By Dr. Arturo Cifuentes, Senior Research Associate, Clapes UC, Catholic University of Chile
Original ideas seldom get a second chance to be implemented successfully. The Chilean pension system is one of the few exceptions. To grasp how much is at stake in the coming months—a third opportunity might never happen—we need to go back to its origins and identify the key events that shaped its evolution.
The Chilean pension system was the brainchild of José Piñera, who, as minister of labor and social security, conceived its basic architecture in 1981. The idea was simple: Workers would contribute 10 percent of their earnings to individual accounts managed by private-asset managers known as AFPs (their Spanish acronym). Upon retirement, they could use these funds to either buy an annuity or choose a scheduled-withdrawal alternative. Initially, each AFP offered a single investment fund, subject to guidelines set by the pension regulator. A 2002 reform introduced broader investment options, known as multifondos. In short, five funds—A, B, C, D and E—were designed to offer different risk-return profiles, ranging from Fund A, the most aggressive, to Fund E, the most conservative.
The Chilean model was received with enthusiasm and curiosity by policymakers and academics alike, and it served as a blueprint for pension reforms enacted in countries as dissimilar as Bolivia and Mexico, Kazakhstan and Bulgaria. Not surprisingly, in Latin America, the reception was far warmer among liberals and free-market enthusiasts—who saw it as a mechanism to give individuals greater agency over their own financial futures, free from the state’s reach—than among left-leaning ideologues, who resented placing these funds beyond the government’s control.
By the early 2000s, however, it had become clear that the system was delivering pensions that most Chileans found disappointing—a sentiment that only grew stronger over time and that was, unfortunately, grounded in reality. In recent years, the median replacement rate for self-financed pensions (i.e., excluding any state benefit) has been below 20 percent, compared with the OECD (Organisation for Economic Co-operation and Development) average, which hovers between 50 percent and 63 percent.
What happened? Where did things go wrong? Was it the concept or the execution?
The short answer is that it was a typical case of a sound idea poorly implemented. The longer answer is more complex: We can identify three factors behind this unfortunate outcome.
The short answer is that it was a typical case of a sound idea poorly implemented. The longer answer is more complex: We can identify three factors behind this unfortunate outcome.
Let us start with the mandatory contribution rate. Notwithstanding the power of compound interest, empirical evidence on asset returns indicates that 10 percent is insufficient to achieve an acceptable replacement rate. The consensus points to a figure closer to 15 percent. This is not a recent financial insight; it is straightforward arithmetic, something every government knew, or should have known, from the outset. Yet no government, regardless of its political orientation, has ever attempted to raise the rate. The reason is obvious: Increasing the mandatory contribution meant reducing the take-home paycheck—a politically dangerous decision with clear implications for the next election.
A second factor concerns the labor market’s formality. Roughly one in four Chilean workers is informally employed, and many move in and out of the formal sector throughout their working lives. Since informal workers typically make no pension contributions, this pattern generates gaps in their savings history—known euphemistically as lagunas previsionales. The unhappy result has been that the average Chilean worker contributes to his or her pension account for only about half of his or her working life, with women showing far more severe gaps than men.
Then there is the Régimen de Inversión de los Fondos de Pensiones: a 76-page document issued by the pension regulator that dictates how AFPs must structure their portfolios. Unfortunately, this document rests on some deeply flawed assumptions: that investing outside Chile is inherently risky; that currency diversification is inadvisable; that real-estate exposure should be minimal; and that alternative assets should be severely restricted. Needless to say, AFPs have repeatedly attempted to challenge this framework’s most damaging provisions, with little success. Its defining feature is that it reduces portfolio construction to a rigid set of asset-class limits, without referencing any investment objectives or portfolio-level risk metrics—a design that investment professionals have widely criticized for depressing multifondos’ returns and, by extension, the pensions that depend on them.
An aside: It is worth noting that during the COVID pandemic, the Chilean Congress (National Congress of Chile) passed legislation allowing workers on three separate occasions to withdraw funds from their retirement accounts. This decision, widely condemned by serious policymakers but embraced with gusto by populist politicians, will inevitably take a toll on future pensions. In total, the three withdrawals amounted to roughly US$50 billion, reducing AFP assets under management from 83 percent of Chilean GDP just before the first withdrawal to 62 percent by 2024.
One redeeming aspect of this otherwise regrettable episode was that when workers went to withdraw their funds, the money was there. Given Latin America’s sorry history of governments raiding private savings—for example, Cristina Fernández’s expropriation of some US$29 billion from private pension accounts in Argentina in 2008—this was no small thing. The fact that Chilean workers could access their savings freely and instantly was a validation of the AFP system’s integrity, whatever its other shortcomings.
In any event, by early 2025, the so-called “pension problem”—that is, the widespread dissatisfaction with the pensions most Chileans were actually receiving—had reached a critical point. In a rare consensus encompassing the left, the center and a significant fraction of the right, Congress scrapped the system’s original architecture and replaced it with an entirely new one. The reform was received as a major legislative achievement across virtually all sectors of Chilean society. A sober assessment, however, calls for restraint. In fact, far from resolving the “pension problem”, the new design should give Chileans plenty of reasons to be concerned about their futures.
