One of the most misunderstood concepts in investing is that of risk. Often, I will advise a retiree to keep virtually all of their assets in equity (ownership) investments, including domestic and international stocks, real estate, and (horror of horrors) commodity futures. The most frequent objection I hear is that they are afraid of the risk. I'm afraid of the risk they will face if they DON'T invest that way.
The most unfortunate event in the history of investment theory occurred when risk was defined as VOLATILITY. By this definition, it is obvious that stocks are risky and commodity futures are insanely risky. But volatility is a lousy definition of risk for the average person.
My favorite definition of risk came from Robert Jeffrey, who defined risk as "the probability of not having sufficient cash with which to buy something important." If you keep your savings in bank accounts your entire working life you are taking an incredible risk. Assuming you save 10% of every paycheck, you'll find yourself at retirement with savings equivalent to about 4 years of earnings, and your bank account, after taxes and inflation, will have earned nothing IF YOU'RE LUCKY, since the actual return after taxes and inflation (also known as the "real" return) is normally negative (this and subsequent numbers derive from various publications of Ibbotson Associates). You might have earned about 1% a year after taxes and inflation had you kept the money in bonds, and instead of 4 years of earnings, you would have accumulated almost 5 years (unless inflation rose unexpectedly over that time period, in which case you'd probably have less than the savings accounts would have produced). If you had kept the money entirely in large U.S. common stocks, it would have grown, on average, to more than 15 years of earnings. Only this last result would likely provide enough for a comfortable retirement. Clearly, the person relying on savings accounts or bonds is the one who is at tremendous risk of not being able to live out their retirement in comfort.
But, of course, equities are volatile, so the results might have been significantly better or worse than that. You wouldn't mind better, but you WOULD mind worse. How much worse? Well, it wouldn't have been wise to keep the equity money entirely in US stocks, ignoring international stocks, real estate, and commodity futures, but if you had done so during the worst 40 years in American stock market history (which you might be surprised to know was the period from 1880 to 1920), your stake would only have grown to equal 7 years of earnings. This, of course, is much, much worse than having accumulated 15 years of earnings, but after doing extensive calculations, I've determined that it is still better than having only 4 or 5 years.
So am I saying stocks aren't risky? Of course not. Though 7 years of earnings is more than 7 years of spending (since you obviously spent less than you earned during your working years if you were able to save and pay your taxes), and you might have social security to supplement it (though you wouldn't have had social security in 1920), and you might have had some growth after retirement had you stayed the course a bit longer (by 1924, the stock market had more than doubled from its dismal 1920 level), there is clearly a possibility that your 40-year faith in the US stock market would not have built a sufficient stake to fund your retirement. And you might not have taken much comfort in knowing you just happened to have been investing during a remarkably bad period (things would not have been as dismal in the worst US stock market periods for an investor whose equity holdings included the other 3 categories I mentioned earlier, but still might have fallen several years short of the 15 year average that sounded so good in my initial summary).
But if a mere 7 years of earnings would have failed to provide for the average retirement, what of the investor who was earning nothing or 1% in bank accounts or bonds? The worst 40 years in stocks beat the best 40 years in bank accounts, and 40 years in bonds last grew to more than 7 years of earnings in the early 1800s when they were priced based on the young US government's status as an emerging market. And let's not forget that the worst periods for interest-bearing investments would have brought the stake down as far as only 3 years of earnings. The 35-year bear market in bonds from 1946 to 1981 brought them down by 65% in real terms, and that was BEFORE considering the income taxes that would have been owed on all the phony, inflation-generated interest income (bank accounts also steadily lost real value over that time period, losing 20% in real terms before taxes). After considering inflation and income taxes, the 1946-81 slaughter in bonds was slightly WORSE than the infamous 1929-32 crash in stocks. Fear stocks, especially over the short term. But don't fool yourself into thinking bonds are safe. For every time period I was able to examine longer than 7 years, the worst real performances of bonds and bank accounts were worse than the worst real performance of stocks.
Now, I might not get much of an argument from you about keeping your assets in equities during your working life (for those thinking a mixture would have been superior, the best, worst, and average 40 years all drop as more interest-bearing investments are added to the mix), but certainly I wouldn't be recommending you stay entirely in equities after retirement, would I?
Not so fast. For one thing, I didn't say you should keep everything in equities during your working life. History lessons are useful, but rearview mirrors aren't windshields. Still, you should carefully consider your reasons for keeping any of your long-term money intended for retirement in investments that have historically had both lower highs AND lower lows.
On to retirement day. Before we make the automatic assumption that you should be shifting heavily into bonds and banks, let's keep our eye on the risk of outliving your income. Assume that, by retirement, you have grown your assets to the point that you only need to take out 5% of your assets in the first year of retirement to live comfortably, and only need the amount you draw to rise each year with increases in the cost of living.
