What Does History Have to Tell Us About The Stock Market?
Today the market dropped over 600 points on the Dow. The broader S&P 500 which I will refer to as the market for simplicity's sake dropped 6.66% to close at 1119.46. With all the breathless commentary about the market's latest swoon, it's useful to put it in historical perspective. What follows are some observations based on every trading day in the S&P 500 since 1980, in Question and Answer format.
Question: Was today's drop a record?
Answer: No, it wasn't in the top 10, but it was the 11th greatest percentage drop in the S&P 500 since 1980.
The market fell over 3 times as much during the market crash of October 19, 1987. Here are the top 12 percentage declines since 1980 with today's decline highlighted in purple:
| Rank: |
| Date | S&P 500 | 1 day % change: |
| 1 |
| 19-Oct-87 | 224.84 | -20.47% |
| 2 |
| 15-Oct-08 | 907.84 | -9.03% |
| 3 |
| 1-Dec-08 | 816.21 | -8.93% |
| 4 |
| 29-Sep-08 | 1,106.42 | -8.81% |
| 5 |
| 26-Oct-87 | 227.67 | -8.28% |
| 6 |
| 9-Oct-08 | 909.92 | -7.62% |
| 7 |
| 27-Oct-97 | 876.99 | -6.87% |
| 8 |
| 31-Aug-98 | 957.28 | -6.80% |
| 9 |
| 8-Jan-88 | 243.40 | -6.77% |
| 10 |
| 20-Nov-08 | 752.44 | -6.71% |
| 11 |
| 8-Aug-11 | 1,119.46 | -6.66% |
| 12 |
| 13-Oct-89 | 333.65 | -6.12% |
Two things about this table:
- Despite the decline, the S&P 500 is still higher than at any time following a similar magnitude decline; it's almost 5 times as high as it closed on October 26, 1987. Since the index does not reflect reinvested dividends, this understates how much an invested dollar would have grown since then.
- The single program-trade-driven decline of the Crash of 1987 really stands out. The second greatest decline was not half this 20% plunge. In fact no other plunge was in double digits, percentage-wise. Note to journalists covering the market: please do not ever compare point declines, only percentage declines. Thank you.
Question: But the market fell sharply last week; is the total decline over the past 3 trading days (Thursday, Friday, and Monday) not one of the greatest in stock market history?
Answer: Yes, the past 3-trading-day decline of 11.18% came it at #7 since 1980:
| Rank: |
| Date | S&P 500 | 3 days: |
| 1 |
| 19-Oct-87 | 224.84 | -26.34% |
| 2 |
| 20-Oct-87 | 236.83 | -20.55% |
| 3 |
| 9-Oct-08 | 909.92 | -13.91% |
| 4 |
| 26-Oct-87 | 227.67 | -11.89% |
| 5 |
| 31-Aug-98 | 957.28 | -11.71% |
| 6 |
| 20-Nov-08 | 752.44 | -11.56% |
| 7 |
| 8-Aug-11 | 1,119.46 | -11.18% |
| 8 |
| 7-Oct-08 | 996.23 | -10.59% |
| 9 |
| 8-Oct-08 | 984.94 | -10.40% |
| 10 |
| 16-Oct-87 | 282.70 | -10.12% |
Notice how October (7 out of 10) dominates this table, especially October '87 (4) and '08 (3).
Excluding non-overlapping 2-day periods, the rank of this 2-day decline rises to #6.
Question: Well, what about a full week of trading (5 trading days)? Wasn't this one of the worst week in stock market history?
Answer: It was bad, but didn't quite make the top 10:
| Rank: |
| Date | S&P 500 | 5 days: |
| 1 |
| 19-Oct-87 | 224.84 | -27.33% |
| 2 |
| 20-Oct-87 | 236.83 | -24.70% |
| 3 |
| 9-Oct-08 | 909.92 | -18.34% |
| 4 |
| 10-Oct-08 | 899.22 | -18.20% |
| 5 |
| 20-Nov-08 | 752.44 | -17.43% |
| 6 |
| 22-Oct-87 | 248.25 | -16.72% |
| 7 |
| 21-Oct-87 | 258.38 | -15.35% |
| 8 |
| 8-Oct-08 | 984.94 | -15.17% |
| 9 |
| 7-Oct-08 | 996.23 | -14.59% |
| 10 |
| 27-Oct-08 | 848.92 | -13.85% |
| 11 |
| 8-Aug-11 | 1,119.46 | -13.01% |
| 12 |
| 23-Oct-87 | 248.22 | -12.20% |
Question: What about 2-week (10-trading-day) periods? Surely, this has been one of the worst declines ever, correct?
