Global Investment Outlook 2021 — Peak Liquidity

2021 was the largest year on record for private capital deployment. Venture funding, buyout activity, and listings all reached levels without precedent, and valuations expanded across nearly every private category.

The temptation is to explain this with sentiment — exuberance, mania, greed. That explanation is emotionally satisfying and analytically weak. The behaviour of 2021 follows from two structural conditions, and once those conditions are stated, most of the year's apparent irrationality becomes rational.

The first was the price of money. Policy rates sat near zero and central banks were still expanding balance sheets in response to the pandemic. When risk-free returns approach zero, the relative appeal of any asset offering a return rises, and the mathematical penalty for cash flows arriving far in the future nearly disappears.

The second was the quantity of money seeking returns. Fiscal and monetary support had put enormous liquidity into the system, and institutional investors faced a genuine problem: meeting return targets in a world where the safe portion of the portfolio yielded almost nothing. Private markets were the obvious destination.

An investor deploying capital in that environment was not being reckless by moving quickly. They were responding to the incentives in front of them. The mistake was not participation — it was the assumption that the conditions creating those incentives would persist.

What abundant capital does to behaviour

When capital is scarce, investors compete on selection: the constraint is finding assets worth owning. When capital is abundant, they compete on access: the constraint is being allowed into the deal. That inversion explains most of what looked unusual in 2021.

Diligence compressed. Processes that historically ran weeks ran days. This was not laziness. In a competitive round, the investor who takes three weeks loses to the one who takes three days. Speed became the product being sold.

Pre-emption replaced process. Rather than waiting for a company to run a round, investors approached earlier and offered terms designed to prevent a competitive process starting at all.

Round sizes and frequency both rose. Companies raised larger amounts, more often, further ahead of need. Where capital is cheap and possibly temporary, taking more than required is a defensible hedge.

Valuation became a means rather than an end. For an investor whose real constraint is deploying a large fund into a limited set of quality assets, paying twenty percent more to secure an allocation is rational if the alternative is not investing at all. Fund economics reward deployment.

When the binding constraint moves from selection to access, paying more is not a failure of discipline. It is what the incentive structure asks for.

The crossover effect

The most structurally significant change in 2021 was who was writing cheques. Hedge funds, mutual funds and other public-market managers moved into late-stage private rounds at scale.

These investors brought a different valuation framework. A traditional venture investor prices a company against other private companies at similar stages, with an eye to ownership and fund construction. A crossover investor prices against public comparables — if a listed peer trades at a given revenue multiple, a private company growing faster arguably deserves a similar or higher one.

That reasoning is internally coherent. Its weakness is that it makes private valuations a derivative of public multiples. When public software multiples compressed in 2022, the anchor holding those private marks in place disappeared. The subsequent repricing was not a change of opinion about the companies; it was the removal of the reference point their prices had been set against.

This is why the 2022 correction hit late-stage hardest and earliest. Late-stage was the part of the private market that had been priced publicly.

SPACs: mechanism, not madness

The special purpose acquisition company boom is remembered as the year's clearest excess. It is more useful to understand what problem it solved.

A traditional IPO requires audited history, predictable metrics, and a process that filters out companies unable to withstand public scrutiny. A SPAC merger allowed a company to reach public markets faster, with forward-looking projections permitted in marketing in a way traditional IPO rules restricted.

For companies that were genuinely ready, this was mostly a routing decision. For companies that were not — pre-revenue businesses in capital-intensive sectors with long paths to commercialisation — it was a route to public markets that would otherwise have been closed.

Two features made the structure fragile. Sponsor economics rewarded completing a deal more than completing a good one, since the sponsor's promote depended on a transaction closing. And redemption rights meant the cash actually delivered at closing was frequently far below the headline trust value, leaving companies public but under-funded.

None of this required abundant liquidity to be conceived. It required abundant liquidity to clear. When that receded, the structure stopped functioning almost immediately.

Where 2021 capital went

Deployment concentrated in a few identifiable themes, each with its own subsequent trajectory.

Software and cloud infrastructure absorbed the largest share, underwritten on the assumption that pandemic-era digital adoption represented a permanent step change rather than a pull-forward of demand. The distinction between those two readings mattered enormously and was hard to test in real time.

Fintech attracted very large rounds, particularly in payments and lending. Lending businesses in particular carried an unhedged exposure to the rate environment that was not always priced.

Consumer and marketplace businesses raised on user-growth metrics at a time when customer acquisition costs were unusually low — partly because competitors' marketing budgets and pandemic behaviour patterns were themselves temporary.

Climate and energy transition began attracting serious institutional capital, in a shift that has proved more durable than most of the year's other themes because it was driven by policy and industrial demand rather than by discount rates.

What separated survivors from casualties

With hindsight, the companies that navigated 2021 and emerged intact shared characteristics that were observable at the time.

