The Nikkei took thirty-four years to reclaim its 1989 high. The lesson was never the crash. It was what the country refused to do afterward, and what that refusal cost a generation of savers who were told to stay the course.
See how long the road back wasYears from a major market's peak to the day it closed above that peak again, in nominal terms. The bars are scaled to Japan. The last row is the one that has not been written yet.
Two things to notice. Every US road was shorter than Japan's, and each was shorter than the last, because the United States recognized its losses faster each time. And the two US roads that began at a valuation like today's, 1929 and 2000, were the two long ones. Dates are index closing highs; dividends would shorten each road somewhat and do not change the ordering.
The 1985 Plaza Accord pushed the yen sharply higher and hurt Japanese exporters. To offset it, the Bank of Japan cut rates to historic lows and left them there. Cheap money went into stocks and land. Banks lent against land as collateral, land rose, collateral rose, and lending rose again. By the end of 1989 the Nikkei stood near 39,000 and Japanese equities traded at roughly sixty times earnings. Companies owned each other's inflated shares through cross-holdings, so the whole system was leveraged to itself.
A new Bank of Japan governor raised rates five times in about fifteen months and the government capped real estate lending. The Nikkei fell roughly sixty percent by 1992. Land prices fell for most of the next fifteen years. This part, the part everyone remembers, took three years.
Banks did not write down their bad loans. They rolled them over, kept insolvent borrowers alive, and starved healthy companies of credit. Serious recapitalization did not arrive until roughly 1998 to 2003, a decade after the peak. Prices began falling, which made every yen of debt heavier and gave every household a reason to wait rather than spend. Stimulus was applied in bursts and reversed before it took, most famously with the 1997 consumption tax increase. And the working-age population peaked in the mid-1990s, so there were fewer workers each year to grow the country out of what it owed.
The crash cost Japan three years. The refusal to recognize losses cost it thirty-one more.
Most versions of this comparison stop at the bubble. The bubble is the least useful half. What made Japan's road long was the aftermath, and on the aftermath the two countries genuinely differ.
Valuation. Japan in 1989 paid about sixty times earnings. The S&P 500's CAPE is above forty, a level seen only in 1929 and 2000.
A story that justifies the price. Japan had a national model that could not lose. The US has AI capital spending. Both require the future to arrive on schedule.
Concentration. Japan's index was bound up in cross-holdings among a few banks and industrials. The US index is bound up in a few technology companies whose earnings depend on each other's spending.
A long stretch of cheap money. The Bank of Japan held rates down to fight a strong yen. The Federal Reserve held rates near zero for most of 2009 to 2021 and bought bonds on top of it.
Loss recognition. US banks were forced to recapitalize inside two years in 2008 and 2009. US bankruptcy law clears dead companies rather than propping them up.
Inflation, not deflation. Japan's debts got heavier in real terms as prices fell. The US has run inflation above target and its borrowing costs have been rising, which reprices assets faster and more visibly.
Demographics. Japan's working-age population peaked in the mid-1990s. The US population is still growing, so there is underlying growth to eventually justify prices.
The reserve currency. Japan funded itself at home. The US can run large deficits because the world buys Treasuries. That is a cushion, not an unlimited one.
Read together: the US is unlikely to get Japan's thirty-four years. It is quite capable of getting 2000's thirteen. For a household drawing income, the difference between a lost decade and a lost three is smaller than it looks, because the withdrawals happen in the first ten years either way.
The comparison is not a feeling. It is four measurable conditions that decide how deep a fall goes and how long the climb back takes: how expensive stocks are, how much of the index rides on a handful of names, how much room the Federal Reserve has to cut its way out, and how much room the federal government has to borrow its way out. On every one of the four, today's starting line is worse than 2000's and worse than 2008's.
| Condition | 2000 peak | 2008 peak | Now |
|---|---|---|---|
| Valuation, Shiller CAPE | 44.2 | 27.5 | 41.6 |
| Top 10 stocks, share of the index | ~27% | not the era's fault line | ~41% |
| Fed funds rate, room to cut | 6.50% | 5.25% | 3.50–3.75% |
| Federal debt, share of GDP | ~55% | ~70% | ~123% |
2000 was a valuation and concentration bubble; 2008 was a credit and leverage bubble, which is why top-10 index weight is marked differently for that year rather than forced into a comparison it was never about. CAPE is the S&P 500 Shiller ratio at each period's market top. Fed funds is the target rate in the months immediately before each downturn began, the room the Fed had available before it needed to start cutting. Federal debt is gross federal debt as a share of GDP, rounded, for the same three points. Full sourcing is in the Sources section below.
The damage from a bubble is set by the leverage and the valuation going in. The length of the road back is set by how fast the system is willing to admit what it lost.
Unlike 1989 Tokyo or 2000 Silicon Valley, the companies at the top of today's index are enormously profitable and are funding their spending from cash flow, not debt. A high multiple on real earnings is a different animal from a high multiple on a story.
