|
Eugene F. Fama Jr., Vice President, Dimensional Fund Advisors
Investors sometimes use bonds to meet current and future income needs. This is especially true in retirement, when cash from bonds can substitute for a regular paycheck. But while bonds pay steady income, using them expressly for that purpose can add risks you didn’t plan on.
Using bonds to meet future cash needs or generate income differs from the more traditional asset allocation approach, where bonds diversify a portfolio and reduce overall risk. In the former approach, bonds are often seen as a de facto insurance policy that “immunizes” retirement goals from investment risk. Fixed income covers ongoing bills and frees up the rest of a portfolio to pursue growth through riskier instruments. If the riskier stuff doesn’t pan out, the thinking goes, you still have the bonds to support you in high style. If the riskier stuff wins big, you might not even need the bonds. Either way you win, right?
Maybe not. The raw pursuit of income can engage inadvertent risks that are especially rough on retirees. And for what? In the end, financial security is about total wealth—the present value of net worth, not marginal income. Understandably, insecurities plague investors looking forward to their last paycheck, so some logic and perspective can help avoid pitfalls.
Getting enough cash out of bonds to pay the bills usually means extending maturities, going lower in credit quality, or both. This increases yield, but it also increases risk. Historically, longer-term bonds have been more volatile (as measured by standard deviation) than shorter-term bonds, and without much added return to show for it.
Does It Pay to Extend Maturities? (Quarterly Data: 1964-2006) |
| |
|
 |
| |
One-Month US Treasury Bills, Five-Year US Treasury Notes, and Twenty-Year (Long-Term) US Government Bonds provided by Ibbotson Associates. Six-Month US Treasury Bills provided by CRSP (1964-1977) and Merrill Lynch (1978-present). One-Year US Treasury Notes provided by CRSP (1964-May 1991) and Merrill Lynch (June 1991-present). One-Year US Treasury Notes provided by © Stocks, Bonds, Bills and Inflation Yearbook™, Ibbotson Associates, Chicago (annually updated work by Roger G. Ibbotson and Rex A. Sinquefield). CRSP data provided by the Center for Research in Security Prices, University of Chicago. The Merrill Lynch Indices are used with permission; copyright 2006 Merrill Lynch, Pierce, Fenner & Smith Incorporated; all rights reserved. Past performance is no guarantee of future results. |
It’s not entirely clear why this is the case. It may be because institutional investors (like insurance companies) use long-term bonds to meet future obligations such as employee pensions. Thanks to actuarial tables, they have a pretty good idea of the size and timing of their payouts, so they simply match the duration of their bonds to these dates. When the bonds mature, the proceeds pay off the obligations. In this limited framework, volatility along the way doesn’t matter. Therefore, the way long-term bonds are priced might not be determined mainly by volatility but by factors related to the liability streams of big market players.
Anyway, that’s just a guess. What we do know is that institutions use bonds in exactly this way, and that it might even make sense—for them. When an employee retires, his pension benefit is typically not adjusted for inflation. It’s a “nominal” liability to the plan that can be met with a nominal bond. If the retiree is youngish, the liability is longer in term, so the plan buys a longer-term nominal bond. The institution’s goal all along is not to provide for the financial security of the retiree but rather to meet the obligations of the plan. As long as the plan isn’t among the relative handful that includes cost-of-living adjustments, inflation is all but irrelevant. The only inflation the plan needs to anticipate is wage inflation: the size of the retirement benefit is a function of the employee’s last pay rate. From there on, the payout is purely nominal.
This would make the institution’s job easier but for one inconvenient glitch. By regulation, every year the plan has to mark-to-market all its assets and liabilities. If assets come up short of its estimated funding needs, the plan has to fund the difference today by contributing new assets. So even a plan with long-term liabilities has an incentive to behave like a short-term investor.
Contrast this with an individual saving for his own retirement. He knows that even if he earns a paltry return this year, he won’t have to add extra money or defer consumption; he has plenty of years to make up the difference. He therefore has the luxury of being a genuine long-term investor. Yet, also unlike the plan, the individual investor’s obligations aren’t “nominal.” His liability stream is his future consumption—which is highly sensitive to inflation. A long-term bond is a lousy way to hedge this liability because it tends to tank with unexpected inflation, right along with the investor’s spending power. So, an individual might be better off with a bond that moves up and down with inflation, like a short-term nominal bond or a Treasury inflation-protected security. These instruments, however, don’t throw off as much cash in dividends.
Should we care whether we draw income from dividends or capital growth? Dimensional’s resident bond guru Dave Plecha answers this with a nice illustration. Imagine two portfolios with the same average return, say, of 5%. Now imagine Portfolio A has a standard deviation of 1% and none of its 5% average return comes from income. Meanwhile, Portfolio B has a standard deviation of 18% and all of its 5% average return comes from income. Which would you want to hold?
Financial theory teaches that for two portfolios with roughly equivalent average return, the portfolio with lower variance will have the greater terminal wealth. The relevant goal for advisors and their clients is therefore to maximize return for a given level of volatility. Like it or not, bonds are not surgical instruments applied to a specific purpose. They are components of an entire plan. To treat them as something else—“fixed income”—can undermine your primary goal as an investor, if not the very tenets of portfolio theory itself.
One reason you want to maximize return for risk is because funding needs in retirement are not entirely predictable. People aren’t insurance companies—they don’t know all their future liabilities. Life expectancy, health care, and other costs can change suddenly and drastically, and investors might need more money than their bonds provide. Holding long-term bonds in pursuit of raw income can therefore risk principal.
Unless you’re a big pension plan, it probably makes sense to hold high-quality short-term bonds with maturities that vary according to changes in the yield curve. This approach pursues expected return instead of income, and in the past has generated less volatility than longer bonds. Plus, short-term bonds seem to offer better inflation protection, a crucial benefit for individual investors.
This still leaves the question of where income will come from absent hefty bond yields. Research by Merton Miller and Franco Modigliani teaches that money is money, whether it comes from a dividend or from capital growth. There’s no reason to prefer one above the other.
With this in mind, the best way to meet monthly cash needs might just be to redeem assets. This approach—call it a “synthetic dividend”—can also help manage taxes and costs. If you redeem any instrument you’ve held for longer than a year, the cash receipts are taxed at the capital gains rates of 15%, instead of at higher dividend income tax rates. You can use these redemptions to rebalance a portfolio, selling shares of whatever piece is over-funded relative to its target weight. Or you can sell investments with embedded losses and offset taxable gains elsewhere in the portfolio. Viewing the portfolio in its totality rather than in pieces opens opportunities to add value.
I realize that drawing cash from investment principal instead of from income can be a tough pill to swallow. After all, regular payments feel secure—especially to investors that are no longer bringing home paychecks and don’t want to chip away at their life savings. But this might be one of those cases where financial principles give us the discipline to beat back emotion and do something that makes more sense. A short-term bond might be better suited to individual investors, even if its monthly cash payout feels less secure.
In the end, you shouldn’t care. The form of your cash flow is less important than the size of your wealth. Investors can draw cash from a higher-yielding strategy and take a lot of extra risk that might not pay off, or they can reduce bond risk and redirect the saved risk to equities, where history suggests the average return is stronger.
One crucial proviso: I’m not advocating that older or retired investors load up on stocks. To the contrary, because retirees depend on their investments more, they should deploy them more conservatively—which is why a portfolio of short-term, high-quality bonds and a focus on a strong tradeoff between risk and return are rules to retire by.
|
|