Raoul Pal says investors should treat 11% annual compounded returns as the real cutoff between preserving wealth and losing purchasing power. In a column published under his own name, Pal argues that global liquidity expands at roughly 8% a year and that another 2% to 3% of inflation lifts the effective hurdle rate to about 11%.
Under that framework, assets earning less than that level may still rise in nominal terms while leaving holders poorer in real purchasing-power terms. Pal says the old allocation model built around bonds, property, gold, and broad diversification no longer works the way many investors assume.
Why Pal uses 11% as the benchmark
Pal starts with a simple view of economic growth. He says there are only three ways a country grows wealth: by adding more labor, by improving output per worker, or by expanding through debt.
For most of the last century, population growth and productivity gains did the heavy lifting. Now, he argues, both engines have weakened. Birth rates started falling decades ago, labor-productivity growth has slowed, and economies rely far more on debt expansion. Debt carries interest, and in Pal’s telling, the politically viable way to service that burden is monetary expansion.
That is the base for his 11% number. He describes global liquidity as the total stock of money and credit in the financial system and says it grows by around 8% a year. Add everyday inflation of 2% to 3%, and the result is a rough annual threshold of 11%.
Pal’s conclusion is blunt: if an asset compounds above 11%, purchasing power can rise. If it compounds below 11%, the investor may be losing ground even if the account balance is growing. He says that changes how assets should be judged. The question is no longer whether something looks safe or well diversified, but whether it can clear that line over time.
Bonds, property, and gold under his framework
Pal describes bonds as a straightforward lending contract. Investors lend capital, receive a fixed coupon, and get principal back at maturity. The flaw, he says, is that the coupon is usually set against ordinary inflation assumptions and does not account for the loss caused by currency dilution.
He uses a 4% government bond as an example. Even if the investor holds it to maturity and receives every promised payment, the real purchasing power of that capital still falls if the underlying currency is losing value at around 8% a year.
On property, Pal takes a more mixed position. He says real estate can still help hedge against currency debasement because buyers use fixed-rate debt to buy a hard asset priced in money that keeps losing value. He also notes that he personally owns property.
Still, he argues that the big wealth effect many households experienced through housing was tied to a long period of falling rates and favorable mortgage windows. In his account, that backdrop is gone. Rates first dropped toward zero and later moved higher, making the old playbook much harder to repeat. He adds that most people now face stretched price-to-income ratios and elevated mortgage costs.
Pal also says real estate has lagged global liquidity growth since 2007. In nominal U.S. dollar terms, property prices may be higher, but he argues that their real purchasing-power value has weakened against the pace of monetary expansion.
His treatment of gold is more careful than outright bearish. Pal says gold has gone through a major rally this year, writing that it rose above $5,500 an ounce in January before pulling back to around $4,600 by late August, when he wrote the piece. That still left it up by about one-third for the year.
Even so, he says gold’s role is mainly to keep pace with central-bank balance-sheet expansion rather than generate compounding growth on top of it. He compares gold to a store of value that can defend purchasing power against monetary debasement. What it does not have, in his view, is a user-adoption curve or an expanding business ecosystem that would support long-run compounding in the way a network asset can.
In short, Pal says gold may preserve wealth, but it does not create new wealth in the same way growth networks can.
Why he puts technology and crypto in a different category
Pal says broad equity exposure has done better than the traditional defensive assets he critiques. Over the last 10 years, he puts the S&P 500 at roughly 13% annualized and the Nasdaq-100 at about 20%.
He pairs those figures with about 12% annualized for gold and 58% to 70% for Bitcoin over 10 years, depending on methodology and start date. Against an 11% hurdle rate, he says the relative spread is clear.
For Pal, the point is not simply that these returns happened. He says the more important issue is whether there is a repeatable structural reason behind them. His answer is that technology assets and crypto assets both benefit from adoption curves that follow an S-shape.
He brings in Metcalfe’s law to make the case. As networks gain users, the value to existing users increases. Adoption is not linear; it tends to start slowly, accelerate sharply, then flatten as a market matures. As long as an asset remains in the steep part of that curve, Pal argues, it has a structural path to outrun money-supply growth rather than relying on speculative bursts alone.
