Crosscurrents | The 40-Year Vintage, Stop Waiting for Godot, Think in Circles

Sanjeev Krishnan | October 6, 2026

The 40-Year Vintage — Why 1980 – 2020 Is the Exception in 300 Years of Returns, and How It Shaped Our View of Risk

Your ability to buy low and sell high is mostly an accident of timing. Most of the time, it has nothing to do with you.

Harvard Business School professors Josh Lerner and Paul Gompers empirically demonstrated the first half of this in their paper ​“Money Chasing Deals?”: the capital flowing into a vintage predicted the price paid. The year you committed mattered more than what you picked, and the same result holds far beyond private equity. A bond fund’s return is mostly set by the yield on the day it was bought. Equities bought in 1982 and equities bought in 1999 produced two entirely different decades.

I call this the conditions inherited. Every investment starts with a set of five conditions beyond the investor’s control: the level and direction of rates, the stability of inflation, the volatility of growth, the price of energy, and the shape of the world order. The investor chooses the asset, and I believe the era (i.e., the vintage year) chooses most of the return. 

Looking across 300 years of data, I see 1980 to 2020 as the strongest investment vintage on record. Not because any one condition was unusually good, but because all five conditions favored investors, moved in the same direction, and stayed that way for a working lifetime.

  • Rates: Over three centuries, nominal rates mostly oscillated within a band, and the direction was unpredictable. From 1981 to 2020, the 10-year Treasury went from roughly 15% to 0.5% in nearly a straight line. 40 years of falling rates repriced most long-duration assets upward, and the multiple expansion was largely attributed to skill.
     
  • Inflation: The historical norm I’ve observed is episodic and unstable, with a decade of relative calm, followed by inflation of 10 – 30% amid war or a failed harvest. From 1980 to 2020, U.S. inflation averaged about 3%, and after the early-1980s inflation shock subsided, it remained in a relatively narrow range for much of the following three decades. The China shock and integration of more than a billion new workers in the global labor force also helped keep it stable.
     
  • Geopolitics: From 1991 through roughly the mid-2010s, the U.S. enjoyed a historically unusual degree of military, monetary, and institutional primacy, which is the clearest unipolar moment in the modern state system. Great-power warfare, already in long-term decline, remained at exceptionally low levels during this period.
     
  • Resources: Resource nationalism has historically been the rule, while food shocks have recurred every decade or two. Energy expenditures accounted for roughly 8 – 14% of U.S. GDP through the 1970s and early 1980s. By 2019, that share was under 6%, and in 2020, under 5%, the lowest on record. The era was defined by cheap molecules, cheap calories, and open trade.
     
  • Technology: From what I’ve observed, general-purpose technologies arrive every 40 – 60 years, often accompanied by a bubble and crash, then take two to four decades to show up in productivity. U.S. labor productivity growth averaged roughly 1.5% from 1973 to 1995, then accelerated to about 2.5% from 1995 to 2000, with Federal Reserve research attributing much of that acceleration to information technology.

Each of these conditions has appeared before, but never all five at once, moving in the same direction for 40 years. That is why I think of 1980 to 2020 as a single extended vintage, and why the period stands apart in modern economic history.

The returns show what the conjunction was worth. From 1941 to 1980, the 10-year Treasury lost 1.9% per year in real terms. From 1981 to 2020, it earned 4.7% per year in real terms. Stocks went from 5.3% real per year in the half-century before 1980 to 8.4% per year in the four decades after. A 60⁄40 portfolio nearly doubled its real return, from 3.8% to 7.2% a year, with a fifth less volatility. Re-rating also contributed materially to the equity return: the Shiller CAPE went from 7.4 in 1982 to 31 in 2020. I don’t think the next 30 years can repeat that trick without pushing the multiple toward triple digits

Consider what that vintage built. Almost every tool we use to define risk was calibrated or widely applied within it. The 60⁄40 portfolio worked because stocks and bonds hedged each other, which they only reliably did for the back half of the period. Sharpe ratios, value-at-risk, and mean-variance optimization were applied to an unusually tame volatility régime. Stock and Watson, who coined the term ​“Great Moderation,” found that U.S. output volatility fell by roughly 40% after 1984 and attributed most of the decline to luck rather than policy. Private equity’s leverage model assumed the cost of debt would fall over the holding period, and Bain estimates that nearly half of the industry’s value creation from 2010 – 2017 was due to multiple expansion. Falling risk-free rates put downward pressure on DCF discount rates for four decades, until in December 2020, $18 trillion of bonds, more than a quarter of the world’s investment-grade debt, carried a negative yield. 

