Matthew Mežinskis: Bitcoin Is Cheap and Nobody Believes It
Matthew Mežinskis returns for the quarterly update. Bitcoin is sitting near the 8th percentile of its 17-year power trend, sentiment is the worst Marty has seen relative to price, and falling cost of capital on Bitcoin-collateralized loans is the signal almost nobody is watching.

↓ Jump to the video and timestamps
Matthew Mežinskis comes back on every quarter, and every quarter I'm reminded why looking at the charts matters more than reading the timeline. We sat down this quarter with Bitcoin hovering around $80,000, a market that had just bounced off its lows, and a sentiment environment that, I said this on tape and I'll say it here, is the most dislocated from actual price I've ever seen. The shallow bear market we've been through doesn't match the doom. It reminds me of the summer of 2015 more than anything else.
Matthew's model, the power law trend he tracks at forkopolis.io/topmoney, puts Bitcoin at roughly the 8th percentile of all observed prices relative to its 17-year trend. That means 92% of the time Bitcoin has existed, the price has been higher relative to where the trend line says it should be. Meanwhile gold just came off what Matthew's regression pegged near its 100th-percentile ceiling. The Bitcoin-to-gold ratio nearly doubled in eight months. And almost nobody was watching that in dollar terms.
We also got into something I've been tracking quietly on the back end: the cost of capital for Bitcoin-collateralized lending is falling, and the amount of credit flowing into that space is rising. I think that's one of the most important fundamental signals out there right now. We closed out with a long digression on the macro picture, the debt, AI infrastructure as a potential black swan, and the sovereign individual thesis gaining ground by the day. Here's what we covered.
Key takeaways
- Bitcoin is near the 8th percentile of its 17-year power trend. By Matthew's model, 92% of all observed Bitcoin prices have been higher relative to trend than where we are now, comparable to the lows of 2015, 2018, and 2022.
- Gold's run looks exhausted on the charts. A 5.7% CAGR since 1969, a recent brush with its 100th-percentile ceiling, and an OLS projection of roughly $8,500/oz by 2050 make the gold-bull case look thin next to Bitcoin's power trend.
- The Bitcoin-to-gold ratio nearly doubled in eight months. What cost 12-13 ounces of gold per Bitcoin in February cost 20 ounces by the time we recorded. Dollar-price watchers missed the whole move.
- The cost of capital on Bitcoin-collateralized loans is falling. From conversations I've been having with companies in this space, non-bank lenders are now extending credit to Bitcoin lending desks at 6-7%, with rates dropping as more capital enters competitively. Most people have no visibility into this.
- The global monetary base picture still has enormous room. Bitcoin's market cap is a rounding error next to $26T in global base money and roughly $158T in broad money, by Matthew's aggregation. The penetration trade has decades to run.
- Sentiment is more broken than it should be. The dislocation between a relatively shallow bear market and historically dire mood is larger than I've seen at this price level. The charts say cheap. History says trust the charts.
Bitcoin at the 8th Percentile, What the Power Law Actually Shows
Matthew tracks a power law model for Bitcoin going back to its earliest price data, 17 years of observations, with a 96% R-squared fit when you plot it log-log. The result is a straight line, which is something you do not see in any other financial asset he tracks. Stocks, bonds, gold, all exponential trends when you run the same analysis. Bitcoin is the only one that gives you a power trend.
Right now, by that model, Bitcoin is sitting just below the 8th percentile. The way to read that: 92% of all observations across Bitcoin's history have shown the price higher relative to where the trend says it should be. The exceptions, the times Bitcoin was this cheap relative to trend, were the 2015 lows, the 2018 lows, and the FTX-era collapse in late 2022. Peter Zeihan went on Joe Rogan around that last one and, as Matthew recalls it, said Bitcoin was $17,000 overvalued at $16,000. Those are the moments you're supposed to remember, because the setup right now is the same percentile-wise.
Matthew also ran through the quarterly candle pattern. Every cycle has produced at least four red quarterly candles. This cycle has had three full red candles so far, which means another weak quarter wouldn't be a surprise. He's not predicting one, but he's not ruling it out either. Could we see the 5th percentile? Maybe. But the trend is the trend, and at this level, the charts are saying something the sentiment is not.
Gold vs. Bitcoin, The Ratio That Tells the Real Story
I tweeted out Matthew's chart the week before we recorded: Bitcoin at the 8.4th percentile, gold at the 93.5th. That split says more than almost anything else I've seen this cycle.
