Latam Daily: Peru Pension Fund Withdrawal … Another One

• Peru: Yearly pension fund withdrawal strikes again!

Mandated pension fund withdrawals have become a yearly event. On Saturday, the government gave its approval to a Congressional Bill allowing for the eighth withdrawal in six years. The whole debate and approval process occurred with enormous alacrity. In hurried sessions, on September 18^th Congress approved the new Law regulating private pension funds. The Executive ratified the Law two days later. The next step is for the Banking Superintendent, SBS, to provide the regulatory framework for the law’s implementation. It has 30 days to do so.

The issue of more immediate macroeconomic interest included in the law is the option for private pension fund affiliates to withdraw up to PEN 21,400 in four monthly parts. But, it was not the only change that was included. The new Pension System Modernization Law (Ley de Modernización del Sistema de Pensiones) also contemplates the following:

• The law extends the option to withdraw 95.5% of funds upon retirement to workers under 40 years.
• The law erases a provision from the previous law that would have required independent workers to begin a compulsory participation in 2028. The new law establishes that this will now be voluntary rather than compulsory.
• The law establishes that affiliates that withdraw their funds will continue to be entitled to the legal minimum pension in the future, which was not true under the previous law.

The new law was always a possibility, but until recently appeared to be relegated to the back burner of Congressional priorities, so much so that until fairly recently there was hope that the issue would not be brought up at all. Instead, the issue was introduced and debated swiftly and the government turned from opposing to favouring the withdrawal practically from one day to another.

Estimates regarding how much funds will be withdrawn run from PEN 26bn to PEN 32bn. In the previous withdrawal in 2024, about PEN 27bn were withdrawn. Assets Under Management (AUM) today total PEN 118bn, just marginally above where it was before last year’s withdrawal, so it seems fair to work under the assumption that a similar PEN 27bn may be withdrawn this time.

Initial impact of the withdrawal

Depending on how long the SBS takes in providing a framework, the most likely date for the first monthly payment to take place would be early November. This should also be the largest payment, as some individual funds will be depleted quickly.

There are four generic ways in which withdrawals have been used in the past: to draw down bank debt, to complement consumption, to augment savings, and to use as a down payment on a home. Based on the withdrawal that took place in 2024, we have estimated how the 2025 funds may be distributed by households in the initial four months following the first withdrawal (table 1).

Nearly two-thirds of savings go into bank accounts in the initial four months. However, they don’t necessarily stay there afterwards. When affiliates withdraw, they receive the funds in a bank savings account. Much of the funds received are maintained in these savings accounts until their owners are ready to dispose of them and distribute the funds to other uses, a process which occurs over time. Typically, the decision to reduce household debt occurs promptly, and the 7.6% debt prepayment estimate is fairly accurate. However, the 23.4% of funds going into consumption is likely to be underestimated over a longer period as drawing upon withdrawn funds for consumption is likely to last for a rather prolonged time. In the meantime, the funds will appear as savings.

Consumption has always risen with withdrawals. The ~20% to be dedicated to household consumption as a first impact seems to be in line with data and consumer surveys linked to previous withdrawals. The lagged impact, as augmented savings spill into greater spending gradually over time, is much more difficult to measure, and gets mixed up with the natural impact of improved household balances (lower debt). However, even if we limit the portion of withdrawal resources used for consumption to 20%, the impact is significant, raising yearly consumption by 0.8pp and yearly GDP by 0.5pp. Given the timing, we expect most of this increase to take place in 2026, although one might expect a particularly frothy 2025 Christmas season as well.

And then there’s the impact on the pension fund system itself, as well as on Peru’s asset markets. This is the eighth withdrawal since 2020, and assets under management have been decimated by these withdrawals. Chart 1 shows the evolution of assets under management before and after withdrawals, together with the possible trend that AUM would have taken, if there had been no withdrawals at all (based on the average long-term yield trend). Note that the no-withdrawal line starts lower than the AUM peak level because of COVID at the time.

The difference is enormous. The current eighth withdrawal of PEN 27bn would potentially lower total AUM to PEN 95bn once the process is over. This compares with nearly PEN 230bn in AUM had none of the eight withdrawals ever taken place. Going forward, affiliates will have to draw their pensions from a pool that is 60% lower than had there been no withdrawals. This is on average. In practice this is likely to mean that most affiliates are likely to have their personal funds largely depleted, and will need to receive the minimum PEN 600 per month pension upon retirement.

The shrinking of the private pension fund system also means that the role that private pension funds have played in local financial markets has become severely impaired, reducing the role that AFPs have had in providing liquidity and breadth to markets. Following the eighth withdrawal, it would take approximately a year and a half for AUMs to return to their pre-withdrawal level. It would take nine years for AUM to return to the level it was at prior to the first withdrawal in 2020. That’s in total AUM terms. In per capita terms, which is what matters for affiliate pensions, the time would be much longer.

—Guillermo Arbe