Thomas Agostino
There is an idea gaining serious traction in capital markets circles, and it deserves more attention from allocators than it is currently receiving: the possibility that artificial intelligence-driven disruption compresses the useful economic life of businesses so dramatically that terminal value—the portion of equity valuation attributable to cash flows beyond Year 5 or 10, begins to erode structurally.
Whether one subscribes to this thesis is beside the point. The more important observation is that the risk of terminal value compression is real, growing, and many institutional portfolios are overwhelmingly exposed. Perhaps less appreciated is that entire categories of investable assets, ones that rarely headline institutional portfolios, are structurally insulated from this risk.
How Much of Your Portfolio Depends on Terminal Value?
Consider what the typical institutional allocation actually owns.
Public equities, which dominate most portfolios, are priced largely on what businesses are expected to earn in Year 10 and beyond. At current S&P 500 multiples of roughly 22 times earnings and technology multiples of 30 to 60 times, the majority of equity value sits not in near-term cash flow, but in the discounted sum of profits that may or may not materialize over the next decade or two.
Private equity operates on a similar assumption. The buyout model depends on building value over a holding period and exiting at a multiple that reflects continued growth. Venture capital is even more explicit: Investors fund pre-revenue companies at billion-dollar valuations precisely because they are buying terminal value. Without it, the model does not work.
Even growth-oriented fixed-income strategies carry implicit terminal value exposure. A corporate bond issued by a company whose business model may not survive the next technology cycle embeds duration risk that is not captured by its credit rating.
Add it up, and the typical institutional portfolio has 60% to 80% of its value resting on the assumption that competitive moats persist, market positions compound, and the cash flows a business generates today will continue, and grow, for decades. That assumption has been reliable since the advent of modern portfolio theory. The question is whether it remains reliable in an era when the cost of disruption is falling and the pace of innovation is accelerating.
Historical Precedent Is Not Comforting
Markets have repriced entire industries when terminal value assumptions broke down, and they did so rationally.
Newspaper companies that traded at 12 to 15 times EBITDA compressed to 3 to 5 times as digital advertising dismantled the print business model. Brick and mortar retailers compressed to 3 to 6 times free cash flow as e-commerce rewired consumer behavior. In both cases, the near-term cash flows were real. What changed was the market’s willingness to extrapolate them forward. Duration risk, not credit risk, was the driver of repricing.
The AI disruption thesis suggests that this pattern, which has historically been applied to one industry at a time, could generalize across broad swaths of the economy simultaneously. Even a partial version of this outcome, where terminal values compress by 30% to 40% rather than disappearing entirely, would represent the most significant structural shift in equity pricing in decades.
Return Streams Exist That Do Not Require a Terminal Value Assumption
This is where the conversation becomes relevant for portfolio construction.
Not all return streams depend on the durability of competitive advantages. Not all of them require a view on what a business will look like in Year 10. Some categories of credit investments generate returns that are contractual, event-driven or tied to physical assets with inelastic demand. These are returns that get collected over months and years, not projected over decades.
Private credit, particularly direct lending, is the clearest example. A senior secured loan to a borrower with tangible collateral, conservative leverage and a two-to-three-year maturity generates contractual cash yield that is indifferent to whether AI disrupts the borrower’s industry in Year 7. The investor is not buying the borrower’s terminal value. The investor is lending against current assets and current cash flow with structural protections that exist independently of the borrower’s long-term competitive position.
Special situations and litigation finance operate on a similar principle. Returns are driven by specific legal or event-driven catalysts: a court ruling, an insurance recovery, a restructuring resolution. These outcomes are binary and idiosyncratic. They have no relationship to technology cycles, equity multiples or the durability of business models.
Opportunistic credit, deployed into stressed or dislocated sovereign and corporate situations with identifiable catalysts, is priced on the gap between current market value and near-term recoverable value. It is, by construction, a short-duration bet on a specific repricing event, rather than a long-duration bet on terminal growth.
