Six Years Later, the Super Cycle Is Still Winning

• Leuthold's 2020 Super Cycle work showed post-war bull runs follow 10 years of revaluation with 20 years of boom.

• Updated through October 2026, the S&P 500 has caught up to the 1968-1997 cycle.

• AI-driven compute demand supports a 2029 run to 10,500.

The bears called time on this bull market six years ago, and they're still waiting…

Back in January 2020, plenty of economists and forecasters thought the economic expansion and the S&P 500 Index's run had grown "long in the tooth." Stocks had tripled in a decade, and valuations were stretched. The consensus said gains were overdue for a reversion to the mean, which implied low- to mid-single-digit returns over the next 10 years.

Jim Paulsen at The Leuthold Group wasn't so sure. He agreed the bears' logic made sense, but he argued the post-war history of the U.S. stock market told a very different, and more optimistic, story. In his view, stocks were in the middle of a "Super Cycle." He admitted he didn't know who'd be right. And he couldn't have known COVID was just weeks away.

The pandemic triggered a flurry of wartime-like spending. Central banks slashed interest rates, governments flooded their economies with stimulus, and billions poured into vaccines and treatments. That money sparked medical and technological innovation at a wartime pace. It let businesses prosper with much of the workforce at home, and it laid the groundwork for today's AI build-out.

Now the skeptics are back, questioning whether all that AI spending will pay off. The latest evidence points the other way. Late last week, Amazon CEO Andy Jassy said AWS capacity is largely sold out for 2027, with a good chunk of 2028 already spoken for. And Anthropic's draft S-1 shows the company generating $1.70 in revenue for every dollar it spends on compute. That's why I think Paulsen's Super Cycle is still very much intact, with room for the S&P 500 Index to run through the end of the decade.

But don't take my word for it, let's look at what the data's telling us...

Paulsen's case rested on a simple pattern. Since 1940, every Super Cycle has followed the same script. The first decade revalues the market. Multiples get crushed, and stocks churn sideways for an average gain of just 1.5% a year. Then the boom kicks in. The second decade averaged 12.9% a year, and the third averaged 10.6%, both with less volatility than the first.

The current cycle has followed that script almost to the letter. The 2000s were the revaluation decade, and they were brutal. Two bear markets left the S&P 500 lower at the end of 2009 than where it started the decade. The 2010s brought the boom, with gains of about 11% a year. And since January 2020, the index has climbed about 13% a year, right through a pandemic recession and the 2022 bear market.

Here's where it gets interesting…

When Paulsen originally published his paper, the current cycle was trailing both historical paths. Today, it's caught up. As the updated chart shows, the S&P 500 now sits just above the 1968-1997 path at the same point in the cycle. The market closed the gap.

Now for the honest part. Valuations are richer than they were at this stage of past cycles. When Paulsen wrote, the Shiller CAPE ratio was 30.5. Today, it's near 41, well above the top of its post-1990 range. And with the 10-year Treasury yield north of 5%, bonds offer competition that didn't exist in 2020.

But history says valuations shape how the final stretch plays out, not whether there is one. In the 1940 cycle, the CAPE ratio drifted lower over the last three years, and stocks gained about 5% a year. In the 1968 cycle, it climbed from roughly 20 to 33, and the market surged. That's where earnings come in. Jassy also said AI labs are using "gobs and gobs of compute" unlike anything he's seen. He doesn't expect the shortage to end in 2027, and he said Amazon could invest a lot more without it being a bad use of capital. That's the kind of earnings power that can support a higher multiple.

So where could it go? The cycle has a little over three years left. If the S&P 500 matches Paulsen's 10.6% average for the final boom decade, it would sit near 10,500 by the end of 2029. If it rhymes with the softer finish of the 1940 cycle, it lands closer to 9,000. And if it tracks the late-1990s surge of the 1968 cycle, it could top 15,000.

That's a wide range, and the ride won't be smooth. Paulsen's history shows the final boom decade has averaged 1.5 recessions and 2.5 bear markets, and the current tech-led pullback is a reminder of that. But don't let short-term hiccups scare you out of long-term gains.

Investors faced the same test in 1998. As the chart below shows, the Nasdaq's run since 2023 has tracked its 1995-1998 climb almost step for step, and today's pullback lines up with the summer 1998 selloff. Back then, the Asian and Russian debt crises knocked the S&P 500 down more than 19% in about six weeks. Those who held on saw it climb nearly 60% over the next 19 months, while the Nasdaq more than tripled. Yes, that run ended in a bust. But that cycle was already 30 years old, while this one still has a few years to go.

