GFD Guide to Best and Worst Investment Periods in the 20th Century

This guide provides information on the 20 best days, weeks, months and years for all of the world’s major stock markets and for the World Index.

Data are calculated on a calendar basis. Weeks are from Friday (or Saturday) to the next Friday (or Saturday), months are from one month’s end to the next, and annual data are from December 31 of one year to December 31 of the next year. Using calendar months and years can obscure large movements in stock markets within a single month or year, but this methodology does provide some consistency in the comparisons.

All calculations of changes are calculated on a percentage basis rather than on an absolute basis. Since the base for each index varies from country to country, and an index such as the Dow Jones Industrials has ranged from a low of the 40s in 1932 to over 11,000 in 2000, a comparison based upon absolute point changes would be misleading.

We provide comments on the data for each country below. Brief descriptions of the data sources are provided, but for more detailed information, see the Encyclopedia of Global Financial Markets.


The Australian data are based upon three primary series. Monthly data back to 1875 are taken from Lamberton’s data on Australian stock markets. The Sydney Stock Exchange began calculating daily data in 1938. Data from 1938 to 1958 are based upon our 45 Ordinary Shares Index, which includes both 34 Ordinaries and 11 Mining stocks. The Sydney All-Ordinaries Index is used from 1958 to 1979, and the ASX All-Ordinaries index is used from 1980 on. The data for Australia is based upon monthly data from 1875 to 1937 and daily data from 1938 to date.

No particular period appears to dominate the returns, as is true in some other countries. October 1987 stands out, containing the worst daily, weekly and monthly data for Australia. The weakness of the Australian market in 1973 and 1974 is clear since these two years rank as two of the three worst in Australia’s history. But bear markets give way to bull markets, and 1975 and 1983 are two of Australia’s best years. Similarly, November 1987 contains several of the best days in Australia’s history as the market bounced back from the October 1987 crash.


No single series is available for all of Canada’s stock market history. We have had to combine three indices to get long-term daily data for Canada. The Montreal Combined index is used from 1938 to 1962, the Toronto Industrials is used from 1963 to 1975, and the S&P/TSX 300 from 1976 to date. For monthly data we have relied upon the Investor’s Index for data from 1914 to 1955, and the S&P/TSX 300 from 1956 to date. The data for Canada are based upon monthly data from 1914 to 1936 and daily data from 1937 to date.

Some of the worst days and weeks occurred between 1937 and 1940. If daily data were available back to 1929, there would be more best and worst daily and weekly periods from the early 1930s in Canada.

For the data that are available, September 1939 has the best day and week for Canada as the market became optimistic about profits from the war in Europe. October 1987 ranks as the worst crash in Canada’s history on a daily basis, but Dunkirk caused the worst week in Canada’s history. The importance of the volatility of the 1930s is seen more clearly in the monthly and annual data where the 1930s are represented prominently.


Several series were used for France’s data. Data are available from the French statistical institute, INSEE, from 1856 to 1894 on an annual basis, and from 1895 to date on a monthly basis. Weekly data are also available between 1929 and 1940 from the Annalist index of 15 French stocks and between 1945 and 1949 from the INSEE. In the 1960s, France had a daily index based upon 50 Blue Chip stocks. This index is used for weekly data from 1962 to 1969 and for daily data from 1965 to 1969. From 1969 to 1990, the CAC General Index is used, and from 1991 on, the SBF-250 index is used. The data for France is based upon annual data from 1856 to 1895, monthly data from 1896 to 1962, weekly data from 1929 to 1940, 1945 to 1949 and 1962 to 1964 and daily data since 1965.

In 1981, Francois Mitterand was elected President of France, and fears over having a socialist President sent stocks tumbling, producing a national stock market shock that was worse than October 1987 and other international crashes. Several of France’s best and worst days and weeks since 1962 have occurred in the past few years.

World War II stands out as producing the most volatile months for France. The uncertainty of this time period produced a roller coaster ride for French investors. The 1940s also produced some huge gains on an annual basis, but these gains should be considered against the backdrop of runaway inflation.


Germany has the most volatile data of any country that is included here because of their involvement in World War I, the hyperinflation that followed in the 1920s and the disaster that World War II created for Germany. The data for the hyperinflationary period are adjusted for inflation and are calculated in gold marks rather than in paper marks. Without this adjustment, all of the best periods would come from the hyper-inflationary years.

1948 also provides an important point for our data. In 1948, Germany reformed its currency, and all financial assets were converted at a rate of 1 Deutschemark for 10 Reichsmark, imposing an immediate 90% loss on all shareholders. However, price controls that had been imposed on stocks between 1940 and 1948 were removed, and the currency reform allowed stocks to seek their true level.