That is why he says the real debate is not whether tech and crypto have already beaten legacy assets. The more important question is whether their user-penetration curves have already peaked. His answer is no, and he ties that view to a coming wave of non-human participants.
Why he thinks traditional diversification is less effective than investors assume
Pal argues that classic portfolio construction rests on a premise that has weakened since 2008. The old model treated bonds, gold, real estate, and equities as separate risk buckets that could offset one another when one area broke down.
In his view, liquidity has become the dominant macro driver across all of them. Bonds are a liquidity trade, gold is a liquidity trade, real estate moves with the liquidity cycle, and index funds are also tied to that same broad force, just in a more polished wrapper.
He does not reject diversification itself. In fact, he says he holds diversified positions too. His objection is narrower: if three of the four asset groups in a portfolio cannot clear the 11% line and the fourth only barely does so under unusually strong conditions, spreading capital across them does not solve the purchasing-power problem.
Pal says the core question is not how many positions an investor owns. It is whether the capital meant to generate long-term gains is actually placed in assets with the capacity to compound above the real hurdle rate.
Inside crypto, Pal leans toward base-layer exposure
Pal does not argue for a one-coin approach inside digital assets. He says application-layer protocols can produce very high returns, and the right project tied to a real business need may even outperform a base-layer network.
The challenge, he says, is identifying those winners in advance. Base-layer public blockchains sit underneath the whole ecosystem and carry settlement activity across applications. In his framing, value accrues to infrastructure even if investors cannot predict which single application will dominate later on.
He also argues that conventional equity valuation methods do not translate cleanly into crypto networks. He points to metrics commonly used in the sector, such as fee multiples, revenue growth, and value-locked ratios, and says they have not shown reliable predictive power in backtests run by GMI across 12 major blockchains. The one factor he says did matter was whether capital entering an ecosystem stayed there over time.
AI agents as the next users of public blockchains
One of the article’s most forward-looking claims is that the next adoption wave may come from AI agents rather than people. Pal says most market-sizing reports still assume that the users of digital ecosystems are human, acting on human schedules with a handful of transactions, payments, and queries per day or month.
He says that assumption is close to breaking. In his account, AI agents are moving toward becoming autonomous economic actors rather than mere tools. He cites industry forecasts he has seen suggesting that the ratio of non-human intelligent identities to human employees inside companies could reach 80:1.
Pal argues that such agents cannot open bank accounts, lack legal identity, cannot visit branches, and are poorly matched with financial systems that stop settlement in the evening and may take three days to move money. They need programmable money and payment rails that run continuously, which he says public blockchains already provide natively.
He names several developments to support that argument:
- Anthropic has open-sourced its Model Context Protocol.
- Google has launched Agent2Agent and released a preview of WebMCP, allowing websites to expose functions directly to agents.
- Coinbase has restarted the x402 protocol so agents can pay each other over HTTP links.
For Pal, those product releases point to future settlement demand whether or not speculative capital arrives at the same time.
No all-in trade, and no leverage
Despite his strong preference for growth networks, Pal says he does not advise putting all capital into a single asset. He writes that markets constantly showcase all-in success stories while the many traders who concentrated, blew up, and disappeared rarely get the same visibility.
His portfolio logic comes down to two variables: how much capital is allocated to a genuine long-term growth track, and how long that capital can stay invested.
He adds a third rule and treats it as non-negotiable: no leverage. Not a small amount, not “careful” leverage, and not leverage justified by stop losses. His reason is survival through deep drawdowns. If an investor can sit through a 50% decline without forced liquidation, the long-term thesis can still play out. With leverage, even a correct 10-year view can be destroyed by a two-week collapse.
What he says ordinary investors should do with the framework
Pal does not end with a fixed asset-allocation template. He says that would make little sense without knowing a person’s debt burden, investment horizon, and risk tolerance.
What he does offer is a screening standard. Keep some traditional assets for security and peace of mind, he says, but allocate a meaningful share to long-duration growth themes and size that exposure so a 50% drop would not force a panic sale. He adds that, after 13 years watching crypto markets, the investors who compound over long periods are often not the ones trading most actively.
His final point returns to the 11% benchmark. If a portfolio compounds below that level, he says, the wealth created by one year of labor buys less freedom the next year. That is why he wants investors to review every holding against the same question: can it clear the hurdle, or is it quietly consuming time and optionality?