Abstract Pension Savings in Israel – Is there Cause for Concern?

A session summary from the 2012 Caesarea Economic Policy Planning Forum

On June 27–28, 2012, IDI and the College of Management Academic Studies convened The 20th Caesarea Economic Policy Planning Forum, which was called The Eli Hurvitz Conference for Economics and Society. Following is a summary of the recommendations of the working group on Pension Savings in Israel. The working group was chaired by Prof. Eytan Sheshinski, Department of Economics, Hebrew University of Jerusalem.

Over the past twenty years, the pension system in Israel has undergone numerous reforms, which have led to substantial changes in the structure of citizens’ long-term savings and to the creation of a new pension system, which is still being developed.

The pension system has shifted from what was largely a defined-benefits (DB) system, in the (pre-1995) “old” pension funds and unfunded pension plans, to a defined-contributions (DC) system, based on cumulative savings, in the “new” pension funds, provident funds, and life insurance plans. In addition, the investment structure was modified and an actuarial balance was computed for the old pension funds, the retirement age was raised for both men and women to bring it in line with longer life expectancies, mandatory pension was introduced for all salaried employees, and greater emphasis was placed on saving for a pension paid in monthly installments as opposed to one distributed in a lump sum.

As part of the reforms, the government reduced the share of designated bonds in old and new pension funds. This process was coupled with increased investment by the funds in such assets as stocks and corporate bonds. Between 2000 and 2011, the proportion of investment by pension funds in stocks rose from 2.9% to 25.9%.

Initially, the shift to investing in stocks allowed the pension funds to maintain high yields and low volatility, since, from 2003 to 2007, the capital market experienced continuous gains. During this period, the new pension funds achieved an average nominal yield of 7–11% percent annum. But in 2008, the global financial crisis struck, toppling stock markets around the world and bringing interest rates down to nearly 0%. That same year, the pension funds, life insurance plans, and provident funds lost 10–30% of their value. Although the world’s stock markets recovered by 2009, and most of the pension and provident funds recouped their losses, the interest environment remained at a low level of between 0 and 3%, as did the yield for “no-risk” assets, even for long-term investments of 20–30 years.

The financial crisis has clarified the existing risks of pension savings. On the one hand, the low interest does not allow pension funds and insurance companies to achieve sufficiently high yields via low-risk savings vehicles to enable employees to increase their accumulated savings; on the other hand, investing in stocks and corporate bonds, which have the potential for high yields, exposes the saver to volatility and high risk (in comparison with such low-risk assets as government bonds and term deposits).

These processes led the members of the working group to address various plans and alternatives that could increase the level of certainty for employees, reduce volatility, and boost pension savings. The members elected to seek alternatives that would not entail significant change in the level of state involvement and would not increase the government’s financial liabilities.

Accordingly, the working group chose to focus on three areas that complement one another (and are not mutually exclusive):

  1. Method of allocation of designated bonds
  2. HACHAM model (life cycle–based savings plans)
  3. Dual role of severance pay—as compensation and as an important component of pension savings

1. Method of allocation of designated bonds

Designated bonds are non-negotiable, index-linked government bonds with a fixed interest rate that are issued solely for pension funds and constitute 30% of pension savings. The current interest rate on these bonds is 4.86%, which is higher than the average long-term market rate. Since it is the state that bears responsibility for the interest payments, this interest is, in effect, a subsidy given by it to the pension funds.

In the context of this paper, the working group chose to discuss the current method of allocating designated bonds, with the aim of achieving more efficient allocation that would maximize the advantages of this instrument.

  • The participants looked at four alternatives, examining the allocation of designated bonds for all pension schemes based on monthly allocations: pension funds, life insurance, and provident funds.

  • The participants proposed examining the possibility of increasing the number of designated bonds by lowering the guaranteed yield so that the budgetary cost to the government would not be affected.