Annual data are available from the Reichsamt from 1856 to 1870 and monthly data from 1871 on. The Reichsamt kept a weekly index of stocks from 1924 to 1944, and these data are included. Unfortunately, no daily or weekly data are available for 1948, otherwise this year would figure prominently in daily and weekly time periods. The Commerzbank Index is used from 1955 to 1959, and the DAX index from 1959 on. The data for Germany are based upon annual data from 1856 to 1870, monthly nominal data from 1871 to 1918, monthly inflation-adjusted data from 1919 to 1923, weekly data from 1924 to 1944, monthly data from 1945 to 1954, and daily data from 1955 to date.

On a daily basis, the past few years have been much more volatile than any period since 1955. Once the interwar weekly data are included, we can see that the volatility of the German stock market in the 1920s and 1930s dwarves all other periods. October 1987 barely makes the top 10 for declines on a weekly basis. The 30% increase for the first week of 1924 must surely be a record.

The stabilization of the German mark in 1923 led to a doubling of the stock market in September 1923. Data from the 1918-1924 period dominates the monthly changes, though July 1948 when the currency reform was introduced in Germany, provides the worst returns for any single month. The volatility of the 1920s is similarly reflected in the annual data.


The Italian Statistical Institute calculated monthly data for Italy from 1905 until 1956. The Edison stock index is available on a daily basis from 1956 to 1972 and the BCI Index of 30 stocks is used from 1973 to date. Data for Italy are monthly from 1905 until 1956 and daily thereafter.

1981 and 1986 stand out as two of the most volatile periods in Italy since 1956 on a daily and on a weekly basis. When monthly data are considered, the 1940s provide the largest number of best and worst months. The best month data are biased by the strong inflation that occurred in Italy during and right after World War II. This also affects the monthly decreases as well as when markets overreacted to contemporary events.

Interestingly, Italy has had five years in which markets rose by over 90%, though three of these years were during the inflationary 1940s. No single period stands out for the worst years in Italy’s history, since bear markets are scattered throughout Italy’s history.


After Germany, Japan has had the most volatile market of any of the countries that are covered here, primarily because of World War II and its aftermath, but also because of the popping of the Japanese bubble in the 1990s. The National Bank Index is used from July 1914 through December 1932, and the Oriental Economist Index is used from 1933 through September 1948. The Nikkei is used from May 1949 to 1952, and the TOPIX index is used from 1953 to date. Data are monthly from 1914 to 1952 and daily from 1953 on.

Clearly, the vast majority of the best and worst daily and weekly time periods in Japan have occurred during the popping of the Japanese bubble during since 1989. On a monthly basis, the period after World War II dominates the record, because of the uncertainty in the market, the inflation, and the recovery that followed. The best months are primarily from the inflationary period between 1946 and 1949. It is surprising how few of the worst months are from the 1990s, given the extent of the decline that has occurred.

On an annual basis, the impact of the 1990s is clearer, with 5 of the 10 worst years occurring since 1989. Two post-war recessions caused a couple of the worst annual declines in Japan’s history. It is also interesting to note that three increases in the Tokyo stock market over 80% occurred in non-inflationary times.


Next to the United States, more data are available from the London Stock Exchange than from any other country. Monthly index data are available for London from 1700 to date, and the Financial Times 30 Industrials provides the longest single daily time series outside of the United States with daily data back to 1930. Scholle’s Index of 20 Ordinary Shares provides weekly data between 1921 and 1926. Daily and weekly data are derived from data for the FT 30 Industrials from 1930 to 1969, and from the FTSE-All Share Actuaries index from 1970 on. Monthly and annual data are derived from a number of broad stock market indices, including the FTSE All-Share Actuaries index at present.

What is surprising about the data from London is the relative absence of volatility from the 1930s. Only two of the best and worst days are from the 1930s. On the other hand, the 1973-1974 bear market and its recovery in 1975 dominate the daily and weekly movements, though October 1987 included four of the worst days. On a weekly basis, the early 1930s show up more strongly, with four of the best weeks from 1931-1932.

The monthly and annual data for London are available over a 300-year period. This allows us to view the ups and downs for the stock market in the past three centuries using data that are not available from any other market. Several events stand out from the 1700s and 1800s. The South Sea Bubble in 1720-1721 produced several of the best and worst months in stock market history. The Napoleonic wars added further volatility, and the South America Mining Stock bubble of 1824-1826 produced similar dramatic rises and falls. The most dramatic events of the 20th Century were the 1973-1974 bear market and its recovery, producing four of the worst months in 1974 and four of the best months in 1975. The October 1987 crash and the decline after the fall of Dunkirk produced two of the other worst events in London Stock Market history.

The dramatic impact of the 1973-1974 bear market on London is also illustrated in the annual data. It should be remembered that this bear market was worse than the 1929-1932 bear. The decline in 1974 and the recovery in 1975 dwarf all other years. Only the 1824 bubble comes close to matching the 1975 rise, and after the bursting of the South Sea Bubble in 1721, 1973’s decline is the steepest.


Three sets of indices are provided for the United States. The Dow Jones Industrials provides the longest series of daily data available anywhere. The current Dow Jones Industrials goes back to 1914. Similar daily series extend the daily record back to 1885.

The S&P Composite is a broader index that includes daily data for 500 stocks back to 1957 and for 90 stocks back to 1928. Using the Cowles Commission data and other series, we have extended the data for the United States back to 1800 on a monthly basis.

The NASDAQ Composite has daily data back to NASDAQ’s founding in 1971. Before that we have used data from the National Quotation Board indices to provide daily data history back to 1950.

Data for the Dow Jones Industrials is daily back to 1885. Data for the S&P Composite is monthly to 1918, weekly from 1919 to 1927, and daily from 1928 on. Data for the NASDAQ/NQB series is monthly from 1938 to 1949 and daily from 1950 on.

The legacy of the Great Depression can be seen in the data from the Dow Jones Industrials Average. The majority of the best and worst days and weeks are from the 1930s. October 19, 1987 holds the dubious record of being the worst day in US stock market history. The recovery from the October 1987 crash is the only up day to make the top 20 since 1940. On the other hand, six of the worst days in Dow history have occurred since 1985. The terrorist attacks of September 11 produced the fourth worst week in Dow history, even worse than October 1987.

The S&P Composite provides a similar record on a daily and on a weekly basis, but the extension of the data back to 1800 allows us to discover events that would have gone unnoticed in the DJIA data. For example, two of the best years in US stock market history were 1862 and 1863 as inflation and expectations of war profits drove stocks up. 1843 saw the Depression that had started in the 1830s come to an end, and in 1879, the United States returned to the Gold Standard, revising confidence in the economy. The recovery from an oversold market in 1933 and 1935, and the bull market of the 1950s produced other bull market wonders.

The years 1930, 1931, 1932 and 1937 all rank among the worst years in US stock market history. Financial crises in 1854, 1857 and 1907 produced bear market years. 1973 and 1974 represent the two worst years in the market since 1950.

Data for NASDAQ/NQB is only available back to 1950 on a daily basis. With the exception of October 1987, the Technology bubble of 1998-1999 and the bursting of the Technology bubble in 2000-2002 produced virtually all of the best and worst days and weeks in NASDAQ’s history.


The MSCI World Index has been calculated since 1969, and we have extended the index back to 1919. We have used daily data for the MSCI World Index back to 1979. Since the World Index is a combination of all of the world’s indices, it can focus on the events that affected global stock markets and factor out purely national events. Since the United States represents almost 50% of the world index, its influence on the index should not be ignored.

On a daily and weekly basis, the crashes of October 1987, October 1998, and August-September 1990 clearly figure in both the best and worst categories. It is interesting that after October 1987, five of the worst weeks since 1980 have all occurred in the 2000-2002 bear market.

Monthly and annual data provide more perspective since they provide over 80 years of history. The 1930-1932 period provides the most number of best and worst months in this index. Recoveries from sell offs dominate the best months, and short-term crashes dominate the worst months.

1933 represents the best year in the World index’s history, but who would have guessed that after that 1954, 1958, 1985 and 1986 would have played such a prominent role? Each of these years is in the middle of a bull market when all of the world’s stock markets were doing well producing some of the overall best years in market history.

The worst years come as no surprise. 1920, 1921, 1930, 1931, 1937, 1946, 1973, 1974, 1990, 2000 and 2001 are all years associated with global bear markets.

REQUEST A DEMO with a GFDFinaeon Specialist

Please type your first name.
Please type your last name.
Please type your phone number in the following format 123-456-7890
Invalid email address.
Please type your company name.
Invalid Input


Our comprehensive financial databases span global markets offering data never compiled into an electronic format. We create and generate our own proprietary data series while we continue to investigate new sources and extend existing series whenever possible. GFD supports full data transparency to enable our users to verify financial data points, tracing them back to the original source documents. GFD is the original supplier of complete historical data.