The terminal point of the vintage was a risk-free rate that guaranteed a loss, and every model built off that curve inherited it. Investors’ understanding of risk was shaped by that exception and then embedded into models, incentives, and governance Some call this the risk miracle. 

A miracle is the right word, as it’s a stark exception to 300 years of return data.

The inherited conditions over the next 30 years look far more like the 300-year norm than the 40-year exception. That means rates with no reliable direction, episodic inflation, several powers setting rules, contested resources, and a new general-purpose technology in its bubble phase. That is closer to the ordinary state of the world, though a system engineered for the exception may read it as a crisis.

Now let’s look at the first years outside the vintage. In 2022, the Fed raised rates by 425 basis points, the most in a calendar year since 1980. The 10-year Treasury had its worst year in 95 years of records. The 60⁄40 portfolio had its worst year since 1937. And the stock-bond correlation turned positive for the first time since 2000, which is to say for the first time in the working life of almost everyone running money today.

As my finance professor used to say, most humans are incentivized to buy high and sell higher. Buying low requires seeing through a volatile inheritance, and most governance structures will not let you do it, because they were built to reward the behaviors that worked in the 40-year vintage.

The next 30 years will test whether we can read the conditions we inherit and redefine risk and reward for a world where volatility is the norm.

Don’t buy high and sell low, assuming the last 40 years. 

Stop Waiting for Godot, Start Reordering the System 

Energy investors have spent 50 years waiting for peak oil. It will not be the binding constraint. Oil, gas, and coal still supply more than 80% of global primary energy, and in 2025, all three set consumption records. Despite all of the geopolitical and price volatility, these commodities are not disappearing. The scarcity is elsewhere: the atmosphere’s capacity to absorb exhaust. I call it Peak 33.

33°C — the setting we built on

A planet absorbing Earth’s share of sunlight should average about −18°C. Earth averages about 15°C. That 33°C gap is the greenhouse effect. It is the only setting civilization has ever known, and every crop calendar, building code, actuarial table, and DCF terminal value assumes it.

For 800,000 years, nature kept the emitters and drains in balance. Volcanoes and wetlands put greenhouse gases in, and oceans and forests take them out. Peak 33 is the point at which the drains can’t keep up and the setting starts to move.

The numbers show that we are there:

  • CO2 has risen from about 280 ppm to more than 425 ppm, the highest level in the ice-core record.
     
  • Fossil fuels emit about 38 billion tonnes of CO2 a year, and oceans and land absorb roughly half of it.
     
  • 2024 was the first calendar year to average more than 1.5°C above pre-industrial levels.
     
  • The remaining 1.5°C budget is about 170 billion tonnes from the start of 2026. At current emissions, that is four years.
     
  • The drains are also weakening. The land sink faltered in 2023 as the Amazon dried and Canada burned.

Why I believe markets miss it: sources price themselves, sinks don’t

Every scarcity that has shaped markets was a source, including land, timber, coal, oil, and chips. A source can be owned, so when it gets scarce, its price rises and capital moves.

A sink can’t do that. The atmosphere is a commons. No one collects rent on its remaining capacity, so its scarcity never shows up as a price. Only about 28% of global emissions are subject to any carbon price, and less than 1% are subject to one consistent with the Paris Agreement.

So the cost arrives as damage instead: missed crop yields, repriced reinsurance, utilities that can’t cover their wildfire liabilities, flooded ports, and summers that break the grid. Markets built to read prices under-read a constraint that has none. This mispricing can persist for a long time.

Where the constraint shows up first

1. On a balance sheet that has to put a number on damage. Insured catastrophe losses have topped $100 billion for six straight years through 2025. The Los Angeles fires alone cost $40 billion, the most expensive wildfire in history. Treasury’s ZIP code study found premiums were 82% higher in the most climate-exposed areas. Florida’s carrier insolvencies show how the sink constraint reaches underwriters, and carriers that can price the risk accurately inherit the market. Water rights, crop insurance, grid reliability, and the cost of capital for long-lived fossil assets follow the same pattern.

2. Where physics wins even without any price on the sink. ERCOT has no carbon price and a legislature that has repeatedly tried to restrict wind and solar. Yet in summer 2025:

  • Solar out-supplied coal.
     
