The Meb Faber Show: Dance If You Like — Just Count Your Steps to the Fire Exit

Contents

1. I was waiting for the AI numbers to fail first
2. The money that popped the Nasdaq was withdrawn in January 2000
3. The two numbers I now watch, and three places he has not convinced me
4. Worth a look
5. The one thing to take with you

Mist wraps the cold water, moonlight wraps the sand; tonight I moor on the Qinhuai, close to the wine houses. The singing girls know nothing of a kingdom lost — across the river they are still singing “Flowers in the Backyard Court.”

—— Du Mu, “Mooring on the Qinhuai” (Tang, c. 9th century; my own translation)

Episode 656 of The Meb Faber Show, published 6 November 2026, has Meb Faber hosting Albert Edwards, global strategist at Société Générale, who has held the same job for 44 years and says he has never been promoted. The S&P 500’s twelve-month forward P/E has already fallen from 23–24 times to below 19, and tech from 33 times to around 20, with AI-driven earnings expectations holding the tape up. The market has priced two or three hikes; Edwards thinks the Fed needs 200 basis points, and that 100 basis points more on the ten-year would be enough to cut off the liquidity equities run on. On a longer horizon he expects US ten-year yields at 10%, a forward P/E he imagines near six by then, and the S&P down 50 to 60% at some stage — that last one he says he is certain about. He also hands you the counterweight himself: quantitative easing had him wrong for twenty years, and the fiscal dominance and money printing he now expects are the same thing that bailed the market out back then.

I was waiting for the AI numbers to fail first

Before this episode my script was simple. For this cycle to end, the AI arithmetic had to break first, so what I tracked was return on capital expenditure, chip depreciation schedules, and the circular deals where a handful of companies invest in each other and buy from each other. I assumed the day the story got falsified was the turning point, and that until then the narrative would hold the tape up.

Edwards put up a chart that stopped me. The ratio of forward earnings to the last twelve months of actual earnings is stretched to a level rarely seen. Historically it only gets this stretched coming out of a recession, when trailing earnings have fallen into a deep hole and the denominator is crushed. This time there is no hole — trailing earnings have done well, and forward estimates are still pulling away from them. The optimism is not borrowed from a low base; it is added on top.

Then he said the thing that moved my whole frame sideways. Across 44 years — the Asian crisis in 1997, the Mexican crisis before it — every bubble came wrapped in a plausible, compelling story. The narrative is the constant. Being present every time, it cannot explain the timing of this one. His own framing: he knows he is called a permabear, but most of his clients have to stay fully invested, they have to participate, and “you’ve got to keep dancing while the music’s playing.” What he is actually calculating is whether we should be in the middle of the dance floor enjoying ourselves, or dancing next to the fire escape.

The money that popped the Nasdaq was withdrawn in January 2000

His alternative answer is liquidity, and the example he gives is a date rather than a concept.

Through all of 1999 rates were rising and bond yields were rising, and the Nasdaq went up anyway; the market ignored it. The reason was Y2K. Central banks were so frightened that everything would shut down at midnight that through the second half of 1999 they hosed liquidity into the system while still raising rates. Once 1 January passed and nothing happened, they decided they had been spoofed and took the money back out — the level of US base money started falling sharply, and that was the timing point for the top. Bubbles are popped by the money being withdrawn; whether the story breaks is the explanation everybody writes afterwards.

Measured on that gauge, the reading now is tight. He cites Michael Howell of CrossBorder Capital — his king of liquidity — who had been bullish until this year and has turned. US money supply is picking up, M2 around 5%, nominal GDP growing at 6% (real closer to 2%), and that is not enough to absorb the issuance now coming, public and private. Equities rise on the surplus liquidity the real economy does not need, and the surplus is being eaten.

