments printing unimaginable amounts of money to deal with their debt. Technology may make trillions of dollars of wealth that already exists far more liquid, dramatically increasing its velocity through the economy. That possibility sits outside most historical endgame models, and it is where this paper begins.
The $195 Trillion M2 Was Never Designed to See
The United States has crossed $40 trillion of federal debt, and that number has understandably become an obsession in macroeconomics. Every additional trillion raises familiar questions about who will buy the bonds, how high yields will have to go, whether interest expense will become unsustainable and whether the Federal Reserve will eventually have to monetize the debt. Almost all of that discussion begins with the liability side of the national balance sheet. The other side deserves considerably more attention. According to the Federal Reserve’s latest Financial Accounts, U.S. households and nonprofit organizations ended the second quarter of 2026 with $195.9 trillion of net worth, supported by approximately $217.8 trillion of gross assets. Against roughly $40 trillion of federal debt therefore sits almost five times as much household and nonprofit net worth.
Something potentially revolutionary is beginning to happen to that wealth. Technology is making it easier to mobilize. For decades, economists have focused on how much money governments and banking systems might eventually need to create to support expanding nominal economies and growing liabilities. Tokenization introduces another source of potential liquidity by reducing the friction separating the enormous existing stock of wealth from the financial system in which money moves. The implications reach well beyond crypto. They touch money supply, monetary velocity, collateral, financial conditions and ultimately the way economists think about the relationship between national liabilities and the assets sitting on the other side of the balance sheet.
The economics textbooks give us a familiar framework for how money works its way through an economy. Money and credit create liquidity and influence financial conditions; easier financial conditions encourage spending and investment; businesses respond with greater capital expenditures and hiring; hiring generates income; income supports consumption; and together those activities drive nominal GDP. The process contains feedback loops at every stage, but the basic framework begins with the availability of money and credit as fuel for expanding economic activity. That is why economists spend so much time watching M2, bank lending, central-bank balance sheets and interest rates. If an economy requires substantially more nominal activity, the traditional framework assumes that money, credit or velocity must expand sufficiently to support it. Tokenization raises a new question: what happens when technology dramatically increases the amount of economic activity that can be supported by the money and wealth already in existence?
What rarely appears explicitly in that framework is time. Every stage of the traditional transmission mechanism was built around an economy filled with human and institutional friction. Humans sleep, markets close, banks close, payments wait for business days, trades wait for settlement, collateral has to be moved, approvals have to be obtained and capital can remain idle while the next transaction waits to occur. A dollar might ultimately move from financing to investment to wages to consumption, but every transition consumes time. The economics textbooks were written around a financial system operating on human schedules, with a limited number of productive and transactional hours available during a 365-day year.
AI agents and tokenization begin changing that constraint. Agents can operate 24 hours a day, seven days a week, 365 days a year, and tokenized financial markets are evolving toward the same continuous operating schedule. An agent does not have to wait until Monday morning to evaluate a balance sheet, identify available collateral, execute a transaction or redeploy capital. As more assets migrate onto always-on rails, the number of economically active hours in a year increases while the friction separating transactions declines. The arithmetic is straightforward: more operating hours combined with less friction creates greater potential economic throughput from the same stock of capital. The traditional economic model focuses heavily on the quantity of money available to the system; the machine-time economy forces us to think much more seriously about how frequently that money and the wealth supporting it can be put to work.
This creates an unusual mismatch between the two sides of the balance sheet. The debt was accumulated on human time, while the economy that services that debt is beginning to move toward machine time. The nominal value of the liability does not change because an AI agent works overnight, but the productive capacity, transaction velocity and capital efficiency of the system supporting that liability can change dramatically. Capital that can be evaluated, collateralized, exchanged and redeployed continuously has the potential to participate in more economic activity during the same calendar year. That is why the transition from human time to machine time belongs directly in the discussion of debt sustainability: the numerator has largely been inherited from the old economy while the denominator is increasingly being generated by a faster one.
The traditional monetary identity provides the simplest way to think about the issue: M × V = P × Y. Money multiplied by velocity equals nominal economic activity. Most monetary debates concentrate heavily on M, particularly during periods of fiscal stress when investors begin worrying about monetization. Tokenization may have its most profound effect through V while simultaneously making the traditional boundary around M less informative. An asset that can be valued continuously, divided into fractional interests, pledged as collateral, settled instantly and converted into a payment instrument by an autonomous agent has very different liquidity characteristics from the same asset sitting behind several days of paperwork, settlement and intermediary approval. Its ability to participate in the financial system has changed, and machine time allows that participation to occur more frequently.
The Definition of Money Is Already Changing
Less than three weeks ago, on September 4, 2026, the Federal Reserve published a note titled New Forms of Money and the U.S. Monetary Aggregates. Its opening definition deserves far more attention than it received: money is a group of safe assets with stable values that households and businesses can use to make payments or hold as short-term investments. The Fed goes on to explain that monetary aggregates classify assets according to their liquidity and function, particularly whether they operate primarily as a medium of exchange or a store of value. That definition becomes increasingly important as technology changes the functionality of assets. Monetary classifications have always reflected the financial architecture of their time, and tokenization is beginning to change that architecture.
