Seen in one light, it appeared the perfect time for a U.S. investor to buy a bond in the late 1890s. Disregarding year-to-year movements in prices that largely cancelled out, there was hardly any inflation over long periods of time. Also, the going rates on all sorts of fixed income investments had generally fallen over the past few decades. So, if you held out, you might find only lower yielding investments in the future.
But, seen in another light, it was a terrible time to buy a bond. The economy was recovering from a deep depression, and a strong economy could lead to some inflation that would eat away at your return. Buying now also meant accepting a fairly low yield, at least by historical standards. In the end, these considerations and a global war no one could reliably predict, meant that these years were the beginning of a two-decade-long malaise for bond investors who chose to lock in their bond investments in the late 1890s.
A Different Kind of Bond Market
At the end of 1898, the seasonally-adjusted average rate of short-term high-quality commercial paper in New York City, a measure of short-term interest rates for corporate borrowers, was at 3.0%. Rates had been a tad lower for a brief period in 1894 but these were still very low and long-term interest rates, or the yields on longer-term bonds, were also very low by historical standards.
At that time, the state of the bond market was measured by the yields on corporate bonds, often issued by railway companies, and municipal bonds. U.S. federal government bonds existed, some of them were issued to build the Panama Canal, but their yields were not great barometers of interest rates. In the late 19th century, banks issued their own banknotes to provide their clients with physical cash and the issuance of these banknotes was used by banks to fund purchases of government bonds which were required to back those very banknotes. This created demand for government bonds that had little to do with the broader bond market and these securities always traded at below-market yields.
At the turn of the century, a 2% Treasury Bond maturing in 1930 was trading hands at a premium to face value such that its actual yield was just 1.55%. By 1913, 80% of U.S. government bonds were held by nationally chartered banks who were happy to issue banknotes, earning 0% interest, to fund purchases of these government bonds, no matter how low the yield. So, the market for federal government debt was small enough and so unique to be of limited use in assessing the condition of the overall market.
Rising Rates (1899-1907)
Rightfully putting government bonds aside, the average yields on New England municipal bonds and the highest-quality corporate bonds – in other words, the safest other bonds available – were still less than 3.2%. At their lows near the turn of the century, bonds issued by the State of Maine and State of Massachusetts were yielding between 3.0% and 3.1%. Pennsylvania Railroad bonds were yielding 2.97%. Even bonds issued by the West Shore Railroad paying a 4% coupon and due over four centuries in the future, in 2361, were trading hands at a premium over face value, a price of 114.75, meaning that these very long-term investments were yielding 3.48%.
However, these numbers more or less mark the top of the market. Bond yields drifted higher over 1899 and the first decade of the 20th century which means that bond prices, which move inversely to yields, sank. Bond yields in Europe had already reached their lows two years earlier and were further along in their reversal by 1900. U.S. bonds defied this trend but only for so long.
The increase in bond yields was attributed to the end of the early-1890s depression and the giving way of a period of deflation in prices for a new era of moderate inflation. Inflation would typically cause investors to demand higher returns on bonds. Compared to a low of around 3%, short-term rates rose to 5% in the winter of 1899-90, fell back to 4% later in 1900, but were once more at 5% by 1902.
When short-term interest rates rise, sometimes, though not always, longer term yields on bonds rise too. This is reflected in lower bond prices for previously-issued bonds as investors must sell these at a discount in order to entice buyers to purchase these securities in a period when higher interest rates and yields are available elsewhere. That said, initially, the rise in short-term rates after 1898 did not cause bond prices to fall. In fact, bond prices peaked in 1902. That year, an index of ten railroad bonds had an average price of 107.0 and a separate average of forty miscellaneous bonds, from railway, utility, and industrial company issuers, stood at 103.6. This was the peak in both data series.
The bounce in short-term rates in 1902 did not stop at 5%. Unlike in early 1900, short term rates did not immediately retreat back to 4%; rather, they kept rising, exceeding 5.5% by mid-1903. From here though, short-term rates did fall back to 4% in 1904. Yet, bond yields did not revert back to prior levels which would have kept bond prices high. After having dropped below 100 for the railroad bond index mentioned earlier and below 95 for the broader index, bond values recovered, but only partially and not back to peak 1902 levels. In any case, interest rates rose again in 1905, erasing most of the 1904 reversal and hovered around 6% for most of 1906.
Panic of 1907
This marks the end of the first of three movements comprising the bond downturn of the early 20th century. If the story ended here, bonds would have still looked lackluster. Short-term rates had doubled, making short-term deposits more enticing and bond prices retreated. However, conditions in the bond market were about to get much worse. The Panic of 1907 caused brief but severe financial distress. Short-term rates reached 7% during the panic.