  • Batteries, which barely existed on the grid in 2021, passed 15 GW.
     
  • Wind and solar met roughly 90% of load growth.

Nobody in Austin asked for that outcome. Solar modules went from about $100 a watt in 1976 to cents per watt today, and battery packs are 93% cheaper than in 2010. The molecule’s cost didn’t fall.

The point: assume no price on the sink and build a new stack

The usual framing gets this wrong. Texas re-ranked gas rather than retiring it. Solar carries the day, batteries carry the evening ramp, and gas carries the cloudy week when batteries are drawn down. The molecules moved from baseload to backup rather than disappearing.

That is hybrid physics. Physics-based technology will increasingly do more work as it moves down a falling cost curve, and hydrocarbons could become the reserve layer of the same stack rather than its competitor.

This is also how environmental physics gets addressed without waiting for a carbon price. The sink doesn’t care whether a molecule exists, only whether it is burned. Every hour the physics layer carries the load is an hour the drains aren’t asked to absorb. Texas is buying those hours on cost, not conscience.

My take on what investors should underwrite

Stop pricing energy as a volume race to own the last barrel. In a hybrid stack, a molecule’s value shifts from energy sold to reliability provided. The spread goes to whoever integrates the stack and whoever underwrites the residual risk.

That leaves two questions:

  1. Where does the unowned cost first land on a balance sheet, and who can price it?
     
  2. How fast can load shift onto the physics layer, and who collects the spread as molecules move into reserve?

Peak oil was a source story, but Peak 33 is a sink story. Responding to it means reordering the system rather than pursuing wholesale replacement.

Think in Circles, Not in Lines

Imagine five investors looking at the same AI build-out in 2022 and seeing five different things.

  1. The momentum investor bought NVIDIA on a reinforcing loop. Better models need more compute, and more compute makes better models. They were spectacularly right…NVIDIA went from a market value of $360 billion at the end of 2022 to $5.7 trillion this month. 
     
  2. The value investor, worried about capex and negative cash flow, called for the correction and is still waiting. 
     
  3. The trader rode belief. Every earnings call that mentioned AI capex lifted the stock, which justified more capex, which lifted the stock again. 
     
  4. The distressed investor was watching the lag. The hyperscalers will spend roughly $700 billion on capex in 2026, roughly 80% more than they spent in 2025, and that capex now absorbs 80% of the big three’s operating cash flow, up from 42% in 2023, so the thesis is that overbuilt compute will arrive late and some of it in trouble. 
     
  5. The real-asset investor wanted the substation, the transformer order book, and the interconnection queue, where 2,061 gigawatts of generation and storage projects are waiting, more than the entire installed U.S. fleet, with a median wait of five years, and where a large transformer now takes three years to deliver, up from less than one.

Five investors, five partial truths. The systems investor saw all five at once and bought the powered land next to the substation and the fiber. Land with the right power, fiber, and entitlements can now trade for hundreds of thousands to millions of dollars an acre. Amazon paid $3.7 million an acre in Virginia last November.

The first two reflections above argued that we have left a 40-year era in which the macro tide lifted every asset and entered one in which the binding constraints are volatile and, in the case of the atmosphere, unowned. 

In that world, problems and opportunities often live in the loops between companies. A grid, an insurance market, a watershed, or a supply chain does not improve because one participant behaves well. It improves when the feedback that governs all the participants changes. A single company’s performance is routinely absorbed, offset, or arbitraged by the system around it. The era of moderation hid that. The era of volatility exposes it.

A business is a function. Capital and inputs go in one end, product and return come out the other. A system is a set of loops. The result of an action comes back around as the cause of the next one, and the cycle starts again.

Much of what appears complex in a market stems from a small number of mechanisms. Part of the output returns as input, and that feedback can only do a few things. It can amplify itself. It can correct itself. It can run through people’s beliefs, changing what the facts become. It can arrive late. And it can hit a wall. Five mechanisms: Reinforcing, Balancing, Reflexive, Delayed, and Constrained. 

Every market you have ever traded is some combination of the five. Most investing styles privilege one mechanism over the others.