The biggest single bite is the IPO calendar. Meb and his producer went looking for a precedent — a company that went public in the US and landed straight in the top ten by market capitalisation — and after the Second World War the only one they found was Ford; before that you have to go back to US Steel. SpaceX just did it, and two more could follow within a year in Anthropic and OpenAI. Inflation-adjusted, those three combined are bigger than every IPO of the 1990s put together. Edwards treats a mega-IPO the way he treats a large bond issue: a draw on the existing money supply, pulling from the same pool, pushing yields up unless someone prints enough to offset it.

Here he makes a correction to the transmission channel that I had not thought about. The textbook chain runs: rates rise, corporate interest costs rise, profits or margins get squeezed, companies cut investment, inventories and employment, the economy turns. US corporates cut that chain this cycle — the large ones locked in near-zero coupons on corporate bonds during and just after COVID and parked the cash on deposit, so higher rates pay them. The sector carrying too much leverage this time is the public sector, which took the excess debt off the private sector’s books after the financial crisis.

With the old road blocked, the road he thinks it travels instead is greedflation. In 2022 he watched the US corporate sector do something he could not find in the record: unit costs went up and unit profits went up at the same time. In every prior cycle rising costs compressed margins, and that compression is the standard start of a downturn. That time, with COVID supply constraints and the Ukraine war pushing raw materials and oil up, companies noticed the stimulus cheques had just landed on doormats and demand was overflowing — so they raised prices past what protecting margins required and expanded margins instead. The St. Louis Fed later traced that profit-led inflation to retail, wholesale and property.

He says it is happening again. Oil has pushed US unit non-labour costs up around 10%, while productivity growth running above 3% keeps unit labour costs near zero, the two offsetting each other — companies have no cost case for raising prices, prices are rising anyway, and margins are soaring. His phrase: corporates kill the goose that laid the golden eggs. If inflation picks up on that, the Fed finds itself behind the curve the way it did in 2022, and further behind: the market is pricing two or three hikes, he is pricing 200 basis points and quickly. Then his tone hardens. Even if the Fed never delivers all of it, the market starting to believe it would put the ten-year 100 basis points higher, and that alone is curtains for equities. You do not have to wait for the AI profit cycle to turn.

And yields rising now means something opposite to what it meant five years ago. Simon White of Bloomberg gives the threshold: 5% is where the stock-bond correlation flips from negative to positive, driven by inflation sitting above 2%. Edwards has history with that line — in 1996 he wrote up his Ice Age thesis, the argument that the West would follow Japan into low growth, low inflation and a decades-long bond bull market, years before Larry Summers called it secular stagnation. Twenty years of his career ran on it, and he has declared it dead himself: we are in a multi-year, multi-decade bear market for government bonds, the correlation has flipped back to positive, and higher yields pull equities down. Anyone holding stocks and bonds together has lost a layer of protection.

He also put up a chart that reminded him of thirty years ago, from Torsten Slok at Apollo: US total factor productivity has fallen from around 3% eighteen months ago to below zero. Total factor productivity counts the capital stock as well, so labour productivity above 3% alongside negative total factor productivity means the capital going in is not coming out as output. It reminded him of how Paul Krugman, back in 1995 when he still wrote economics, described the East Asian miracle: the miracle was mobilising the factors, how well they were used was another matter, and total factor productivity growth was poor throughout. Looking at the chart he thought of Thailand and Malaysia in the mid-nineties — a mirage. The returns may arrive further down the road; on the liquidity clock, there is a roadblock before that.

The two numbers I now watch, and three places he has not convinced me

Two slots on my watchlist changed. Where “return on AI capital expenditure” used to sit, I now watch the direction of margins when costs rise — costs up and margins up together is his signal. Where “did earnings beat” used to sit, I now watch the issuance calendar for the next few months, government paper plus mega-IPOs. Both are monthly, both are checkable, and neither waits for anyone to admit the story broke.