Long term bond yields did not rise as much. Still, prime-quality corporate bond yields reached 4.2% in November 1907 as compared 3.73% the prior November. This was still meaningful. In fact, the going yield demanded by investors would not need to rise by much for longer-term bonds to depreciate meaningfully. A 30-year 3.25%-interest bond would price at a discount, at 91.4 per $100 of face value, in order to yield 3.73%. However, if the yield demanded by investors for equivalent bonds rose to 4.20%, as they had in 1907, the bond would depreciate to 84.0. This drop from 91.4 to 84.0 may not seem like much but a financial institution holding these bonds, perhaps a pension fund or insurance company with its own long-term liabilities, might look severely underwater in their bond investments with a decline like this.
In a manner reflective of the above example, bond prices fell during the Panic of 1907; on average, railroad bond prices and the broader basket of bonds fell below 90 cents on the dollar. Cumulatively through November 1907, a 30-year 3.25% bond would have depreciated by 17.5 since 1899 on a par value of 100 in order to keep up with the higher bond yields. Approximately 12 points of this decline would have occurred in 1907 alone.
Potential Bottom (1907-1916)
In early 1908, after the panic of the prior year had subsided, bond prices recovered. The index of railroad bonds returned to 94.3 and the broader market to 91.5 in February 1908; by the following autumn, they were all the way back to 98.5 and 97.7 respectively. Helping drive this was a stunning drop in short-term rates. After reaching 7.5%, rates of high-quality commercial paper had fallen to 3.35% in December 1908. These were the lowest interest rates in New York’s commercial paper market since early 1899, the year that the bond bear market began. That hypothetical 30-year 3.25% bond that had fallen by 17.5 in the years leading to November 1907 would have recovered by 7.4 points by February 1909, based on comparable yields elsewhere.
It would seem like the start of new beginnings if rates could stay at low levels. That did not happen though. Rates did rise back to nearly 5% by the end of 1909. From here, short-term rates gyrated between 4% and 6% between 1910 and 1914. These movements, and the resulting change in bond prices, were not as severe as those before 1907. Prime commercial paper rates never returned to the high levels of the 1907 panic, even at the outbreak of war in 1914. But, all the same, they did not return to the 1898 or 1908 lows either.
As rates moved around, the expected correlation with bond prices was observed. When short-term rates moved higher, it tended to be associated with bond prices that were near the peak of the cycle and likely to move lower in their subsequent movement. When short-term rates fell, it tended to be associated with bond prices that were near the lows of the cycle and likely to move higher. In these years when rates moved around but never stayed very high or very low for long, it could have been difficult to say whether the bond bear market of the preceding decade was over or not. Bond investors and issuers would have to wait and see.
The start of the First World War in 1914 was financially disruptive but this soon gave way to a flood of money in the United States, the gold imports into the country in 1915-16. Also, the creation of the Federal Reserve was understood to allow more credit to be created for a given volume of bank reserves. The new central bank was very accommodative in the early war years. In these circumstances, short-term rates fell to 3%, their lowest levels since 1899. Yet, perhaps because of the unusual circumstances of war, bond prices did not rise meaningfully. Yields on longer term securities did not move much in response to the sharp drop in short-term rates. Even as late as 1916, it still wasn’t clear if the bond bear market was over.
War and Inflation (1917-1920)
The First World War forever changed the bond market in several countries, including the United States. For one, there was tremendous growth in government financing needs. In the U.S., the stock of government debt was previously small enough and held by banks rather than capital market investors, that they had comprised a small part of bond holdings by non-bank investors. But Treasury bonds, now offering higher rates, soon became a competing destination for investor capital in large volumes.
In addition to large scale government borrowing, interest rates rose. Short-term rates jumped from around 3% in November 1915 to 4.28% by April 1917, the month the U.S. entered the war. This happened despite the Federal Reserve’s discount rate being kept in a 3.0-3.5% range. From here, interest rates rose further and bonds sank for the duration of the war.
As short-term rates rose, so too did yields on long-term borrowings. This can be tracked clearly in the various war bond issuances of 1917-19. A $2 billion loan was issued at 3.5% interest in 1917, followed later the same year by a $3.8 billion 4% issue, and then two issues summing to $11.2 billion in total at a 4.5% rate in 1918. While a sentimental appeal to buyers meant these bonds were issued with their par value intact, all of these bonds traded at discounts as soon as they were issued.