    • This expansion is expected to have an impact on the capital market and on government financing methods, but these were not considered as part of this paper.
    • In particular, it should be ascertained that this move will not lead to an increase in the state’s bond inventory and debt-raising needs beyond the debt necessary to finance the deficit (as stipulated in the fiscal rules), and will not undermine the debt management policy of the state.
    • It should also be ascertained that expanding the non-negotiable bonds market will not cause severe harm to the negotiable bonds market, in terms of liquidity and level of negotiability.

The following are the alternatives that were considered:

  • Alternatives A and B propose allocating designated bonds based on the age of the employee so as to increase certainty in the years leading up to, and following, retirement age.

    • Alternative A would allocate designated bonds starting from age 60, at a proportion of 55% of the savings portfolio and with a real yield of 4.86%.
    • Alternative B would allocate designated bonds starting from age 40, using a graduated model, with a reduced yield of 4.2%, so that the proportion of the bonds would increase with the age of the investor.

  • Alternatives C and D would allocate 70% of the savings to designated bonds up to an income ceiling, so as to create a new basic-pension layer, between the layer of Israeli social security and that invested in the free market.

    • Alternative C limits the allocation of designated bonds to an income ceiling of NIS 1,500, with a real yield of 4.86%.
    • Alternative D limits the allocation of designated bonds to an income ceiling of NIS 3,000, with a real yield of 4.0%.

In the paper, the members of the working group present a simulation that compares the various alternatives based on three wage levels: NIS 4,000, 8,000, and 16,000.

  • An analysis of the results shows that increasing the proportion of designated bonds at the expense of lowering the guaranteed yield has a positive impact, in terms of lessening volatility and raising the yield for pension savings (based on the assumed yields of the simulation).

  • It was found that Alternative D is preferable for intermediate wage levels and below, while Alternative B is more suited to wage levels that are intermediate and above.

  • The simulation examined the alternatives in terms of the period up to retirement age. Thus the advantage of Alternatives A and B—in which emphasis is placed on the stability of pension payments during retirement itself—is not reflected in the comparison.

  • The members did not examine the effect of expanding the proportion of designated bonds on government methods of financing or on the capital market.

  • In light of these points, the members did not reach a consensus regarding a recommended alternative. Several members of the working group supported Alternative D, others favored Alternative B, and the remainder felt that, due to the limitations noted above, it would not be appropriate to formulate any recommendation at this time.

2. HACHAM model (life cycle–based savings plans)

Under the HACHAM model, the risk level of the investment of pension savings is affected by the age of the employee such that the level of risk is reduced as he or she approaches retirement age. In this way, young employees with a long investment horizon can invest in pension products that have a high risk level and high potential yield, while investors with a short investment horizon will invest in conservative products and avoid volatility in their pension savings.

  • It emerges from the findings of the working group that, compared to a portfolio with static management, the HACHAM model makes it possible to reduce the risk level of the older investor without diminishing yield (and at times, even raising it) and without increasing the overall risk level of the savings plan.

  • The members also based themselves on publications of the OECD, which supports the adoption of strategies of this type, particularly in countries like Israel, where pension savings are mandatory and defined-contribution pensions are the norm.

  • Accordingly, the members of the working group support the implementation of the HACHAM model as the default option for pension savings in Israel.

3. Dual Role of Severance Pay

Severance pay is a form of compensation to an individual who has been dismissed from his workplace, as well as a means of helping him to maintain an income during the transition between jobs. Taken together, these monies constitute over 30% of pension savings. In recent years, we are witnessing a growing trend whereby investors withdraw their severance pay before reaching retirement age. There are several reasons for this tendency, among them: changes in the labor market and frequent movement between jobs, shortsightedness of investors, and taxation reforms that force employees to decide on the future of their compensation monies immediately upon their dismissal.

The members of the working group did not reach a consensus on the way to address this issue, but they did agree that suitable changes must be implemented so that the bulk of the severance pay is channeled into savings for retirement age.