Answer: It was up there, but didn't make the top 10. It came it at #15. Over the past 10 trading days, the market declined 16.30%.
| Rank: |
| Date | S&P 500 | 10 days: |
| 1 |
| 19-Oct-87 | 224.84 | -31.47% |
| 2 |
| 26-Oct-87 | 227.67 | -26.41% |
| 3 |
| 10-Oct-08 | 899.22 | -25.88% |
| 4 |
| 27-Oct-87 | 233.19 | -25.86% |
| 5 |
| 20-Oct-87 | 236.83 | -25.81% |
| 6 |
| 9-Oct-08 | 909.92 | -24.75% |
| 7 |
| 28-Oct-87 | 233.28 | -23.57% |
| 8 |
| 15-Oct-08 | 907.84 | -21.81% |
| 9 |
| 22-Oct-87 | 248.25 | -20.98% |
| 10 |
| 23-Oct-87 | 248.22 | -20.20% |
| 11 |
| 21-Oct-87 | 258.38 | -18.89% |
| 12 |
| 29-Oct-87 | 244.77 | -17.88% |
| 13 |
| 8-Oct-08 | 984.94 | -16.94% |
| 14 |
| 20-Nov-08 | 752.44 | -16.85% |
| 15 |
| 8-Aug-11 | 1,119.46 | -16.30% |
However, it was only about half of the 31.47% 10-day decline ending on October 19, 1987.
Question: Do major 2-week sell-offs indicate buying opportunities or sell signals? In other words, how did the market perform in the past AFTER it sold off this much?
Answer: There are countless ways of trying to answer this question - perhaps the easiest is simply to look at the 14 cases in which the market's 10-day decline was greater than the 16.30% of the past 10 days and measure how it rose or fell over the next 10 days, 20 days, and 250 days (remember, these are trading days, so represent 2 weeks, a month, and a year, roughly):
|
|
| Average: | 2.94% | 2.03% | 18.54% |
|
|
|
|
|
|
|
| Date | S&P 500 | 10 days: | Next 10d: | Next 20d: | Next 250d: |
| 19-Oct-87 | 224.84 | -31.47% | 13.75% | 9.75% | 22.41% |
| 26-Oct-87 | 227.67 | -26.41% | 6.81% | 6.73% | 24.25% |
| 10-Oct-08 | 899.22 | -25.88% | -2.50% | 3.53% | 18.49% |
| 27-Oct-87 | 233.19 | -25.86% | 2.49% | 5.66% | 21.64% |
| 20-Oct-87 | 236.83 | -25.81% | 5.91% | 2.62% | 16.33% |
| 9-Oct-08 | 909.92 | -24.75% | -0.20% | -0.55% | 16.23% |
| 28-Oct-87 | 233.28 | -23.57% | 3.70% | 4.64% | 21.00% |
| 15-Oct-08 | 907.84 | -21.81% | 2.45% | -6.12% | 18.21% |
| 22-Oct-87 | 248.25 | -20.98% | 2.51% | -3.30% | 12.54% |
| 23-Oct-87 | 248.22 | -20.20% | 0.88% | -2.51% | 11.58% |
| 21-Oct-87 | 258.38 | -18.89% | -3.65% | -4.97% | 6.98% |
| 29-Oct-87 | 244.77 | -17.88% | 1.53% | -1.81% | 15.37% |
| 8-Oct-08 | 984.94 | -16.94% | -8.95% | -3.27% | 7.08% |
| 20-Nov-08 | 752.44 | -16.85% | 16.43% | 18.00% | 47.49% |
| 8-Aug-11 | 1,119.46 | -16.30% |
|
|
|
This is a busy table but it shows the market rebounded on average 2.94% over the 2 weeks following a decline of the past 10 days or greater, a snap-back that was greater and more consistent (only 4 down periods) for the subsequent month (or 20 trading days) which averaged 2.03% and was down roughly half the time. Notice, however, that the market rose every year (250 trading days) following 10-day declines of the last 10 days or greater with 5 years climbing over 20% and only 2 rising less than 10%.
Someone reading this table closely might notice that it may overstate the returns an investor would receive since one could only really invest once at some point in October, 1987, and hold for the next year, and 9 of the greatest 10-day declines occurred in this month, so since we now know that the world didn't end following the Crash of '87 and the market was sharply higher a year later, this may skew the average. Nevertheless, look at the stunning 47.49% advance that followed a year after the 10-day decline ending November 20, 2008.
