"This is well and good," you might say, "but stocks fluctuate. What are the odds that stocks will outperform bonds or T-bills over various periods of time?" Good question. It is one thing to see that stocks outperform over the long term, but as Keynes once remariced, "In the long term were all dead." How do they do then for somewhat shorter periods?

Column 4 shows the percent of times stocks beat bonds for periods varying from one to 30 years, and column 5 provides the same information for T-bills. As you can see, its a slam dunk for stocks after four years. Holding stocks, you have an 82% chance of doing better than bonds after inflation, and an 84% shot of outperforming T-bills after 48 months, moving up progressively to a 100% chance of outperforming bonds, and a 94% chance of beating T-bdls in 15 years. Beyond 15 years the odds favoring stocks surge to 100%. For longer periods, stocks are clearly the least risky of these three categories of investments.

Lets look at how stocks and bonds have performed over other periods in the past. Table 14-2 shows the probabilities of stocks outperforming bonds and T-bills after inflation over different intervals between 1802 and 1945. Three periods are analyzed, 1802 to 1870, 1871 to 1945, and 1946 to 1996. In each period the probabilities of stocks outperforming bonds and bonds beating T-bills is measured from 1 to 30 years. The table indicates the probabilities of stocks outperforming bonds or T-bills increase with time. More importandy, the odds are higher for any period from 1 to 20 years in the post World War II period than in the previous 145 years.

For 5 years the odds of stocks beating T-bills and Treasury bonds rise from about 70% in the two earlier periods to 82% after Worid War II. In the 1945 to 1996 period stocks had a 100% chance of outperforming T-bills for all 20-year periods, but by that time they clearly outperform even without the increasing inflation of the postwar period. Stocks had an 87% chance of beating T-bills after 20 years over the 1802 to 1870 period, which increased to 99% for the 1871 to 1945 time span.

Do you sense something wrong here? T-bills, the "risk-free" investment of modern portfolio theory, are providing far worse returns over time than the "risky" ones-common stocks. But wait-there is more.

Enter again the "second horseman" for bonds and T-bills-taxes. After taxes, the returns on fixed-income securities drop dramatically. For all their allure, buying long-term governments is about as safe and profitable as having been heavily margined in stocks just before October 19, 1987. We saw in chapter 13 that if an investor in a 50% tax bracket put $100,000 into long Treasury bonds after Worid War II, he would have only $39,200 of his purchasing power left in 1996. Thats right, inflation and taxes had eaten up over 60% of the investment.

Holding Portfolio for...

Stocks Beat Bonds Stocks Beat T-bills Bonds Beat T-bills

1 year

2 years

10 years

20 years

30 years

The results were equally bad for T-bills. Lower tax brackets dont help the T-bill or Treasury bond buyer that much.* Finally, had the investor put $100,000 into blue-chip stocks, the "risky" investment with the same 50% tax rate after inflation, the portfolio would have appreciated to $913,000 over the 50-year period. The capital invested in equities would be worth 23 times as much than if placed in T-bills or bonds.

Table 14-3 is identical in format to Table 14-1, but shows the retums after both inflation and taxes for stocks, bonds, and T-bills for periods of 1 to 30 years through the postwar period. Stocks compound at almost

See endnote #7, chapter 13, for a description ofhow taxes were calculated.

Table 14-2

Frequency of Stocks Outperforming Bonds & T-Bills, Bonds Outperforming T-Bills, Inflation-Adjusted 1802-1996

Table 14-3

Compounded Returns After Inflation and Taxes 1946-1996

5% annually adjusted for inflation and taxes. In 10 years the investor has increased her capital by over 54%; in 20 it is up 138%.

With bonds or T-bills its a dirge. The longer you hold them the louder the organ plays. After 10 years, both will cost you over 16% of your capital; after 20 years, 30%, and so it goes. The last two columns again demonstrate the probabilities of stocks outperforming bonds and T-bills from periods of 1 to 30 years. After 4 years there is better dian an 85% chance that stocks will outperform bonds and T-bills, and the odds build up significantly with dme. By 15 years the odds are 100% that youll do better in stocks than government bonds and 92% that youll do better than T-bills. Investing in govemment securities, then, considered by most to be almost "riskless," is a classic losers game.

Common stocks, as Tables 14-1 and 14-3 indicate, though a more volatile asset in any short period of time, provide much higher retums than T-bills and bonds over longer periods.

Is There Something Wrong with This Picture?

As Appendix A indicates, the starting point of modem portfolio theory is the retum an investor receives on a riskless asset, normally a Treasury bill. The investor then selects a portfolio made up of risk-free and more risky assets, naturally measured by their volatility, to get the optimum mix for himself. Trouble is, the "risk-free asset" of academic theory is One ofthe riskiest assets out there over time.