Matthew's gold model is a 55-year exponential trend going back to 1969, with a CAGR of 5.7%. That's the rate gold has compounded at since Nixon closed the gold window. This winter, by his regression, gold was pushing up toward the 100th percentile of its own long-run trend, an event roughly comparable to the January 1980 spike to $850 that took two decades to recover from. Matthew called the top publicly and in real time.
The Bitcoin-to-gold ratio tells the story even more cleanly. When you plot Bitcoin's price in ounces of gold, the log-log line goes straight, a power curve, not an exponential. That's the fingerprint of Bitcoin's network effect dominating the relationship.
Back in February and March, you could buy one Bitcoin for 12-13 ounces of gold. By the time we recorded, it would cost you 20 ounces. That's nearly a doubling of the ratio in eight months, in a period when most people thought Bitcoin was going sideways or down. Matthew's point: the trend line on the Bitcoin-to-gold cross is growing every quarter. He thinks 10 ounces per Bitcoin is probably permanently off the table at this point, barring a major black swan. The OLS projection for gold out to 2050 sits around $8,500 an ounce, which is fine, but compare that to where Bitcoin's power trend is pointed.
You don't need alchemy. You don't need fusion reactors turning mercury into gold at some enormous cost per gram. Just look at the trend.
The Monetary Base Race, Where Bitcoin Actually Sits vs. Global Central Banks
Matthew's monetary base chart is one of my favorite things he's built. It stacks the top five central bank money supplies against Bitcoin's market cap, so you can watch the race in real time.
By his compiled data from public central bank sources, the current order is: Chinese yuan at roughly $6T, US dollar at $5.5T, euro at $4.6T, yen at under $3.5T, and British pound at around $1T. Bitcoin's market cap bounced off approximately $1.2T at the end of Q2 and was back up well above that by the time we recorded, trading, as Matthew puts it, as a top-10 stock even though it isn't a stock.
What I find interesting eye-testing the chart is the divergence. The yen and euro are moving one direction. The yuan, dollar, and Bitcoin are moving the other. The yuan's trajectory is an open question, China has the fastest-growing debt load of any major developing economy, which puts a ceiling on how long that trend holds.
On the Fed's side of things, Matthew tracks the Fed's proportion of US federal debt as a monetization indicator. The peak was around 20% in 2021, during the pandemic QE surge. Now it's 11.3%, by his reckoning. That gap matters because it means there's historical precedent for the Fed to double its balance sheet if it needs to act fast. Nobody should hope for that. But it's worth knowing it's within the range of what's happened before.
Broad money globally Matthew estimates at roughly $158T, and he's explicit that this excludes eurodollars, which are their own enormous category. The stablecoin market, Tether around $182B, combined with USDC puts the total at roughly $250B, is a rounding error against those numbers. The Genius Act, Bessent's front-end Treasury strategy, the theory of hot-swapping long-duration debt for short via stablecoin demand rails, all interesting to watch, but the numbers are still tiny. The penetration trade for Bitcoin has a long way to go before the math gets crowded.
The Quiet Bullish Signal, Falling Cost of Capital on Bitcoin-Backed Loans
This one doesn't show up on any chart I've seen published, and I think it's one of the most important things going on in Bitcoin right now.
From conversations I've been having with companies operating in this space, non-bank lenders are now extending credit to Bitcoin lending desks at 6-7%, and that rate is falling. The driver is competitive pressure, more capital is entering the space, and as it does, the marginal cost of borrowing against Bitcoin collateral drops. Credit funds, and some banks, are getting comfortable with it.
Most people have zero visibility into this. But think about what it means structurally. If you can borrow at 6-7% and Bitcoin's power trend is running at roughly 40% trailing 12-month growth, there's substantial coverage for your collateral before you'd ever have to scramble. That's the Saylor bet applied at a lower cost of capital, and it's happening in the non-bank world before it's really happened in the banking world.
Matthew connected this to a bigger question he's been asking about the power-vs-exponential debate. If enough fiat capital gets levered against Bitcoin, if Bitcoin becomes just another financialized collateral asset like gold or Treasuries, the price action could shift from a power trend to an exponential trend. He thinks that would break Bitcoin to the upside, not the downside. But it also raises the harder questions: more centralization, more KYC, a future that looks more like TradFi than cypherpunk. Those are open questions, and we don't have clean answers. What I'd say is we're not anywhere near that regime shift yet. At $26T in global base money and $158T in broad money, Bitcoin is still negligible as a collateral asset at scale. The Aven Bitcoin Visa card is one product that shows where this is heading at the consumer level, borrow against your stack, no rehypothecation, BitGo custody, but the institutional version of that story is only starting to develop.