Why should Chileans be worried? Let us look first at what this reform did. The following three key points deserve special attention.
Why should Chileans be worried? Let us look first at what this reform did. The following three key points deserve special attention.
Mindful of the fact that the 10-percent contribution rate was insufficient—a valid point—the new scheme introduced a 7-percent increase, to be borne entirely by the employer, not the employee. This additional contribution, however, will be phased in gradually, reaching the full 7 percent only in 2033. There is, however, a significant caveat: Only 4.5 percentage points of that 7 percent will flow into the worker’s individual pension account; the remaining 2.5 percentage points will go to finance a set of collective social benefits. Let us put aside for the moment the implication that the gradual phase-in effectively means that Chileans will contribute inadequate amounts to their pensions over the next eight years. A more fundamental concern is that the reform has, in practice, dismantled the pure private-capitalization model and replaced it with a hybrid system that incorporates substantial elements of intergenerational redistribution.
A second element of the reform is the replacement of the multifondos system with a generational-funds scheme. The law mandates a minimum of 10 such funds, though preliminary proposals from the regulator suggest the final number may be closer to 15. Workers will be automatically assigned to a fund based solely on their birth year; neither their individual risk tolerance nor their financial circumstances will factor into the assignment. A peculiar aspect of this change is that the multifondos system was already, in essence, a generational scheme—albeit one with only five options. Under it, workers were expected to migrate gradually from Fund A toward Fund E as they aged, with some latitude to adjust based on personal circumstances. The new setup does not represent a conceptual departure; it simply strips workers of the freedom to tailor their investment vehicles to their individual situations.
A third troubling element concerns the investment regulation that will govern the generational funds—a document that is still in preparation. Its design was awarded through a public auction, with only one valid bidder: Mercer Consulting. The firm has until September 2026 to deliver its non-binding proposal; the regulator is free to disregard any or all of its recommendations. The resulting document will then be opened to a public-comment period—a process that, based on recent experience, offers little comfort. On December 30, 2025, last year’s final workday, the regulator published a major regulatory proposal and gave the market until January 16 to respond: 17 calendar days spanning Christmas, New Year’s Day and the onset of the summer season.
A deeper concern, however, is substantive rather than procedural. A careful reading of the bid documents distributed to prospective consultants leaves little room for optimism: The regulator appears to have drawn no lessons from 24 years of experience with the multifondos. The same structural flaw that undermined that system—attempting to control risk through asset-class limits rather than portfolio-level risk metrics—is once again the template. The bid documents signal that the generational funds will operate under a rigid set of asset-class constraints, with no reference to any portfolio-level risk metrics and with provisions that will likely penalize investments in alternative assets, real estate and non-peso-denominated instruments.
Why does this matter? Because the evidence that asset-class limits are an inadequate tool for managing portfolio risk is not merely theoretical—it is written into the multifondos’ own history. A study published in The Journal of Retirement (Pagnoncelli, Redroban and Cifuentes, Summer 2023), based on the monthly return data of the five multifondos from October 2002 to June 2020, reached a damning conclusion. Employing three rank-order metrics, the authors confirmed that the multifondos’ investment regulations succeeded in creating five funds with different risk profiles—Fund A was indeed riskier than Fund B, Fund B riskier than Fund C, and so on. The problem was that the returns had no coherent relationship to those risk profiles. Investors in riskier funds took on greater risk exposures but were not compensated accordingly; worse yet, over extended periods, the funds’ return ordering was precisely the opposite of what the regulator had intended. The authors traced this outcome directly to the investment policy itself—specifically, to the decision to rely solely on asset-class limits without incorporating any portfolio-level risk constraints. Regrettably, the regulator appears poised to repeat exactly the same mistake with the generational funds.
Taken together, these three concerns point in the same direction. Although celebrated as a historic breakthrough, the 2025 reform has diluted the private-capitalization model, stripped workers of investment choices and appears set to entrench the very regulatory flaws embedded in the multifondosinvestment regime.
Chile’s pension system was given a rare second chance to be implemented correctly. Whether this reform constitutes a genuine improvement—or merely a politically expedient reshuffling of the same structural deficiencies—will ultimately be decided by the investment regime that emerges in the upcoming months. Original ideas seldom get a third chance.
Dr. Arturo Cifuentes is Senior Research Associate at Clapes UC, Pontifical Catholic University of Chile. He served for three years as a member—and later President—of the investment committee for Chile’s Sovereign Wealth Fund (US$25 billion). He also participated in the Financial Regulation Reform Commission appointed by the Chilean Minister of Finance (2010–2011).
Dr. Cifuentes has held academic appointments at Columbia Business School, the University of Southern California School of Engineering, and the University of Chile. As a consultant, he has advised government entities and private-sector firms across the United States, Asia, and Chile.
He has authored numerous academic papers on finance, portfolio management, art markets, applied mathematics, and engineering, while many of his opinion pieces have been published by the Financial Times. Following the subprime financial crisis (2008–2009), he was invited twice to testify before the U.S. Senate as an expert witness.
Dr. Cifuentes holds a PhD in Applied Mechanics and an MS in Civil Engineering from California Institute of Technology.