To repeat, even if I become tiresome: I'm not recommending that your equity holdings be entirely in US stocks, since you should be able to reduce your volatility without reducing expected returns with wider equity diversification (this is an article, not a book, so forgive me for not discussing that in the present piece). But let's assume US stocks are your only equity choice. For your debt choice, let's use US Treasury bonds (TIPS, or Treasury Inflation-Protected Securities, would be better, but if I'm going to be unfair to the equity side, I get to be equally unfair to the debt side, and credible long-term data is limited for the alternative choices).
With thanks to the American Association of Individual Investors, which did the calculations from the Ibbotson database of returns since 1926, I can tell you that your chance of having your retirement assets last 30 years (and there is a very good chance of at least one member of a retired couple living that long) on a portfolio consisting entirely of US bonds was only 17% (meaning you had an 83% chance of going broke during that time). Okay, okay, that wasn't fair, because you weren't considering a 100% bond (or, worse, bank account) portfolio. You just wanted to lighten up on the stocks at that point. Financial journalists are fond of recommending 50-50 splits for retirees between stocks and bonds, and this reduced the chance of going broke to only 24% over that time period. Much, much better: this popular approach succeeded three-quarters of the time. But was it the best? Well, increase that stock stake to 75% of the total, and the chance of going broke dropped to only 17% (which was the chance of NOT going broke for the "cautious" retiree entirely in US Treasury bonds). Just for laughs, what would happen if you threw caution to the wind and put 100% into large US stocks? The probability of your running out of money in those 30 years was 15%, the best of the bunch. The safest choice.
Is it the safest choice for the future? Well, you can criticize a reliance on historical evidence all you want, but before you use what John Templeton called the four most dangerous words in investing, "It's different this time," show the same skepticism toward predictions of a different future as you show toward predictions of a similar one. There are reasons in human nature and fundamental economic theory for the relationship between stock, bond, and savings account returns, and those who claim the near-7% real returns reported by Ibbotson since 1926 represent unusual circumstances need to explain why the Cowles Commission study of 1871 to 1925 also reported real stock returns of nearly 7% and a preliminary study derived from the New York newspapers between 1802 and 1870 reported real stock returns of nearly 7%. A study of the complete 20th century returns of 16 different countries by London Business School Professors Dimson, Marsh, and Staunton revealed a global stock average of 6% (countries which are leveled in wars do suffer some permanent economic losses, but the harm affects stocks, bonds, and banks and doesn't really offer a safe haven other than to be globally diversified). For the world as a whole, the 20th century was an ugly and violent one, and if it muddled through two world wars and a cold one and still averaged 6% returns to owners of businesses, why would it be significantly less in the 21st unless there was a global disaster so great that any reasoned discussion became pointless?
Now, I've worked with retirees long enough to know that few of them are willing to keep EVERYTHING in equities, and since the chance of going broke appears to be about the same between 75% and 100% equities, I'll usually declare victory once I can get them up to 80% or so. Yet, if there really isn't much of a risk difference to the retiree, the difference between 75% and 100% equity is enormous in terms of the expected legacy they might leave for their children and grandchildren or valued charitable organizations.
So, would a 100% US stock portfolio be risky for the retiree in the example cited? Absolutely! A 15% risk of going broke in the next 30 years is not acceptable to most: we insure our homes against fire losses that have only a 1 in 1200 chance of occurring. I'd NEVER be satisfied with recommending to a client a portfolio that left a 15% risk of their going broke if we could do better. I'd see how much the risk could be reduced by adding other categories of equity with similar expected returns but which had their best and worst performances at different times than US stocks (this is what makes international stocks, real estate, and commodity futures so important to consider). I'd look at the flexibility of their spending needs to see if they might either reduce their initial draw or be prepared to do so in the event of poor periods of performance. If necessary, I'd consider the use of immediate variable annuities (see the appropriate piece on variable annuities elsewhere on the wiki) which might still be invested in equities but guarantee payments for life. I'd look at the merits of a reverse mortgage.
But the one thing I would never do to a retiree who was at signficant risk of outliving their income is recommend they add investments which actually INCREASE THAT RISK. Recommending a 100% equity position would NOT mean I believed it had a 100% probability of success, or that I was certain it would outperform bonds and bank accounts 100% of the time. It would mean that, in the choice of risks that had to be taken, I chose to take the smaller one over the larger one.
One final note: I've mentioned the use of TIPS for the debt portion of a portfolio, when a client's emotional needs require it or if they have no desire to leave an estate, to cite two possible reasons. I've had several people insist to me that my arguments about bonds being devastated by inflation (such as that which occurred between 1946 and 1981) don't apply to TIPS. But this is only true for earnings that aren't subject to income taxes. If TIPS guarantee inflation plus 2%, and a person is in the 25% Federal tax bracket, an inflation rate of 6% results in a real return after taxes and inflation of nothing (the total return of 8% will be reduced by 2% for taxes to equal the inflation rate). And if the TIPS are in a Roth IRA, free from any taxes, earning a 2% real return, they still don't do much to improve the chances of someone wanting to draw out 5% without slowly going broke.
