More and more pension funds are looking at dynamic allocation strategies to cope with the ongoing volatility in financial markets. Assumptions underlying the orthodox static approach to asset allocation are being questioned in an environment where pension funds are forced to be ever more concerned about the volatility in their funding ratio.

Japan’s US$120bn Pension Fund Association has outlined an interesting systematic dynamic asset-allocation strategy with a unique outcome reflecting the unique character of its liabilities. The PFA manages the accumulated contributions of members from disbanded or bankrupt employee pension funds, and for staff who have changed jobs and left their original company pension fund.

Pension funds traditionally determined their asset allocation based on a long-term asset liability study. The only dynamic feature in this framework is the rebalancing mechanism to restore the actual weights of the various asset classes to the policy weight.

Policy weights are reviewed only once every three years or so to reflect the latest state of the economy and the fund’s position in terms of liability profile, asset values and the resulting funding ratio. More recently however, the implicit assumption in this static approach to asset allocation, that the level of risk taken onboard on a daily basis can be determined without regard for the prevailing funding ratio of the pension fund, has come under scrutiny.

Clearly the increased volatility in financial markets and the now widespread practice of marking to market both assets and liabilities have made the funding ratio of pension funds more volatile. At the same time, International Accounting Standards are moving towards recognising these volatile numbers unsmoothed and unfiltered into the sponsor’s financial statements.

Under these circumstances pension funds are forced to consider a more agile and dynamic approach. Deriving the level of investment risks to be taken systematically from the fund’s prevailing funding ratio is one method gaining traction. The higher the funding ratio, the more risk can be taken with the understanding that if markets move against the fund and the funding ratio declines, risk is taken off the table and a minimum level of funding is protected.

These dynamic strategies are inherently focused on down-side risks where traditional asset-allocation methods tend to be based on a mean-variance framework. In a mean-variance approach the standard measure for risk does not distinguish upside swings from downside deviations from the mean.

Due to the mechanism of adding risk when risk is paying off and lowering risk when returns are negative, dynamic strategies work well when markets are trending in one direction or another, whilst they suffer when markets are zigzagging up and down.

They have worked very well over the past few years for several European pension funds: with long interest rates on a persistent trend downward, may funds have seen their funding ratios drop severely with the exception of those who chose to dynamically adjust the maturity profile or allocation to long-term bonds or both to put a floor under their funding ratio.

Similarly equity markets, after an extended period of trending downward as the credit crisis unfolded, since March 2009 turned around to start trending higher and the consistency of these trends proved fertile grounds for systematic dynamic allocation strategies.

Such systematic dynamic allocation strategies are most often discussed in the context of protecting a minimum funding ratio, such as the 105% required by the pension fund regulator in the Netherlands, with the fund’s starting point being one of surplus.

It is therefore interesting to take note of the publication by Japan’s Pension Fund Association where it outlines its systematic dynamic allocation strategy implemented as it aims to recover its solvency from a level of underfunding.

The solution arrived at by PFA does reflect some very unique characteristics of this large fund. Given it looks after already funded and vested benefits only, they do not receive any further contributions nor is there an incentive to create large buffers in the fund in order to limit the cost of contributions in future: its singular aim is to meet its liability obligations to the highest degree of certainty possible and the only tool for PFA to make good on its benefit promises is by investing smartly.

PFA arrived at an asset allocation by funding ratio through a multi-period stochastic optimisation designed to find the allocation with the lowest downside deviation in the funding ratio reflecting its unique liability profile. Given the large size of the fund and the resulting difficulty to move light-footedly from one allocation to another, only five categories of funding ratio and corresponding asset allocation are recognised: on starting at a funding ratio below 100% and rising in 5% increments to a funding ratio above 115%.

The noteworthy outcome of the PFA analysis is that, contrary to the typical dynamic allocation strategy, equity exposure declines as the funding ratio increases. Below 100% a 40/60 split between equity and bonds is optimal, with the equity allocation declining by 5% points as the funding ratio improves to the same degree. By the time the funding ratio exceeds 115%, the equity-to-bond split will have reached 20/80.

Systematic dynamic strategies have sometimes been pointed at as the cause of the Black Monday crash in 1987. Pouring money into the market as they trend higher, but subsequently running for the exit as markets dive exacerbate markets swings, are highly pro-cyclical and cause instability, is the claim.

At least this criticism seems not to hold for PFA, because their strategy of lowering equity exposure as their funding ratio improves will cause them to counterbalance some of the forces coming from more orthodox systematic dynamic investors. •

Oscar Volder CFA, is General Manager, Business Development at BNP Paribas Investment Partners in Japan.

Topics