Back in 2020, the consensus said this run was long in the tooth. Paulsen said history might have other plans. Six years, one pandemic, and more than 4,000 S&P points later, history is winning the argument. The Super Cycle isn't over. It's entering its final act, and those have rewarded investors who stayed in their seats.

Five Stories Moving the Market:

The United States ​and Canada are working on a deal to ease U.S. tariffs on Canadian steel and autos, as the two governments try to repair relations that soured after trade talks broke down in August – Reuters. (Why you should care – a return to more normal trade between the two nations would lower input costs for U.S. automakers and manufacturers that rely on cross-border supply chains, easing pressure on their profit margins)

Anthropic’s IPO filing showed revenue jumped to $16.3 billion in the first half of 2026, a staggering 1,300% increase from a year ago amid booming demand for its Claude chatbot and coding services; Anthropic spent $8.9 billion on computing power and infrastructure in the six months ended June 30, a 283% increase from the same period a year earlier – Bloomberg. (Why you should care – revenue is finally starting to exceed spending on an adjusted basis)

Bank of America, Blackstone, Citi, and Morgan Stanley will provide $63 ​billion in financing to buy Broadcom's AI chips and networking gear for Anthropic, in ‌one of the largest debt deals tied to the AI buildout; Broadcom is providing "residual value support" to lenders, who are providing first lien loans or senior debt, effectively providing a significant guarantee in the financing structure – Reuters. (Why you should care – Wall Street’s willingness to fund a deal this size shows capital is still flowing into the AI buildout, which supports demand for chips and networking gear)

Pacific Investment Management Co. bought a $4 billion slice of a huge SpaceX financing deal that will help expand its artificial intelligence ambitions; Elon Musk’s SpaceX said it signed a roughly $40 billion equipment-financing deal with lenders led by Apollo Global Management to help facilitate purchases of Nvidia graphics processing units and related equipment – Bloomberg. (Why you should care – Pimco taking a $4 billion slice shows big institutional lenders still want AI exposure, undercutting recent worries that financing for the buildout is drying up, which is good news for Nvidia and other chip suppliers)

Senior officials from Saudi Arabia, Pakistan, and Turkey met with Iranian President Masoud Pezeshkian to try to prevent a new round of hostilities in the war with Iran – FT. (Why you should care – an interim peace deal would ease the risk premium in oil prices, which would take pressure off inflation and bond yields and give stocks some breathing room)

Economic Calendar:

Earnings: V, KO, BA, STX, SPGI, GLW, HCA, WM, FCX, AMT, JCI, UPS, HOOD, SHW, ECL, HLT, ITW, RCL, OKE, NXPI, PCAR, BKR, PYPL, TER, IQV, VTR, EQT, ACGL, CNC, EXR, CNP, FE, FFIV, XYL, PPG, HUBB, CMS, EXE, ESS, CSGP, IVZ, UHS, BXP, PNR

Japan - BOJ Core CPI for September (1 a.m.)

U.S. - Building Permits for September (8 a.m.)

U.S. - ADP Weekly Employment Change (8:15 a.m.)

U.S. - Durable Goods Orders for September (8:30 a.m.)

Canada - Wholesale Sales for September (8:30 a.m.)

U.S. - S&P CoreLogic Case-Shiller Home Price Index for August (9 a.m.)

U.S. - FHFA House Price Index for August (9 a.m.)

U.S. - Conference Board Consumer Confidence for October (10 a.m.)

U.S. - New Home Sales for September (10 a.m.)

U.S. - Richmond Fed Manufacturing Index for October (10 a.m.)

U.S. - Dallas Fed Services Activity for October (10:30 a.m.)

Treasury Auctions $95 Billion in 6-Week Bills (11:30 a.m.)

Treasury Auctions $54 Billion in 52-Week Bills (11:30 a.m.)

Treasury Auctions $70 Billion in 5-Year Notes (1 p.m.)

ECB's Lagarde (President) Speaks (1 p.m.)

U.S. - American Petroleum Institute Crude Oil Inventory Data (4:30 p.m.)

RBA's Bullock (Governor) Speaks (8:15 p.m.)

Australia - CPI for the Third Quarter (8:30 p.m.)