They treated the raise as runway, not as validation. A company that raised a large round in 2021 and spent it towards profitability entered 2022 with options. A company that raised the same amount and scaled its cost base to match the new valuation entered 2022 needing to raise again into a closed market.

They kept the burn multiple honest. Capital efficiency was unfashionable in 2021 precisely because it was unnecessary. The companies that maintained it anyway found that the metric became the primary underwriting standard about eighteen months later.

They avoided structure in exchange for headline price. Where a lower clean valuation and a higher structured one were both available, taking the higher number often meant accepting liquidation preferences or ratchets that materially changed outcomes for common shareholders at exit.

They understood the valuation was a market clearing price, not a measurement. A 2021 mark reflected what one investor would pay under one set of conditions. Treating it as an appraisal of intrinsic worth led directly to painful conversations in 2022.

The coordination problem underneath

The most useful way to understand 2021's behaviour is as a coordination problem rather than as a failure of judgement, and the distinction matters because it determines what could have been done differently.

A failure of judgement implies participants believed things that were not true. Correcting it requires better analysis.

A coordination problem implies each participant acted rationally given what others were doing, and the aggregate outcome was poor anyway. Correcting it requires something no individual participant can do alone.

2021 was substantially the second. Consider the position of an investor who correctly believed valuations were elevated:

• Declining to participate meant systematically losing the competitive rounds and being left with the companies nobody else wanted — adverse selection, which is arguably a worse outcome than paying up for a good business.
• Maintaining a slower process meant losing to faster competitors, since as the 2021 US venture report describes, speed was the primary competitive weapon.
• Insisting on better terms meant the same.
• Deploying more slowly meant explaining to LPs why committed capital was sitting idle during the most active market in the industry's history.

Each individually rational response to the environment made the environment worse. More speed meant less diligence. Less diligence meant more capital deployed with less scrutiny. More capital deployed meant higher prices.

What this implies about the available discipline is narrower and more honest than the usual prescription. An investor could not fix the market. What they could do was:

• Size positions to survive the condition ending, rather than to maximise participation while it lasted.
• Prefer clean terms to structured ones, accepting a worse headline price for a better actual position.
• Prefer runway to burn in the companies they backed, since runway is what converts a valuation problem into a survivable one.
• Maintain vintage discipline, deploying consistently rather than accelerating into the most active year.

"Do nothing during a boom" is not implementable. "Deploy at a constant rate regardless of how good the market feels" is, and it is the only reliable defence against a coordination problem you cannot exit.

Reading 2021 correctly in hindsight

There is a hindsight trap in writing about a market peak, and it is worth naming.

It is easy to describe 2021 as obvious folly. That framing implies the correct action was to abstain — but an institutional investor who sat out 2021 entirely would have failed their mandate, missed genuine winners, and had no defensible way of knowing when the conditions would change. "Do nothing during a boom" is not an implementable strategy.

The more useful reading is that the conditions were unusual and identifiable in real time, even though their end date was not. Near-zero rates, extraordinary liquidity, and valuations anchored to public multiples were all observable facts in 2021, not hindsight discoveries. What could not be known was timing.

That distinction — knowing a condition is unusual without knowing when it ends — is the practical problem investing actually poses. The response is not abstention. It is constructing positions that survive the condition ending: preferring runway to burn, clean terms to structured ones, and businesses whose value does not depend entirely on cash flows a decade away.

What was observable at the time

A recurring defence of 2021 decisions is that the conditions were only visible in retrospect. That is largely untrue, and separating what was observable from what was genuinely unknowable is useful for the next occasion.

Observable in 2021, from public data:

• Policy rates were near zero and central bank balance sheets were at record levels. Both published, both weekly.
• Public software revenue multiples were at record levels. The Bessemer Cloud Index charts this publicly and free, and it was the reference private rounds were being priced against.
• Non-traditional investor participation in venture rounds was at a record share. NVCA/PitchBook reports this as a named category in its free quarterly summary.
• Round sizes and valuations by stage were at records, reported by Carta from its own cap-table records, free and quarterly.
• Time between rounds had compressed, which meant companies were raising before they had reached the milestones the prior round had been underwritten against.

Genuinely unknowable in 2021:

• When the conditions would end. Inflation was widely expected to be transitory, and the case for that view was not unreasonable given the supply-chain explanation available at the time.
• How far and how fast rates would rise.
• Which specific companies would prove durable. This is the ordinary uncertainty of the business and is not a 2021 phenomenon.

The distinction matters because it identifies where the discipline was available. An investor could know the conditions were unusual. They could not know the end date. The correct response to that combination is not abstention and not full participation — it is constructing positions that survive the condition ending, which is a different objective from maximising exposure while it lasts.

The most useful single observable was the gap between round frequency and milestone achievement. A company raising twelve months after its last round, at a substantially higher valuation, without having reached the milestones that round was underwritten against, is being repriced by market conditions rather than by progress. That is checkable at the individual transaction level, requires no market view, and was the clearest available signal that price was being set by capital availability rather than by performance.