Rethinking Asset Allocation

The Second Side My colleagues Tom Idzorek and David Blanchett have drafted a paper called, "LDI Misapplied: Income Portfolios and Liability-Driven Investing." That title probably doesn't mean much to you. (If it does, please let me know, and I will use you as a consultant for future columns.) Its subject, however, is straightforward: asset allocation for retirement portfolios.

Conventional asset allocation addresses what investors own. In contrast, liability-driven investing, or LDI for short, considers both sides of the balance sheet. The assets exist for a reason, after all--to offset future expenditures. With LDI, the investor estimates those expenditures as thoroughly as possible and then attempts to hold assets that will generate cash that will pay them down. LDI is a matching exercise: It pairs assets with liabilities.

LDI got its start with defined-benefit plans. Because the plans’ future payments were specified and the aggregate life span of a group of retirees is easier to forecast than that of one person, pension managers knew almost exactly how much money they would need to disburse and when. If they could find assets that would reliably generate such cash at such times, then their plans’ needs would be addressed. Their plans would, as the parlance went, be “immunized.” (Even now, after all these years, I can’t write that word without envisioning a needle.)

The Appeal of Bonds The solution was obvious: Treasury bonds. If the plan did not feature cost-of-living increases for its payouts, as was generally the case with corporate pensions, then the Treasuries could be mapped dollar-for-dollar to the benefits. Plans with COLA features were trickier before the invention of Treasury Inflation-Protected Securities, but, barring dramatic changes in the inflation rate--which have not occurred in recent decades--they could effectively be immunized. Then, if the plan had excess assets, besides those that were pledged to pay its debts, management could invest as it wished--which generally meant aggressively.

In summary, moving from conventional asset allocation to LDI meant that a pension plan would likely--

1) Adopt a more-conservative asset mix, by owning more bonds;

2) Prefer long to short bonds (to match the duration of the obligations, most of which were due well into the future);

3) Invest any surplus assets aggressively.

Lost in Translation In recent years, some practitioners have advised using the LDI approach for retail portfolios for retirement planning. After all, what applies to a pension plan also applies to a single employee. The pension plan must meet future obligations, and it can do so with greater precision if it structures its assets so that they have matching payouts. The individual, similarly, needs a stream of cash flows with which to pay retirement costs and would very much like more rather than less certainty when receiving those checks. So, why "optimize" assets without considering the investor's liabilities? Why not follow the LDI approach, so that retirees can enjoy the financial security that they crave?

Well, yes, say Morningstar’s authors--but only to a point. They argue that most who have used the LDI technique with retail investors have done so too literally. Although there are some similarities between the situations of pensions and an investor’s retirement plan, there are also two major differences--

1) Most pensions have very little risk capacity. If they promise more benefit than their investment portfolios can deliver, then they must defray the shortfall with corporate monies. That is bad for the stock price at the best ... and ruinous at the worst. In contrast, individual investors have greater flexibility. Should their portfolios perform badly, they may be able to work longer (if they have not yet retired) or may be able to avail themselves of other assets.

For example, a married couple who owns a fully paid house and expects to stay in that house during the first decade of retirement could accelerate that timing if their investments disappoint. They could move several years earlier than planned, thereby increasing the size of their financial portfolio. Disappointing? Sure. But they would be happier than the CEO of the company that declared Chapter 11 because it couldn’t fund its pension.

2) Retirees don’t have fixed liabilities. If push comes to shove, they can postpone that trip, keep the old furniture. They won’t be sued because their portfolio doesn’t deliver the cash amount that was promised, to the penny. Adhering to a fixed, guaranteed payment schedule is a desirable attribute for retirees but a legal duty, enforced by strict regulations, for pensions.

As a result, state the authors, the LDI portfolio can look very different when created for individuals rather than a pension fund. Whereas pensions that use LDIs invest largely in bonds (unless the pension has a substantial surplus, in which case it may invest more aggressively), the retiree who adopts an LDI approach might conceivably place half or more of his assets into stocks. For pension funds, LDI is often a driving principle; it determines the portfolio. For individual investors, on the other hand, LDI is a tool. It is one input among several.

The Upshot What this means in practice will vary depending both on the investor's circumstances and on the forecasts of expected asset-class returns. In very general terms, though, supplanting the traditional asset-allocation approach will lead to--

1) Substituting long bonds (either fixed or inflation-protected) for short/intermediate bonds. Shorter-duration securities please optimizers, and thus asset-only allocations, because of their lower volatilities. But LDI cares not about standard deviations; what it monitors is how assets fluctuate in response to liabilities, and long-term bonds make for the closer fit.

2) Bigger changes for conservative portfolios than for aggressive ones for the simple reason that conservative portfolios own more bonds, where the largest effects occur.

3) Smaller international allocations. Again, asset-only approaches reward lower volatility, which diversifying internationally can provide. And again, LDI cares not. All things being equal, it prefers domestic securities because the dollars provided by the assets match the currency of most of the liabilities.

The best asset-allocation practice for individual investors is neither to follow the traditional, asset-only approach, nor to switch completely to liability-driven investing. Both views should be used, as assets exist to meet liabilities. However, as the nature of individual liabilities differs from that of pension liabilities, in that the first is elastic and the second is not, the translation from the institutional to the retail practice cannot be blindly done. It must acknowledge the financial flexibility of the individual.

Just in Time! Tuesday's column on the possibility that 401(k) plans might be "Rothed," by having their contributions taxed upfront, barely beat its deadline. The next day, the White House released the broad details of its tax-reform plan, and that proposal was absent. Along with home-mortgage deductions and charitable donations, 401(k) contributions retained their tax-sheltered status, being regarded as a middle-class perk that had become too entrenched to be withdrawn.

There remains the chance that this proposal will be advanced as the debate proceeds. The White House proposal cuts more taxes than it raises, and the main path by which it does increase revenues, that of eliminating the deduction for state and local taxes (including real estate taxes), affects some states far more than others. There will be outcry and the demand for horse trading. Although I don’t expect 401(k) contributions to be among the horses offered, that is a possibility.

John Rekenthaler has been researching the fund industry since 1988. He is now a columnist for Morningstar.com and a member of Morningstar's investment research department. John is quick to point out that while Morningstar typically agrees with the views of the Rekenthaler Report, his views are his own.