So to be rigorous, let's broaden our net and look at all declines that were roughly equal to the past 10 days' decline of 16.30%, say declines greater than 14%:
|
|
| Average: | 3.01% | 3.93% | 21.02% |
|
|
|
|
|
|
|
| Date | S&P 500 | 10 days: | Next 10d: | Next 20d: | Next 250d: |
| 19-Oct-87 | 224.84 | -31.47% | 13.75% | 9.75% | 22.41% |
| 26-Oct-87 | 227.67 | -26.41% | 6.81% | 6.73% | 24.25% |
| 10-Oct-08 | 899.22 | -25.88% | -2.50% | 3.53% | 18.49% |
| 27-Oct-87 | 233.19 | -25.86% | 2.49% | 5.66% | 21.64% |
| 20-Oct-87 | 236.83 | -25.81% | 5.91% | 2.62% | 16.33% |
| 9-Oct-08 | 909.92 | -24.75% | -0.20% | -0.55% | 16.23% |
| 28-Oct-87 | 233.28 | -23.57% | 3.70% | 4.64% | 21.00% |
| 15-Oct-08 | 907.84 | -21.81% | 2.45% | -6.12% | 18.21% |
| 22-Oct-87 | 248.25 | -20.98% | 2.51% | -3.30% | 12.54% |
| 23-Oct-87 | 248.22 | -20.20% | 0.88% | -2.51% | 11.58% |
| 21-Oct-87 | 258.38 | -18.89% | -3.65% | -4.97% | 6.98% |
| 29-Oct-87 | 244.77 | -17.88% | 1.53% | -1.81% | 15.37% |
| 8-Oct-08 | 984.94 | -16.94% | -8.95% | -3.27% | 7.08% |
| 20-Nov-08 | 752.44 | -16.85% | 16.43% | 18.00% | 47.49% |
| 8-Aug-11 | 1,119.46 | -16.30% |
|
|
|
| 23-Jul-02 | 797.70 | -16.28% | 7.76% | 17.52% | 22.70% |
| 7-Oct-08 | 996.23 | -16.16% | -4.13% | 0.96% | 4.44% |
| 22-Jul-02 | 819.85 | -16.08% | 1.80% | 15.96% | 21.16% |
| 27-Oct-08 | 848.92 | -15.39% | 8.28% | 0.34% | 27.17% |
| 19-Nov-08 | 806.58 | -15.34% | 4.79% | 9.76% | 37.66% |
| 2-Mar-09 | 700.82 | -15.24% | 7.57% | 12.37% | 57.60% |
| 16-Oct-08 | 946.43 | -15.06% | 0.81% | -3.71% | 15.38% |
| 21-Sep-01 | 965.80 | -14.80% | 10.93% | 11.15% | -12.68% |
| 18-Nov-08 | 859.12 | -14.58% | 1.35% | 5.27% | 29.12% |
| 17-Oct-08 | 940.55 | -14.44% | 3.00% | -7.15% | 16.59% |
| 14-Oct-08 | 998.01 | -14.43% | -5.76% | -9.93% | 7.83% |
| 23-Feb-09 | 743.33 | -14.42% | -8.99% | 10.71% | 49.22% |
| 19-Jul-02 | 847.75 | -14.28% | 1.95% | 9.56% | 15.80% |
| 21-Nov-08 | 800.03 | -14.07% | 13.71% | 8.95% | 36.86% |
Notice the market did even better here, rising 21.02% on average the next 250 trading days, and falling only once (2001).
Question: Are these numbers that big of a deal? Doesn't the market usually rise on average over any given year?
Answer: Yes, they are a big deal because although the market does have a tendency to rise, it's nothing close to 20%+ a year. Here are the averages for all rolling (overlapping) 10, 20, and 250-day periods for comparison:
|
| Next 10 days | Next 20 days: | Next 250 days: |
| Average, 1980-8/8/11: | 0.36% | 0.73% | 9.48% |
| Average following 10-day declines of 14% or more: | 3.01% | 3.93% | 21.02% |
| Average following 10-day declines greater than the last 10-day decline: | 2.94% | 2.03% | 18.54% |
So if the past is any indication, and it's the only one we have, sharp declines are followed by above average returns in subsequent 10, 20, and 250-day time periods. Remember these are averages, and the market can do whatever it wants. Caveat emptor. Also, just because the market was higher a year later does not mean it didn't decline further along the way (I did not look at that here, only the next 10, 20, and 250 days of trading). It may in some cases have made a new low 60 or 80 or 120 days later, but that would not show up in these numbers.
Question: Does this mean that one should buy a sharply declining market?
Answer: Not necessarily (I certainly don't, waiting instead for indications the market has bottomed and is trying to rally). But the historical data seem to indicate that by the time a market has sold off dramatically to be newsworthy, much of the damage has already been done and the market tends to rise subsequently. Of course, there are exceptions, such as the 73-74 bear market (not in this time span of course), in which the market inexorably ground down for the better part of 2 solid years (followed a third very strong year, then more backing and filling).
Question: But how can the historical record help if the global economy is so much different today? The United States is holding high levels of debt, leadership in Washington does not seem to be able to do anything about it except blame the other side, and the economy is very weak with relatively high unemployment. China wasn't even a global player much less the second largest economy and the Euro with all its troubles didn't exist until a few years ago. Maybe we're going into a double dip recession or something worse.