Momentum investing is a Reinforcing strategy built on buying what is getting bigger because it is getting bigger. Value investors look for the Balancing mechanism, waiting for a correction before buying below the level. Sentiment trading is Reflexive, seeking to anticipate what people are about to believe. Distressed investors look for the Delay, then buy at the bottom when the lag resolves. Real-asset and infrastructure investing is Constrained, centered on owning the scarce thing everyone has to pass through.

Each style sees one mechanism clearly and can lose money when another takes over. The momentum investor gets killed by the correction, the value investor by the reflexive rally he refused to believe in, the trader by the lag, the distressed investor by the wall he never saw. To me, style rotation is one mechanism handing the market to another. In 2022, value beat growth by 22 points, and in 2023, growth beat value by 31 points. In the three months after the March 2009 low, the stocks’ momentum strategies were short, and those they owned rose 163%, while the prior winners they owned rose only 8%. And over the twenty years ending in 2025, which is long enough for all five mechanisms to show up, 93% of large-cap U.S. managers trail the index.

A real system runs on all five, all the time, and a strategy built on one of them is likely wrong four-fifths of the time and hopes its fifth comes around often enough.

Systems investing is the discipline of looking for all five at once, and I think the clearest institutional example is not a climate fund. It is Apollo.

Traditional private equity is a line. Capital comes in, is allocated, returns are distributed, the fund is done, and you raise the next one from zero. That line now stalls as money waits to return to investors. Bain reports the average buyout fund now takes 20 months to raise, up from 11 months before the pandemic. Distributions have run below 15% of NAV for four straight years, an industry record, and 32,000 unsold companies hold $3.8 trillion of unrealized value. 

Athene, Apollo’s insurance arm, is a loop that runs all five mechanisms simultaneously. 

  • The Reinforcing mechanism is the engine. Retained spread grows equity, equity supports more annuities, annuities fund more loans, and that cycle has compounded Athene to more than $400 billion of perpetual-capital AUM within Apollo. 
     
  • The Balancing mechanism is actuarial liabilities, rating agencies, and regulators pushing back when the search for yield gets aggressive. 
     
  • The Reflexive mechanism is the market’s belief that private investment-grade credit is safe, which draws other insurers into the same structure and, so far, has confirmed that belief. 
     
  • The Delayed mechanism is the one to watch. I believe a 2024 origination will not reveal itself as good or bad until 2028, and the whole structure is a bet that the Balancing mechanism is working during the lag. 
     
  • The Constrained mechanism is origination capacity, which is why Apollo built and bought a stable of origination platforms rather than waiting in line for someone else’s.

The result of running all five is that Apollo stopped being a fund manager that re-raises from zero and became a platform with what Apollo calls perpetual capital. Athene now holds roughly 40% of Apollo’s trillion-dollar asset base. Perpetual capital is about 70% of its fee-generating assets, and in 2025, Apollo earned more from spread ($3.4 billion) than from fees ($2.5 billion). More than half of its core earnings came from the balance sheet, not from managing other people’s funds, and that spread funds the next turn without a fundraising cycle. 

The market is not yet convinced. Apollo trades at about 12 times earnings, compared with roughly 18 or more for Blackstone, and is down this year amid private credit fears. That is a Reflexive mechanism too, and which way it resolves depends on the Delay. Whether the Delay eventually exposes the Balancing mechanism as weaker than advertised is a real question. But it is a systems question, and it will be answered by a systems investor, not by a value investor waiting for a P/E to revert.

I would argue that the systems (loop) investment opportunity does not depend on calling it impact or choosing between return and purpose. A system that leaks value, whether as an unpriced externality or an unowned constraint, is a system in which someone will eventually stand where the output comes back around and collect it. The line-thinkers cannot see where that is, because a line has no place for the output to return to. The loop-thinkers can.

Think in circles, not lines. Next week, I’ll look at where the leaks are and how to build the toll. Change systems, think in loops, and create alpha. 

—

The views and opinions expressed are those of the S2G team members identified and do not necessarily reflect the views of S2G. This content is for informational purposes only, should not be taken as legal, business, tax or investment advice, or be used to evaluate any investment or security, and is not directed at any investor or potential investor in any investment vehicle sponsored by S2G. Investing involves risk, including the risk of loss. Specific companies mentioned in this podcast are for educational purposes and should not be construed as an endorsement of any kind. Please note that S2G may maintain investments in some of the companies discussed on this podcast. Any past performance discussed is not indicative of future results. For more important information, please see s2gin​vest​ments​.com/​d​i​s​c​l​o​sures.