Problems With Constant Compound Interest Whether for Governments or Retirees

This is a continuation of an irregular series which you can find here. Maybe if I were more scientific, I would have called it “All Exponential Growth Processes Run Into Constraints and Threats,” or if I were more poetic, “Nothing Lasts Forever — Nothing Grows to the Sky.”

Regardless, simple modeling is the bane of long-duration financial calculations. I remember talking with some friends who served on a charitable board with me about some investment grade long bonds (11-30 years) that I had purchased for a life insurance client that yielded 7-9% in late 1999. They said to me that it was foolish to lock up money for so long in bonds when you could earn so much more in stocks. My three comments to them were:

• It's prohibitive for life insurers to hold equities.
• At current levels of the market, the yield of these bonds more than compensates for the possibility of capital growth in equities (valuations are stretched).
• The risk in the bonds is a lot lower.

And I said we ought to shift our charity’s asset allocation to more bonds, as we were invested past the maximum of our guidelines in equities. They looked in the rearview mirror and said that we were doing fabulous. Why change success?

I was outvoted; I was a one-man minority. There are a lot of people who would have loved to make that change in hindsight, but done is done. I ended up leaving the board a year later over a related issue.

Now, don’t think that I am advising the same in 2011. We may be headed for significant inflation or deflation; it is difficult to tell which. Bonds offer little competition to equities here. Commodities and cash may be better, but I am reluctant to be too dogmatic. If the economy turns down again, long Treasuries would be best.

Here’s the difficulty: Most people have been trained to think at least one of a few things that are wrong:

• That we can use simple models to forecast future outcomes.
• That average people are capable of avoiding fear and greed when it comes to investing.
• That financial markets are random in the sense that last period’s return has no effect on the returns of future periods.
• Over long periods of time, average investors can beat long Treasuries by more than 2% a year. (Corollary to the idea that the equity premium is 4-6% versus 0-2% a year over high-quality bonds.)
• That financial markets are expressions of what is going on in the real economy.
• That the real economy tends toward stability.
• That government actions make the real economy more stable.

I’m prompted to write this because of two articles that I ran across in the last day: Retiring Boomers Find 401(k) Plans Fall Short and Stay Out of the ROOM (registration required).

I’ve written about this before in many places, including Ancient and Modern: The Retirement Tripod. And yet, when I wrote about these issues 20 years ago, one of the things that I tried to point out was that as the demographic bulge retired, it would be difficult for homes and asset markets to throw off the returns necessary, because there would not be enough buyers for the assets/homes. If a large portion of the population wants to convert assets into a stream of income — guess what? They are forced sellers, and yields that they will get will be compressed as a result.

In a situation like that, those who are better off and can delay turning all of their assets into an earnings stream should be disproportionately better off. As with corporations, so with individuals/families: Those with slack assets and flexibility are able to deal with volatility better than those for whom the environment must be stable/favorable for the plan to succeed.

Now, The Wall Street Journal article points at the problems of 401(k) plans. What it says is true, but the same is true of other types of defined contribution and defined benefit plans. When assets underperform, and/or investors make bad choices, guess what? The pain has to be compensated for somehow:

• 401(k)s: They will work longer, maybe all of the rest of their lives, and cut back on expenses and dreams.
• Non-contributory DCs: Maybe the employer will ask them to kick in voluntarily, or he might give more. Also same as 401(k).
• Private sector DB plans: Employers may contribute more, or they may terminate them.
• Public sector DB plans: Taxes may rise, spending cuts enacted, forced contributions to retiree plans negotiated, plans terminated for a 457 plan, partial plan termination, job cuts, funny accounting practices (worse than the private sphere), brinksmanship over debts, etc.

Note that one of the answers is not “take more risk.” First, risk and return are virtually uncorrelated in practice. Only when enough people realize that might risk and return become positively correlated. Second, there are times to increase and decrease risk exposure. Typical people won’t want to do that, because of euphoria (the example of my friends above) and panic. The time to add to high-risk assets is when no one wants to touch a high-yield bond. More broadly, always look for asset classes that throw off the best cash flow yields, conservatively estimated, over the next 10-plus years. Be sure and factor in the likelihood for economic regime changes and capital loss, inflation, deflation, etc.

Good asset allocation marries the time horizon of an investor to the forecasts for future returns, conservatively stated, and considers what could go wrong. At present, investment opportunities are average-ish. I would be wary of stretching for yield here, or raising my risk exposure in equities. Stick with high quality.

And, for those who are retired, I would be wary of taking too much into income. I have a simple formula for how much one could take from an endowment at maximum:

• 10-Year Treasury Yield
• Plus a credit spread — 2% if spreads are sky-high, 1% if they are good, 0.5% if they are tight.
• Less losses and fees of 0.5% — higher if investment expenses are over 0.25%.

Not very scientific, but I think it is realistic. At a 3.5% 10-yr T-note yield, that puts me at a 4% maximum withdrawal rate, given a 1% credit spread. This attempts to marry withdrawals to alternative uses for capital in the market. You may withdraw more when opportunities are high, and less when they are low. (But who can be flexible enough to have a maximum spending policy that varies over time?)

Now some of the advanced models that calculate odds of retiring successfully are a step in the right direction, but they also need to reflect demographics, time-correlation of returns, regime-shifting returns/economics, etc. Things don’t move randomly in markets; that doesn’t mean I know which way things are going, but it does mean I should be cautious unless the market is offering me a fat pitch to hit.

These statements apply to governments as well, and their financial security programs. In aggregate, investments can’t outgrow growth in GDP by much, unless labor takes a progressively lower share of national income. (And who knows but that the pressure on union DB plans to earn high returns might lead to takeovers/layoffs in private firms …) The real economy and the financial economy are one over the long haul, but can drift apart considerably in the intermediate term.

In summary, any long promise/analysis/plan made must reflect the realities that I mention here. We’ve spent years on the illusions generated by assuming high returns off of financial assets. Now with the first Baby Boomers trying to retire, the reality has arrived — sorry, not everyone in a large birth cohort can retire comfortably. Wish it could be otherwise, but the economy as a whole can’t generate enough to make that proposition work.

I don’t intend that this series should have more parts, but if a relevant idea strikes me, I will write again.