The Bank of Japan waited years to cut to zero. The Federal Reserve cut to zero within months in 2008 and again in 2020, and it has shown it will buy assets in size. A lender of last resort that acts early shortens every road.
Japan had no growth to grow into. The US working-age population is still expanding, productivity is rising, and the economy has absorbed every valuation excess in its history within a generation.
The Shiller ratio crossed 30 in 2017 and the index has roughly doubled since. A measure that was early by nine years is not a measure most people can invest by.
Every one of those is true, and every one of them was also true, in its own form, in March 2000: real profits at the top of the index, a Fed that had just proved it would act, the strongest demographics in the developed world, and a valuation measure that had been "wrong" for years. The road back still took thirteen years. The objections shorten the road. They do not remove it, and none of them tells a household how many of those years it can afford to spend drawing income from a portfolio that is underwater.
I will say plainly what the chart above leaves open. I believe a reset is coming that will be larger than any in my lifetime, larger than 2008, and that it will be measured the way Japan's was: not by how far the index falls in a year, but by how many years it takes to come back. I do not know the date and I will not pretend to. What I know is that the starting valuation is the second highest in American history, that the last two roads that started from here were the two long ones, and that very few of the households I talk to have a plan for the wait.
That is the part I can work on. Not the timing. The wait. How many years of income can you fund without selling stock at a loss? What happens to the pension election if the market is down forty percent the year you retire? Which of your assumptions were made by a 65-year-old who has never lived through a decade where the index went nowhere? Those are answerable questions, and answering them is the whole difference between a household that rode out Japan and one that ran out of Japan.
How many years of spending could you fund without selling a single share at a loss?
What return did your retirement plan assume, and was that number set during the last fifteen years?
If the market were forty percent lower the year you retire, would you still make the same pension election?
What share of your portfolio is in the ten largest companies in the index, counting every fund that holds them?
Who in your life has lived through a decade where the index went nowhere, and have you asked them what it was like?
A Portfolio Preparedness Review looks at valuation, concentration, withdrawal timing and pension elections together, and tells you how many years of a flat market your plan could absorb. No product, no obligation.
Start the reviewRobert J. Shiller, U.S. Stock Markets 1871 to present and CAPE ratio, Yale University— the historical CAPE series, record high of 44.2 in December 1999, and median.
S&P 500 Shiller CAPE Ratio, GuruFocus— 41.59 as of August 1, 2026; historical median 16.06.
Nikkei Stock Average, Nikkei Inc.— closing high of 38,915.87 on December 29, 1989, first exceeded on February 22, 2024.
Federal Reserve History, Federal Reserve Bank of St. Louis— the Plaza Accord of September 1985 and the 1929 crash; the Dow's September 3, 1929 close was first exceeded on November 23, 1954.
Bank of Japan— policy-rate history 1989 to 1990 and the bank-recapitalization programs of 1998 to 2003.
FRED, Federal Reserve Bank of St. Louis— S&P 500 closing highs of March 24, 2000 and October 9, 2007, both first exceeded on March 28, 2013; federal funds rate 2009 to 2021.
Federal Debt: Total Public Debt as Percent of GDP, FRED, Federal Reserve Bank of St. Louis— gross federal debt near 55% of GDP in 2000, 73.2% at Q4 2008, and roughly 123% as of Q1 2026.
Top 10 S&P 500 Stocks By Weight, Forbes, citing S&P Dow Jones Indices— the top 10 holdings near 27% of index weight at the 2000 peak, versus roughly 41% in 2026.
Bailey Financial Services, Inc. is a state-registered investment adviser. This page is educational commentary and reflects the opinions of Wilder Bailey as of the date shown. It is not a recommendation to buy or sell any security, not a forecast of when any market will rise or fall, and not personalized advice; nothing here projects a return or predicts a date. The characterization of Japan's aftermath as a refusal to recognize losses, and the reading of the United States as more likely to repeat 2000 than 1989, are judgments, not measurements, and reasonable people read the same history differently. Historical patterns are not predictive of future outcomes.
The recovery periods shown are measured from index closing high to the first close above that high, in nominal terms and excluding dividends; total-return recoveries were shorter in every case. Valuation figures are drawn from the sources listed above as of the dates given and will change. The Shiller CAPE is one measure of valuation among several and has remained elevated for extended periods without a decline following.
Bailey Financial Services is not affiliated with, endorsed by, or compensated by Yale University, Nikkei Inc., the Federal Reserve Bank of St. Louis, the Bank of Japan or GuruFocus. Citing a source does not imply its author agrees with any conclusion drawn here. Investing involves risk, including the possible loss of principal.
The three-way table compares conditions at three different points in time using measures that are each internally consistent but not perfectly like-for-like across eras — index composition, GDP accounting methods, and debt definitions have all changed since 2000. Figures are rounded and reflect the sources cited above as of the dates given. Labeling today's starting point as worse than 2000's or 2008's on these four measures is a comparison of conditions, not a forecast of what follows from them; higher readings on these measures have preceded both sharp declines and, at times, continued gains.