The Economies Built on Things AI Can’t Disrupt
There is a dimension of this framework that most allocators overlook: public credit in emerging markets.
The conversation related to terminal value compression is, at its core, a conversation about the vulnerability of businesses built on intangible advantages. Software moats, network effects, brand premiums, data advantages are the competitive positions most susceptible to disruption because they exist in the digital domain where AI operates most powerfully.
In our opinion, emerging market economies are fundamentally different in character. The sovereign and corporate credit universe in emerging markets is dominated by physical assets and tangible economic activity: energy infrastructure, commodity production, transportation networks, agricultural exports and basic industrial capacity. A hard currency bond issued by a state-owned oil company, a sovereign with revenues tied to copper exports, or a corporate borrower operating toll roads and ports are all backed by economic activity that even a better, large language model cannot unbundle overnight.
This is not a claim that emerging market credit is immune to risk. It is a structural observation about what type of risk it carries. An EM sovereign bond does not depend on the issuer maintaining a competitive moat for 15 years. It depends on the country’s willingness and ability to service its debt over a defined maturity, backed by revenues from physical economic activity with inelastic global demand. A bond matures at par. The investor is buying current yield and a contractual claim, not a discounted stream of speculative future cash flows.
The yield premiums available in EM public credit reflect complexity, access barriers and perceived sovereign risk, rather than terminal value assumptions. For an allocator concerned about the durability of growth-based equity valuations in developed markets, EM credit offers something structurally distinct: income derived from the physical economy, at a premium, with no embedded bet on perpetual competitive advantage.
The Multi-Asset Credit Framework as a Structural Hedge
A portfolio that combines these elements (public performing credit tied to physical economy issuers, private lending secured against tangible collateral, event-driven special situations with binary catalysts and opportunistic credit priced on near-term recoverable value) is not merely diversified in the conventional sense. It is structurally insulated from the specific risk that the terminal value thesis describes.
Each component generates returns on its own timeline and through its own mechanism. None require the investor to hold a view on whether today’s market leaders will still be dominant in a decade. None of them depend on the persistence of intangible competitive moats. The return sources are contractual, physical, legal and event-driven, rather than growth-dependent and terminal-value-reliant.
This does not mean such a framework is without risk. Credit risk, sovereign risk, liquidity risk and execution risk all exist and must be managed. But the category of risk that dominates the current valuation debate—the risk that long-duration growth assumptions prove too optimistic because disruption shortens the competitive life of businesses—is simply not present.
What Allocators Should Be Asking
The debate about AI and terminal value may resolve in any number of ways. Perhaps disruption accelerates to the point where equity markets undergo a fundamental repricing. Perhaps the thesis is self-defeating: If terminal values compress, the capital that funds AI development dries up, disruption slows, and moats become durable again. Perhaps the truth lies somewhere in between, with structurally higher equity risk premiums and more volatile cycles based on shorter-lived competitive advantages.
Regardless of which scenario unfolds, the portfolio construction question is the same: How much of your return depends on assumptions about the distant future, and how much is generated by what exists today?
A multi-asset-credit framework, particularly one rooted in emerging markets where the underlying economy is physical rather than digital, does not need terminal value to justify itself. The returns are already being generated through contractual yields, tangible collateral, legal catalysts and enduring global demand for physical goods and infrastructure.
The question is not whether terminal values will collapse. The question is whether your portfolio was built to perform if they do.
Thomas Agostino is a managing director at Gramercy Funds Management and a portfolio manager of the Gramercy Multi-Asset Strategy.
This feature is to provide general information only, does not constitute legal or tax advice, and cannot be used or substituted for legal or tax advice. Any opinions of the author do not necessarily reflect the stance of ISS STOXX or its affiliates.
Tags: Artificial Intelligence, disruption, Emerging Markets, market risk, Portfolio Construction, Risk