Higher bond yields on newer bonds diminished the appeal of older issues at lower rates. This sent bond prices lower. Inflation also may have been responsible for some of the depreciation of longer-term bonds. An index of wholesale prices that stood at 100 in 1896 and 150 in 1910 reached 340 by 1920. Thirty-year bond yields reported in Durand’s Basic Yields of Corporate Bonds rose from 4.05% in February 1917 to 4.75% the following February.
Municipal bond yields increased much less, from an average level of 4.23% in 1917 to 4.57% in 1918 but these bonds were exempt from federal income taxes which reached a peak rate of 77% during the war. As a result of a changing value proposition on account of their tax-exempt status, these bonds were not as representative of the movements in the bond market as corporate bonds were. Of these, the bonds of railroad companies had fallen below 75 and the broader index of issuers below 85. These were remarkably low levels for what should be safe investments.
Short-term rates fell with the armistice of November 1918. Yet, even peace could not reverse the bear market. The Federal Reserve lifted its rediscount rate to 4.75% in 1919 and to 7% in 1920. Interest rates surged to 8% that year. A new Treasury bond issuance in 1919 was sold to investors with a 4.75% rate and the very next year, short-term Treasury certificates were yielding as much as 7.75%.
Longer term yields did not rise as much but were still lifted from 4.95% in January 1920 to 5.56% just four months later. Bond prices reached their low in May 1920. By now, railroad and industrial bonds were trading hands at an average price of around 70 and the lowest coupon rate securities of the wartime government bonds were changing hands below 90. A long-term Lake Shore & Michigan Southern Railway 3.5% bond due in 1997 was trading in 1920 at a price of 65 and this was by no means exceptional. N.Y. Central Railroad and Morris and Essex Railroad bonds, also 3.5% bonds of around the same term, were also trading in the mid-60s. The period between 1917 and 1920 saw the steepest decline in bond prices of the bear market, bond prices had fallen by 23.6% on average in these three years.
Interest rates eventually fell in peacetime but only after 1920. Bond prices began their recovery from 1920-22. In the summer of 1922, bond prices had recovered to the levels of 1917, the year the U.S. entered the war. In fact, 1922 had been one of the best for bond investors in U.S. financial history, with a price appreciation of 10% on federal government bonds complementing interest income to deliver a total return of 15%. This was a single-year return approximately equivalent to the cumulative total return of the prior five years taken together. The market had finally turned around for good.
Lesson
In 1920, the U.S. bond market was just beginning to exit a bear market that had lasted about two decades. It may seem that these cycles in the bond market are incongruous with and far longer than those in the stock market. To some extent, this is an illusion. Because bond market movements are smaller year-to-year than those in the market for common stock, the short-term fluctuations are not dwelt on for long. So, widespread recognition that a transition from a bull market to a bear market in bonds, or vice versa, is usually reserved for once in a generation reversals that truly inaugurate new eras that themselves last decades, or at least a couple of decades.
The same may not be true of investments in the stock market but this is down to perception rather than actual performance. Over very long periods of time, even on the scale of a century or more, most of the returns on stocks are concentrated in a few decades. The 1950s-60s, 1980s-90s, and 2010s-20s are responsible for the lion’s share of the last century’s worth of U.S. stock returns. There are also periods of a decade or two when stock returns can remain abysmal – consider the 1930s-40s, 1970s, and 2000s. That there are far longer cycles beyond the year-to-year fluctuations that can summarize investment outcomes is just as true of stocks, but the less excitable ‘glacial’ nature of bond investing makes these long-term trends more noticeable in that market.
More from the Tontine Coffee-House
Read about the two panics that marked the beginnings and ends of different phases of the 1896-1920 bond bear market, the Panic of 1907 and the July Crisis. Consider subscribing to this blog’s newsletter or checking out book recommendations, which include many of the sources often referenced in my posts.
Further Reading
1. Homer, Sidney, and Richard Eugene Sylla. “Detail of the First Bear Bond Market: 1899-1920.” A History of Interest Rates, 3rd ed., Rutgers University Press, 1991.
2. McQuarrie, Edward F. “The US Bond Market Before 1926: Investor Total Return From 1793, Comparing Federal, Municipal and Corporate Bonds Part II: 1857 to 1926.” SSRN Electronic Journal, SSRN 3269683, Sept. 2019.
3. Persons, Warren M., and Edwin Frickey. “Money Rates and Security Prices.” The Review of Economics and Statistics, vol. 8, no. 1, Jan. 1926, pp. 29–46.
Consider Subscribing:
