Saturday, March 21, 2009

The 1929 Stock Market Crash Revisited

Worse Than Worst-Case Scenario?


What if we had a repeat of the stock market crash of 1929 -1932? In a previous worst-case post, I developed a worst-case scenario with the DJIA (Dow Jones Industrial Average) closing in the neighborhood of 4100. An alert reader suggests that is optimistic! He maintains that, if we were to repeat the 1929 crash, the Dow would fall to 2100.

Why the difference? The short answer is because my previous worst-case posts used very high-level, year-end data. The point I was trying to make in the earlier posts was that it would be prudent to at least consider the possibility of the Dow at levels that seemed unimaginable at that time. In this post, we will analyze daily data.

The 1929 (Great Depression) Stock Market Crash Graph



Daily chart of 1929-1932 (Great Depression) stock market crash

As you can see in the chart above (click to enlarge), the Dow peaked in September of 1929 and continued "steadily" downward through 1930 and 1931 until it bottomed in July of 1932; the Great Depression continued for some years afterwards. (Note: To see how this period fits in the bigger picture, see 100 Years of Stock Market History. If you are not familiar with log graphs like this one, see About Stock Market Log Graphs) The bottom line? A $10,000 investment in 1929 was reduced to about $1100 in 1932 -- a loss of almost 90%! Some initial observations:
  • It took almost three years to reach the bottom. People think of the 1929 crash as being a dramatic drop. Well, it was; the Dow dropped almost 50% in just over two months. However, at that point, the decline was far from over in terms of either magnitude or duration. The market continued to decline for another two and a half years -- through all of 1930, all of 1931, and half of 1932.
  • If we look only at year-end numbers, the Dow dropped from 248 to 60, a drop of about 76%. However, the drop from the September 1929 high of 381 to the July 1932 low of 41 was an incredible 89%; looking only at year-end numbers misses about 13% of the drop.

Secular Bear Markets and Bear Market Rallies!


What I didn't fully understand until I looked at the daily chart was how agonizingly frustrating it must have been. The market changed direction about a dozen times -- typically every one to three months. During this approximately three-year period:
  • There were at least six "false starts" that might have led an investor to believe that the bear market was finally over. Five of the rallies would qualify technically as bull markets -- the market rose more than 20%. All of these were "bear market rallies;" the resulting stock market highs were lower than the previous highs, and the following lows were lower than the previous lows.
  • It's probably worth pointing out that each of the seven major declines was a bear market on its own; each was greater than 20%; six were greater than 35%. Think about it. In a period of less than three years, a typical investor lost more than 35% of his assets six separate times. Three of the declines were more than 40% (another was 39%).
  • There were only two periods when the market maintained its direction for more than three months:
    • The initial 2-month 50% drop was followed by a 5-month, 50% rally that was probably very reassuring, but ultimately just the first of many "head fakes" (reminder: it takes a 100% increase to erase a 50% loss)
    • A four-month bear market that took the market from 89 to the final low of 41 -- a greater than 50% loss, and from a point where you have to believe many people felt things could not possibly get any worse; this decline was even greater than the more famous initial crash, and was the final capitulation.
Note: For an analysis of the relative contribution that earnings, dividends and valuation made to the drop in prices during this historic three-year period, see Analyzing the Stock Market Crash 1929-1932.

How Many Years After the Crash Did It Take To Recover to Pre-Crash Levels?


You can see from the 100-year chart that, on a year-end close basis, it took until 1954 for the stock market to reach its pre-crash peak year-end close of 300; from 1928 to 1954 is 26 years. Using inflation-adjusted year-end closing prices, it took until 1955. However, here's a case where you might want to look at more granular numbers.

The pre-crash peak daily close was 381.17 on September 3, 1929. On a nominal basis, the market had still recovered by year-end 1954 (to 404.39). However, now the inflation-adjusted picture changes. Using my inflation calculator, I calculate that 381 in 1929 is roughly 4909 in current dollars. In current dollars, the first year-end close above 4909 is the close of 5106 in 1959. So, it probably took roughly 30 years to recover to the daily peak. (I say "probably" because I haven't looked at inflation-adjusted daily/monthly prices in the 50s)

What If We Have a Repeat of the Stock Market Crash of 1929-1932? --- An Even Scarier Graph


The question the reader raises is, what if we had another stock market crash "just like" 1929.

Graph of 1929-1932 stock market crash compared to 2007-2008-2009-2010

The chart above (click to enlarge) compares the stock market crash of 2008 to the crash of 1929. The solid line is actual daily Dow closing prices from January 1, 2007 through March 17, 2009. The dotted line is 1929-1932 closing prices moved to 2007-2010, and multiplied by about 37; the 2007 high was (14165 divided by 381=) about 37 times the 1929 high. If you compare the graphs of the two crashes, you can see that the paths were significantly different. For example, in the 1929 crash, the "crash" occurred in the same year as the peak; the 2008 crash occurred the year after the peak.  However, coincidentally, in both cases March was approximately 50% from the peak.

