What is really important concerning the attractiveness of bonds as long-term investments is whether a similar trend can be expected in the period ahead. It seems to me that if this whole inflation mechanism is studied carefully it becomes clear that major inflationary spurts arise out of wholesale expansions of credit, which in turn result from large government deficits greatly enlarging the monetary base of the credit system. The huge deficit incurred in winning World War II laid such a base. The result was that prewar bondholders who have maintained their positions in fixed-income securities have lost over half the real value of their investments.
As already explained, our laws, and more importantly our accepted beliefs of what should be done in a depression, make one of two courses seem inevitable. Either business will remain good, in which event outstanding stocks will continue to out-perform bonds, or a significant recession will occur. If this happens, bonds should temporarily outperform the best stocks, but a train of major deficit-producing actions will then be triggered that will cause another major decline in the true purchasing power of bond-type investments. It is almost certain that a depression will produce further major inflation; the extreme difficulty of determining when in such a disturbing period bonds should be sold makes me believes that securities of this type are, in our complex economy, primarily suited either to banks, insurance companies and other institutions that have dollar obligations to offset against them, or to individuals with short-term objectives. They do not provide for sufficient gain to the long-term investor to offset this probability of further depreciation in purchasing power.
~ Philip Fisher, "Clues from the Past", Common Stocks and Uncommon Profit, 1958