The Macro Stress Test, Debt, AI Infrastructure, and the Black Swan Setup
The September 2019 repo market seizure is the historical parallel Matthew keeps coming back to when he looks at the current Fed balance sheet trajectory. Back then, the market seized up because the system was trying to normalize after years of QE, and the result was a forced expansion of asset purchases. The same normalization attempt is happening now, with one critical difference Matthew flagged: back then rates were being targeted lower. Now they're rising. It's almost the opposite scenario.
That matters because it narrows the playbook. If the black swan comes, whatever form it takes, the Fed has room to act. Matthew's chart shows 11.3% ownership of federal debt, well below the 2021 peak of 20%. They could double the balance sheet and stay within historical precedent. Nobody wants that. But the option exists.
On AI infrastructure specifically: I raised Anthropic's S-1 filing as a live test case. The question is whether any of these companies can actually price in the public markets, and what the signal is if they can't. There are hundreds of billions, maybe trillions, of dollars locked into the AI build-out trade right now. If the IPO window closes, that's a stress event. Matthew's comparison to 2019 holds there too, in 2019, the yield curve inverted in August and almost nobody connected it to what was coming. The market didn't price the risk until much later.
The debt picture underneath all of this is the one Matthew quotes John Williams on: if you cut every single line item in the federal budget except Social Security, Medicare, and Medicaid, you'd still be running a deficit. That's the math. The Trump accounts being extended to anyone under 40 is the beginning of an attempt to hot-swap Social Security out for an equity-index portfolio, I've been watching that develop and I think it signals the third rail is finally starting to weaken. Whether it actually leads anywhere is another question.
Dropping forward guidance at the Fed is the one thing I'll say I genuinely like about where things are heading institutionally. Markets figuring things out for themselves instead of tea-leaf-reading the FOMC is healthier than the alternative. Warsh seems more independent than some feared. We'll see how long that holds.
Zooming Out, The Sovereign Individual Frame
I've been thinking about this a lot lately. The Sovereign Individual thesis is getting stronger by the day.
The populist outsider arriving to run on immigration and tax reform. Asymmetric drone warfare making trillion-dollar naval assets potentially vulnerable to a cheap drone strike. The entitlement third rail weakening. Trump accounts as a prototype for unwinding Social Security's monopoly on retirement security. All of it points in the same direction: the industrial-era nation-state is losing its monopoly on the things it used to exclusively provide, and individuals are finding their own exits.
Matthew lives in the Baltics and sees the drone warfare reality up close in a way most Americans don't. Air raid alerts for his kids that are drone warnings, not tornado warnings. His read on the Russia-Ukraine conflict is shaped by geography in a way most podcast commentary isn't. I'll say for clarity: I am very anti-war and would like to see a resolution to this as soon as possible. The guests I've had on who discuss Trump's foreign policy approach aren't pro-Russia, they're amenable to Trump's posture on the conflict, which is a different thing. And Matthew's point that Europe building out its own defense capacity and security infrastructure, rather than depending on the US umbrella, is a genuine net positive regardless of where you land on the rest of it. That part I agree with completely.
The Fourth Turning crowd says no resolution to the current disorder until the 2030s. Maybe. My hope is simpler: that we get pushed to the precipice, look around, and collectively decide to focus on the productive stuff. We have the tools for a legitimate golden age right now if we'll take it. Realpolitik gets in the way. But Bitcoin keeps running on its trend line regardless of what the news cycle decides to do this week.
That's the point of doing this with Matthew every quarter. Look at the chart. Touch grass. Log off X every once in a while. We're going to win.
About Matthew Mežinskis
Matthew Mežinskis is a Bitcoin analyst and the creator of the Crypto Voices podcast and the monetary base and power law research published at forkopolis.io. He has been tracking Bitcoin's power law trend, the global monetary base comparison, and the Bitcoin-to-gold ratio for nearly a decade and is one of the longest-running recurring guests on TFTC. He is based in the Baltic region of Europe.
