Appropriate Retirement Planning Assumptions

I became intensely interested in peak oil and energy/climate related issues in 2005.  The way it happened was as follows.  I was in the process of leaving my then job (with the ill-fated Nevis Networks) and restarting my episodic consulting practice.  I was working on a financial planning spreadsheet for how much I had to save in order to plan for retirement.  As everyone who has ever done such a thing knows, a very key input variable to the exercise is how much you think your investments will grow over time.  I had pulled some value from some investment book or other that I now forget, and was wondering about what risks there were in extrapolating such an assumption out over several decades.

I started googling around the Internet trying to understand big societal risks, and somehow discovered ASPO.  I had read the original Campbell-Laharrere piece in Scientific American in 1998, but somehow it was the second exposure that took - I began to intensively explore the issues, and have been ever since on-and-off.

I was reminded of this history lately, because over the weekend the WSJ published a story Pension Gaps Loom Larger.  The essence of is that many large pension funds are continuing to make the assumption that they will be able to earn 8% a year on their investments over the long haul going forward. I strongly recommend reading the whole thing, but here's a flavor:

Many of America's largest pension funds are sticking to expectations of fat returns on their investments even after a decade of paltry gains, which could leave U.S. retirement plans facing an even deeper funding hole and taxpayers on the hook for huge additional contributions.

The median expected investment return for more than 100 U.S. public pension plans surveyed by the National Association of State Retirement Administrators remains 8%, the same level as in 2001, the association says.

The country's 15 biggest public pension systems have an average expected return of 7.8%, and only a handful recently have changed or are reconsidering those return assumptions, according to a survey of those funds by The Wall Street Journal.

Corporate pension plans in many cases have been cutting expectations more quickly than public plans, but often they were starting from more-optimistic assumptions. Pension plans at companies in the Standard & Poor's 500 stock index have trimmed expected returns by one-half of a percentage point over the past five years, but their average return assumption is also 8%, according to the Analyst's Accounting Observer, a research firm.

The rosy expectations persist despite the fact that the Dow Jones Industrial Average is back near the 10000 level it first breached in 1999. The 10-year Treasury note is yielding less than 3%, and inflation is running at only about 1%, making it tougher for plans to hit their return targets.

Return assumptions can affect the size of so-called funding gaps—the amounts by which future liabilities to retirees exceed current pension assets. That's because government plans use the return rates to calculate how much money they need to meet their future obligations to retirees. When there are funding gaps, plans have to get more contributions from either employers or employees.

The concern is that the reluctance to plan for smaller gains will understate the scale of the potential time bomb facing America's government and corporate pension plans.

The basis for such assumptions is that in recent history they've been more or less met in the U.S.  For example, I found this next graph here, and the orange curve shows what would have happened had you been invested in the Standard and Poors 500 large cap stocks, and reinvested all your dividend payments at the time you got them:

Clearly you would have done pretty well, overall, and in fact the 1950-2009 total real return averaged 7.0% - I didn't go check, but very likely the 1950-2001 period was more like 8%.  So are these pension fund managers justified in extrapolating this into the future?

One obvious problem - if you just look at this data from a straight time-series statistics perspective, it's highly autocorrelated.  There are giant bull and bear market periods that last decades and that dominate the variability of the data.  The implication of that is that you actually have very little statistical power to estimate the trend (for example, the sixty years above contain less than two bull-bear cycles, so in an important sense you are averaging less than two data points, rather than, say, 60 annual data points).  So your ability to extrapolate this series into the future is highly suspect.

But thinking further about this - an obvious question is: why did the US stock market grow much faster than the economy over this period?  From 1950 to 2009, the US economy grew at an average of 3.9% (but slowing down through the period), and corporate profits were basically roughly flat as a share of the economy.  Viewing the stock market as basically a right to share in those profits, why did the value of that right grow twice as fast as the economy?  And is such a situation likely to continue?

Well, I can think of a handful of hypotheses for why stocks would have done better than the economy in recent decades.

1. Interest Rates: they have been declining since about 1980. You would expect this to result in increasing stock prices - since stocks and bonds are to some extent competitive investments, if bond yields go down, stocks do not need to yield as much to be attractive to investors, and thus prices may rise relative to earnings.
2. Demographics: as the baby boom has aged and put more focus on planning their retirement, you would have expected them to put more money into the stock market, thus boosting stock prices.
3. Declining risk premiums as the perception that "stocks are better in the long term" spread through a larger fraction of the public, meaning people were willing to pay more for stocks, relative to their earnings.
4. Increasing income inequality.  As a larger share of national income has gone to the top few percentiles of the income spectrum, they will be more likely to invest it rather than spending it.  This will tend to inflate asset prices relative to goods and services (and thus, for example, increase stock returns more than the growth in the overall economy).

So, going forward, I think we have the following problems with an eight percent assumption based on US stocks:

• Underlying economic growth in the US has been slowing for decades.  Given a) the need to deleverage, b) the likelihood that we will have to gradually transition away from oil in coming decades, and c) ongoing competition from Asia, there are reasons to be pessimistic that this trend will turn around soon.
• Interest rates are now very low, and can probably not drop a whole lot more.  Rising interest rates in the future are a risk to stock prices.
• The baby boom are now starting to retire and will be pulling money out of the market in coming decades.  My generation (I am at the leading edge of what is usually called Gen X) is smaller than the baby boom.
• Increasing income inequality is eventually likely to produce a backlash (though it's certainly hard to say when).  Anything like the democratization in income from the 1930s to the 1960s would likely hit asset prices hard.
• I think beliefs like "stocks are better in the long-haul" based on econometric observations of the past are subject to the self-falsification curse of economic science.  Such things are only true until everyone comes to believe them, and then the very fact of that belief negates their truth (eg by bidding stock prices up to a height from which they can only go down).

Given all these things, I think the future growth prospects for US investments are extremely uncertain.

Given that, my current growth assumption for my own investments is zero.  I am now working on the conservative assumption that my retirement has to be funded on a strict cash basis.