The bottom line is, if the Dow again declined by 89.2% from its peak daily close, as it did in the 1929 - 1932 crash, the Dow would go to 1532! Note that this is even worse than the reader's worse-than-worst case. So, it appears that the reader was also being optimistic.

Some Final Comments


This is NOT a forecast. You can't generalize based on what is essentially one observation. On the other hand, I'm not going to ignore it. Could it happen again? Of course -- it's the future.

Related Materials

The 25 Best & Worst Years in Stock Market History
The Impact of Large Stock Market Losses: What percent gain is required to offset losses this large?
100 Years of Stock Market History (graph): where the crash fits in the big picture.
Stock Market Returns Since 19XX: Contrasts returns-to-date of investors who bought before the crash with those who bought afterwards.
Dow Price/Earnings (p/e) Ratio History Since 1929 - Yearly Graph. A contributing factor to the crash.
Looking Forward
Projecting Stock Market Returns Estimating future market returns.
Three Scenarios for the Economy (and the Stock Market): scenarios going forward.
For a list of other popular posts see the sidebar on the left. For an index of stock market posts by subject area, see the blog header.

If after reading this post you need an antidote, try Barton Biggs' article -- in the March 16, 2009 issue of Newsweek it was entitled "Brighter Days Ahead."
The source for my daily data is here

Share This Article

Delicious
Bookmark this on Delicious
See the "Share" option in the Menu Bar at the top of the page to share via Facebook, Twitter, etc.

Copyright © 2009. Last updated 8/15/2011

4 comments:

  1. Can you please update the graph per the latest trends? Do you still think we're headed for < Dow 2K? Also, wouldn't it be better to focus on the differences between '29-'32? Are the fed and govmt actions siginficant enough? Do you see the same kind of mass-panic that we had back then?

    ReplyDelete
  2. Anon,

    Thanks for your comment.

    First, a clarification. I didn't mean to suggest that I thought we were headed for a repeat of 1929-32. As I said at the end of the post, "This is NOT a forecast." Rather, it was only a possibility -- one of several worst-case scenarios. And, as I said in the first worst-case post, worst-case scenarios are unlikely, almost by definition.

    I'm concerned that continually updating the graph might send the wrong message and contribute to any misconceptions that it is a forecast. (Besides, the main point was the bottom line. If we were ever to go that low, the actual path to get there would be completely different.) I think your other questions point the way to a more effective way to monitor the situation -- in effect to monitor differences AND similarities between important economic and psychological factors. In a sense, that was the point of the Three Scenariospost. That post identifies three generic scenarios for the economy (and the stock market) -- Business as Usual, Headwinds, and Snowball. It also includes a summary of the factors that I think will ultimately determine which scenario plays out (including the ones you mention). Periodically, I have been updating that post with links to relevant articles to help readers stay current as the situation evolves.

    A short response to your specific questions: No, I do not see the same panic as in 1929. I think the government's actions have been significant, and more effective than in 1929. However, I'm not sure what "significant enough" means. Enough to stop the panic? Yes. Enough to completely eradicate the problems? Probably not. However, I expect there will be additional government action as we go forward. As a result, I still believe "Headwinds" is the most likely scenario -- though we will likely continue to see some elements of the other scenarios as well. At some point, I may do an updated Three Scenarios post. In the meantime, you might find that post, and the linked-to articles interesting.

    I hope this helps.

    ReplyDelete
  3. This occured in the eary 30's becuase credit was locked up and nations became protectionist.

    It may not be credit this time but let's be very careful to watch: oil prices, commodity prices and currency fluctuations. As the world is more tightly inegrated these fundamentals are almost like a tax or could drastically hold back a recovery.

    One other thought about these false starts. People rush to the share market on a flase boom to leverage the market to be able to deleverage their other assets which are falling. The lucky ones take profits and the market falls again.

    ReplyDelete
  4. Hi Observations-- I remember the markets of l950 on and am interested in your Wors Case repeat of l929 chart--however, it can gain more if you used your 100 year chart for "comparison" and standarxd Technical Analysis charting methods. In this computer age, it would be easyu for you to chart the uper-lower bands and place the Twenty Nine Crash is right perspective witdh room for another 20 or more years "ahead room" for a potential glimmer of the Future. I can do this on paper and you willdo even a better "prophecy" on screen.
    The main thing you will find, I believe, is that the lower 2000 parallell line will show thes= 100 year trend (relative to l929 bottom) is of course still up. But if your hypothetical big 2012+ Crash were to come, the question would be... "Would the 100 year trend show "the new bottom, say about 5000 DJI, adjusted... or would there be a lower band violation that would send the Dow even lower than the bottom of l929.. say 1000 or 3000. I'd lovd to see your Tech. work or this and maybe gain some insight as to what to do with my own stocks and life style. I want my remarks published for all to see, but you have my email on your screen and might want to drop me an e-line if you want some more comments on the 2000-2012 DJI technical pattern being formed--perhaps a double bottom at 7000 or a W breakout past 13000 heading for 20,000DJI in a few years. SOH

    ReplyDelete

No spam, please! Comment spam will not be published. See comment guidelines here.
Sorry, but I can no longer accept anonymous comments. They're 99% spam.