Sources mentioned
- forkopolis.io/topmoney: Matthew Mežinskis's live charts on the Bitcoin power law percentile, monetary base comparison, and Bitcoin-to-gold ratio, the primary data source for this episode
- September 2019 repo market stress (Federal Reserve Bank of New York): the repo market seizure Mežinskis uses as the historical parallel for the current Fed balance sheet trajectory
Watch the conversation
Timestamps
- 0:00 - Intro
- 4:06 - Bitcoin price and sentiment disconnect
- 7:05 - Power law percentile: Bitcoin at the 8th percentile
- 13:29 - Bitcoin-to-gold ratio
- 18:49 - Global monetary base update
- 25:22 - Stablecoins, Tether, and the Genius Act
- 31:35 - AI infrastructure and the IPO question
- 34:08 - ColdCard vulnerabilities and Bitcoin security hardening
- 43:38 - Warsh and the Fed: no forward guidance
- 46:17 - Social Security hot-swap and Trump accounts
- 53:28 - Sovereign individual thesis and drone warfare
- 56:57 - Bitcoin's path to $100T and the power-vs-exponential question
- 1:35:46 - Bitcoin collateralized lending and falling cost of capital
Sponsors
- Cash App: For a limited time, new customers can get $21 added to their balance. Just use code TFTC10 when you sign up, and send at least $5 to a friend in the first two weeks. cash.app/app/TFTC
- Square: For up to $200 off eligible Square hardware. square.com/go/tftc
- Bitkey: Use code TFTC10 for 10% off the new Bitkey. bitkey.world
- Aven: aven.com/bitcoin
- CrowdHealth: joincrowdhealth.com/tftc
- Simple Mining: simplemining.io/tftc
- Salt of the Earth: drinksote.com/tftc
Frequently Asked Questions
What is the Bitcoin power law percentile and what does it tell you?
Matthew Mežinskis tracks Bitcoin's price against a power law trend built on 17 years of price data, with a 96% R-squared fit when plotted on a log-log scale. The percentile tells you where the current price sits relative to all historical observations against that trend. At roughly the 8th percentile, 92% of all recorded Bitcoin prices have been higher relative to the trend than they are now, placing current prices in the same territory as the 2015, 2018, and 2022 lows.
Is Bitcoin cheap right now compared to its historical trend?
By Matthew Mežinskis's power law model, yes, Bitcoin is sitting near the 8th percentile of all historical price observations relative to trend. That makes it historically cheap on his framework, comparable to the major cycle lows of the past decade. This is Mežinskis's model and his attribution, not a claim derived from independent consensus analysis.
How does the Bitcoin-to-gold ratio work and where does it stand?
The Bitcoin-to-gold ratio measures how many ounces of gold it takes to buy one Bitcoin. When you plot this ratio on a log-log scale, it follows a power curve, the same shape as Bitcoin's dollar price, rather than the exponential curve that gold itself follows. As of the recording of this episode, the ratio had moved from roughly 12-13 ounces per Bitcoin in February and March to around 20 ounces, nearly doubling in eight months even as dollar-denominated Bitcoin price seemed flat to many observers.
What is Matthew Mežinskis's Bitcoin price model?
Mežinskis publishes his power law model and monetary base comparisons at forkopolis.io. The model fits a power regression to Bitcoin's full price history on a log-log scale, producing a straight trend line with a 96% R-squared. He tracks where the current price sits as a percentile of all historical observations against that trend and uses it as a long-run valuation tool, not a short-term price target.
How does Bitcoin's market cap compare to the global monetary base?
By Mežinskis's aggregated data from public central bank sources, the top five central bank monetary bases total roughly $20T+, with Bitcoin's market cap in the $1.5-1.7T range, meaningful but still a fraction. He estimates global broad money at roughly $158T, excluding eurodollars. On his power law projection, Bitcoin's market cap could approach the size of the global monetary base sometime in the late 2030s, though he's clear that's a model output, not a certainty.
What is the cost of borrowing against Bitcoin collateral right now?
From conversations I've been having with companies operating in the Bitcoin lending space, non-bank lenders are now extending credit to Bitcoin lending desks at roughly 6-7%, with that rate falling as more capital enters the market competitively. This is my own sourcing from the field, not a published data point. The direction of travel, more capital, lower cost, is what I find significant as a structural indicator of Bitcoin's maturation as a collateral asset.
Could Bitcoin shift from a power law to an exponential trend, and what would that mean?
Mežinskis raises this as a genuine open question. If enough fiat capital gets levered against Bitcoin, through collateralized lending, institutional products, and financialization, the price action could be pulled into the exponential growth pattern that governs most financial assets. His view is that this would break Bitcoin's price to the upside, not the downside. The risk, in his framing, is that it pulls Bitcoin into a more centralized, KYC-heavy financial future rather than the permissionless network it is today.