The Fed specifically examined tokenized bank deposits, tokenized money-market funds and payment stablecoins. Tokenized deposits already fall within existing monetary aggregates. Tokenized money-market funds currently resemble savings instruments, although their expanding use within decentralized finance could increasingly give them transactional characteristics. Stablecoins create an even more interesting classification question because their eventual treatment within M1 or non-M1 M2 could depend upon how they are actually used. The institution responsible for measuring America’s money supply is therefore already confronting a basic reality of the tokenized economy: technology can change an asset’s economic function enough to challenge the boundaries economists use to define money.
There are important limits to how far that idea can be extended. Money derives its special status from safety, stable value, broad acceptability and immediate availability. A tokenized Treasury fund still operates through a fund structure and redemption mechanism. A corporate bond still carries credit risk. An equity can still decline sharply when liquidity is most valuable. Transfer speed and economic liquidity are related concepts with very different meanings, particularly during periods of financial stress. Tokenization improves the plumbing connecting assets to the financial system while the quality of the underlying asset continues to determine how much liquidity the system will provide against it. This distinction becomes central once tokenization moves beyond deposits, stablecoins and money-market funds and begins reaching the much larger universe of household and institutional wealth.
When $195 Trillion Becomes More Mobile
Consider what makes up household wealth: homes, equities, bonds, private businesses, retirement assets, investment funds and other financial and real assets. Economics has historically treated these assets very differently from money because accessing their value requires friction. A homeowner with a $2 million house and $20,000 in a checking account may have substantial net worth while maintaining relatively little immediate purchasing power. Accessing $100,000 from the house generally requires a sale, refinancing, home-equity loan or another borrowing arrangement involving documentation, underwriting, intermediaries, fees and time. The economic distinction between owning wealth and using wealth has therefore been substantial, and our monetary framework developed around that reality.
Now imagine that same house inside a mature tokenized financial architecture. Ownership can be digitally verified, economic interests can be fractionalized, valuations can update continuously and regulated lenders can potentially lend against the property through programmable infrastructure. A portion of its value could be pledged without requiring the entire asset to be sold, while an AI agent managing the household balance sheet could continuously determine the cheapest source of liquidity available to meet an obligation. The homeowner receives a $10,000 bill, and the agent evaluates cash, securities, credit facilities and tokenized collateral, accesses the optimal source, converts the required amount into the appropriate payment instrument and settles the obligation. The economic distance between the $2 million house and the $10,000 payment has suddenly become much shorter.
Collateral economics will determine how powerful this mechanism becomes. A lender will continue to require a loan-to-value cushion because asset prices fluctuate, markets can become illiquid and legal claims require enforceability. A Treasury might support borrowing at a relatively small haircut, while equities, private credit and real estate would support progressively lower advance rates depending on their volatility and market depth. Tokenization can nevertheless improve several variables that influence those haircuts by making ownership clearer, valuations more continuous, collateral monitoring easier and settlement faster. Even modest improvements in the financing capacity of a $195 trillion household balance sheet can produce very large changes in potential liquidity.
This is where scale becomes important. The thesis does not require $195 trillion of household net worth to become M2. It requires some portion of an enormous existing pool of wealth to become easier to borrow against, pledge, trade, fractionalize or convert. If technology makes even an incremental percentage of that balance sheet more economically mobile, the potential liquidity effect becomes measured in trillions of dollars. The same process extends beyond households to corporations, institutions and eventually global assets. A financial system in which a much broader universe of wealth can continuously participate as collateral requires less idle transactional money to support a given amount of economic activity.
From Money Supply to Liquidity Supply
This leads to a distinction that may become increasingly important for economists: money supply and liquidity supply are different measurements of economic capacity. M2 measures a defined collection of monetary assets, while effective liquidity describes the purchasing power that economic actors can actually mobilize. Historically, the two have been sufficiently connected for M2 to provide useful information about monetary conditions. Tokenization can weaken that relationship by allowing assets outside M2 to support transactions more efficiently through collateralization, borrowing and rapid conversion. Two economies with identical M2 could therefore possess very different levels of effective liquidity if one can mobilize its broader stock of wealth far more efficiently than the other.
The implications for velocity could be equally important. Corporations hold cash partly because future obligations are uncertain and converting other assets into money involves friction. A company that historically required $1 billion of immediately accessible cash might eventually operate with a smaller cash buffer if the remainder of its balance sheet can continuously generate liquidity through automated repo, collateralization, sales or conversion. Households could behave similarly as reliable access to their broader balance sheets reduces the need for precautionary cash. Money that previously remained idle can stay invested, while assets that previously sat outside the transactional system can become conditional sources of purchasing power. The same stock of conventional money can consequently support a greater volume of nominal economic activity.