Answer: Whenever the market has sold off in the past, the news about the economy or whatever triggered the sell-off was scary and unique. We had never had a program-trading driven crash until we did. We never had an Asian contagion or a mortgage-backed security implosion until we did. It's always different but some things remain constant: markets tend to sell off, panic, and rally to markedly different news in markedly similar ways.
That being said, those predicting the end of the United States, Western Civilization, and a crash of the United States economy, blah, blah, blah, have been doing this every time the market sold off. The fact that one day they might be right is not helpful (since in that case it doesn't matter how you invest), but the odds are we will muddle through this the same way we have through past crises. The United States economy has been moribund and written off many times before but we remain the world's largest economy with advantages that the EU - which has a single currency but no single, harmonious monetary policy as we do - especially its northern members - can only dream of right now. There are many advantages of being a declining power, the most notable of which is that the United States may wind down its massive military expenditures, more than the rest of the world combined, which would mean all that money would be freed up to invest in schools, roads, and infrastructure. The losers of World War II, especially Germany and Japan, grew far more rapidly and had higher stock market returns than the winners.
Question: But isn't the United States debt burden strangling any economic growth?
Answer: There is no evidence that running a surplus is correlated with economic growth or that running deficits is correlated with economic contraction (except for a lagging relationship, that is, a growing economy, all things being equal, tends to increase government revenues (taxes), thereby lowering deficits, whereas a contracting economy contracts revenue collected at the same time demand for federal government services for everything from unemployment insurance to bailing out of state and municipal governments - surges. The Tea Party repeats this idea so often you would think they would have data to back it up but they don't.
This economic contraction is unique in that after an initial bailout of Wall Street under Bush to the tune of $700 billion and a subsequent $700 billion of stimulus money disbursed under the Obama administration, the government is slashing rather than expanding assistance to the economy. This almost always is disastrous economically but it has not been done since Herbert Hoover in this country (and arguably in Japan after they prematurely stopped stimulus spending the in the early 1990s, leading to another decade of economic stagnation).
There are two separate dramas unfolding in the markets. The first is the dramatic slowdown in the rate of growth including revisions of past growth announcements. That has been the main cause of the sell-off in the stock market, the best leading economic indicator of future economic growth.
The other drama involves debt, but that is a political drama that will probably have a side effect of prolonging the Great Recession or turning it into another depression as Washington cuts off the spigots. (Private industry is actually creating jobs, although at an anemic pace; it is the dramatic, accelerating shedding of state and federal government jobs that is leading to stubbornly high unemployment.) As long as people are piling into US Treasury bonds, willing to lend our government money for 10 years at less than 3%, the idea of a debt crisis is ludicrous.
Greece, Spain, Portugal, and Italy are suffering debt crises and their cost of borrowing reflects that. The United States may face a limit to how much it can borrow but it most decidedly has not reached that limit if our cost of borrowing is any indication. The Chinese can lecture us about spending less on social security programs but they are still first in line to gobble up a billion dollars of US debt every day at ridiculously low interest rates.
As Warren Buffet recently pointed out, the United States has always and will always repay its debt. The imbalance between taxes and government expenditures is almost entirely accounted for by 3 reversible factors, 2 of them immediately amenable to change: the Bush tax cuts which have driven federal revenue as a percent of GDP to record lows, the Bush wars in Afghanistan and Iraq, and the aftermath of the 2008 financial collapse triggered by Bush era deregulation of, among other things, mortgage-backed securities.
Question: But didn't Standard & Poors downgrade our debt? Doesn't that mean the United States does have a debt problem?
Answer: No, it means it has a political paralysis problem. The behavior of Congress, particularly the Tea Party freshmen who refused to negotiate in good faith and created a down-to-the-wire drama by doing something no one had ever done before - threatening to renege on government programs and services already promised if they didn't get their way - was cited by Standard & Poors, not necessarily the level or manageability of the debt, which could disappear in a few years if the United States government agreed on a combination of tax increases and decreases in massive military spending. Entitlement reform is a longer term issue but even small tweaks would change projections from red to black. Only a minority of federal spending is discretionary, but that is where all the effort of both President Obama and the Tea Party has been focused to date. All other options were ruled out a priori by the Tea Party and the GOP. One could drive discretionary spending to zero and not balance the budget. One could roll back tax rates to the level of Clinton and end one of the two ongoing wars and not touch discretionary spending and go into surplus. The self-imposed ideological straitjackets are going to make it virtually impossible for the government to take decisive, concerted, sustained action to stimulate the economy and fill in the slack until private industry can get on its feet again and start hiring. The Tea Party has made stimulus a dirty word, so until those who got us into this mess are voted out or impeached, it's unlikely Congress will allow the government to function in any meaningful way.