Before the quibbles, one belief a lot of people carry, me included: that bond fund was bought as the hedge, so when stocks fall it should rise. In 2022 it fell alongside stocks and I filed that as a one-off caused by extreme inflation. In Edwards’ frame it was a regime change — our generation’s intuition about stocks and bonds offsetting each other grew up inside his Ice Age, fed by very low inflation and yields grinding down. Go back to the first half of the twentieth century and the years when bonds failed to hedge equity drawdowns are everywhere. For shock absorbers in this environment Edwards points at inflation hedges — commodities, value stocks, that family — and my own reading is that one asset should not be asked to be the growth engine and the insurance at once. That was a free lunch of the zero-yield years.

On his conclusions, three things leave me unconvinced.

First, total factor productivity gets revised heavily, a single negative print cannot carry much weight, and his whole East-Asian-miracle analogy hangs on that one chart. Second, the 5% correlation threshold comes from one analyst’s empirical estimate; it is useful, and I have not seen it tested across periods and countries. Third, two time scales are mixed together. His near-term script is yields spiking, the Fed tightening until the economy tips into recession, then yields falling back inside that recession — he quotes Ray Dalio on rates making a series of higher lows and higher highs. The 10% ten-year is “a couple of cycles” away. Heard together, those two layers sound like a forecast for next year.

The deeper question he handed me himself. What derailed the Ice Age thesis was quantitative easing: “you talk about swimming naked when the tide goes out, but there was so much QE you couldn’t tell who was swimming naked.” The script he expects now is fiscal dominance — the Social Security fund draining to zero within a few years, benefits then having to fall by about 25%, no politician anywhere willing to touch it, and central bankers conscripted into printing. But printing is the very thing that had him wrong for twenty years. The same event is both his reason for a 10% ten-year and the mechanism that pushes yields back down and lifts asset prices, and he does not settle the order. On the politics he is concrete: selling bitter medicine to an angry electorate takes a crisis first, as when Britain had to call in the IMF in 1976 and Labour prime minister James Callaghan stood in front of his own party congress to say they could no longer spend their way out, and was roundly booed. The American fragility he points to is maturity structure: 18% of federal debt sits in three-month bills, 35% matures inside a year, and interest payments are already 4% of GDP, the highest in the G7.

One counter-intuitive detail to close on. This is a man who believes in the debasement trade and in precious metals, and in January he wrote that gold’s monthly RSI at 95 and quarterly above 90 was the most overbought market he had seen in his lifetime; with gold back near 4,000 the monthly RSI is still above 70. Being long the structural logic and expecting the technicals to punish it coexist in him. He wore a flowery shirt for the show — “this is my autumn leaves” — because what he talks about is so depressing the clothes have to be cheerful; when Meb asked where people could follow his writing, he said, “I wouldn’t recommend that.”

Worth a look

• The Meb Faber Show, episode 656, 6 November 2026 — the source for this piece
• Torsten Slok at Apollo on US total factor productivity, and the chart behind it
• Paul Krugman’s 1995 essay on the East Asian miracle and factor accumulation
• Michael Howell (CrossBorder Capital) on global liquidity versus issuance
• The St. Louis Fed’s work on the sources of profit-led inflation in 2022
• US Treasury statistics on the maturity structure of federal debt, and the annual trustees’ report for the social insurance trust funds

The one thing to take with you

One idea: you mostly do not get to decide when the music stops, but where you stand on the dance floor is yours to choose. After 44 years what Edwards cares about is distance — stay at the party either way, and standing by the door versus standing in the middle changes what the end of the song costs you.

One exercise for today, which I have tried myself: pick something you are all in on and cannot walk away from — a job, a lease, a project you have carried a long time — and write three lines on paper. First line: if you had to exit tomorrow, what is the one thing you would have to finish. Second line: how many days that thing takes. Third line: what first step of it fits in twenty minutes today. Then go do the third line. When I did it, something I thought I could wind down at any time turned out to need three weeks.

Raw Markdown

This article is an educational discussion of investment method. It is not advice to buy or sell any individual security, offers no target prices, and does not analyze any current holding. Investing carries risk; make your own decisions or consult a qualified professional.