AI agents amplify this process because machines manage liquidity differently from humans. People manage balance sheets intermittently, maintain excess cash for convenience and rarely optimize collateral across every asset they own. An autonomous financial agent can monitor valuations, borrowing rates, collateral requirements, payment obligations and market liquidity continuously. It can determine which asset should fund an obligation, which collateral should support a loan and where excess capital should reside every second of every day. This is the practical meaning of moving from human time to machine time. A financial system that historically operated around business hours, settlement windows and human decision cycles begins operating continuously, allowing capital to be reused and redeployed more frequently within the same year.
Stress periods will continue to distinguish money from other forms of wealth. Haircuts rise when volatility increases, credit lines contract when lenders become cautious and market liquidity can disappear precisely when borrowers need it most. Faster collateral systems could even accelerate margin calls and forced selling during a crisis. Those risks reinforce the importance of asset quality, market depth, legal enforceability and prudent collateral management within the tokenized system. They also provide a framework for thinking about the evolution of effective liquidity as a spectrum rather than a binary distinction between money and everything else. Treasuries can sit close to the monetary end of that spectrum, while volatile securities, property and private assets occupy progressively more conditional positions.
Rethinking the $40 Trillion Debt Question
This brings the discussion back to government debt. The conventional fiscal argument focuses on a $40 trillion federal liability and asks where the money will come from to finance it. Finance, however, ultimately matches liabilities with pools of capital, and the pool sitting on the other side of America’s federal debt is enormous. U.S. household and nonprofit net worth alone approaches $196 trillion, while corporations, institutions and global investors add vastly more financial wealth. Tokenization can increase the mobility of that capital by reducing settlement friction, improving collateral efficiency and connecting assets to programmable financial rails.
Greater capital mobility does not determine where that capital ultimately flows. Treasury demand will still depend on yields, inflation expectations, fiscal credibility, relative returns and risk preferences. What changes is the financial architecture through which wealth can reach competing opportunities. A household’s home equity may support borrowing that eventually reaches financial markets; tokenized securities can serve as collateral for liquidity that moves elsewhere; corporations can keep more capital invested while meeting obligations dynamically. As the friction separating pools of wealth falls, the system gains additional pathways for matching capital with liabilities. That matters when assessing whether growing government debt automatically implies an equivalent requirement for central-bank money creation.
The machine-time distinction adds another dimension to this debate. Debt is a stock measured at a point in time, while the economy’s ability to service that debt depends on flows generated through time. A $40 trillion liability does not become larger because markets close at 4 p.m., yet an economy capable of transacting, allocating capital and deploying digital labor around the clock can potentially generate greater economic throughput during the same 365-day period. The relevant comparison therefore extends beyond the stock of debt versus the stock of money. It increasingly includes the velocity and productivity of the financial and economic system supporting that debt.
This is why the question facing macro investors may gradually change. Instead of focusing exclusively on how much new money must ultimately be created to accommodate rising debt and nominal GDP, economists may increasingly need to ask how efficiently the enormous existing stock of global wealth can move toward assets, collateral and obligations, and how many times that capital can be productively deployed during a year. The answer will depend on technology, collateral quality, legal systems, market depth and the willingness of lenders to provide financing. Tokenization changes several of those variables simultaneously, while AI adds an entirely new class of economic actors capable of moving capital continuously across the rails.
None of this addresses how the debt is ultimately serviced and outgrown, which is a separate question from how it is financed. I believe that answer lies in a tax base that looks very different from today’s, one that includes longer productive lives and the output of agents and humanoids that never appear in a payroll report. That is a paper for another day. This one is about the bridge that gets us there.
When Wealth Becomes Money
For centuries, wealth and money were separated by friction, and our monetary aggregates were built for that world. Money had immediate purchasing power and stable value. Wealth needed time, intermediaries and paperwork before it could pay for anything. Tokenization compresses that distance. A house remains a house, an equity remains an equity and a bond still carries its credit and duration risk, yet each becomes easier to value, divide, pledge, finance and transfer. Asset quality will still decide how much liquidity the system extends against any of it, and stress will still separate money from everything else. But across a $217 trillion household balance sheet, even a small improvement in what can be mobilized is measured in trillions, and none of it has to sit idle as conventional money first.
That is why the liquidity expansion ahead could look very different from the ones in the history books. Every prior escape from the endgame ran through M: Draghi, quantitative easing, the COVID response. The next one may run through V. Central banks may not need to create every incremental dollar through balance-sheet expansion, and M2 may tell us less and less about the purchasing power actually available to the economy. Add an economy increasingly operating on machine time, and the same capital can be put to work more times within a single year. None of this makes the debt disappear. It buys time for the larger transformation of the tax base that longevity and digital labor will eventually force, and it does so with wealth that already exists rather than money that has yet to be created. That is a story for a future paper. The bears will come out of their caves again, and the math will look unsustainable again using today’s variables. The variables are changing. That is what happens when wealth becomes money.
