I Have Seen This Market Before: Bernstein on Bubbles, CAPE, and Quitting While Ahead

I have seen investors brag about being one hundred percent in equities before, and the air smelled exactly like this at the end of 1999. Bonds were for suckers, return forecasts were written in the sky, and nobody wanted to look cautious. Sitting across from me is a man who wrote those days into a book, and he makes no secret of his deja vu. Prices may not repeat, he argues, but behavior does: the crowd recites the same lines, shrugs at the same risks, awaits the same ending.

Is the spirit of 1999 back?

The most concrete proof of that parallel is the opportunity once offered by inflation-protected bonds . When his book was written these securities were two years old and paid a 4 percent real yield — a risk-free asset delivering close to the historical average of equities, while expected stock returns had sunk toward 1 percent. Today the same securities pay around 3 percent in real terms, which the speaker still calls generous. Equity valuations, meanwhile, are as stretched as they were then; the risk-reward scale has tilted back toward the risk-free asset, not a crash prophecy but a cold accounting.

On valuation, the gauge on everyone's lips is the Shiller CAPE ratio, price divided by ten years of inflation-adjusted average earnings. According to YCharts the ratio stood at 40.58 in September 2026, down from 41.12 a month earlier and above last year's 38.58. GuruFocus put the September 1, 2026 reading at 40.77, against a long-run average of 32.57 and a median of just 16.06. The speaker's warning is blunt: the indicator is not econometrically stationary, and anyone who fled stocks on its 1990 reading missed a 35-year rally; a high CAPE demands caution, never market timing.

Why central-bank headlines get ignored

Asked about last week's rate move, the speaker displays a startling indifference, and grounds it in principle: whatever everybody knows is already in the price. According to Reuters, the Fed lifted its policy rate by a quarter point on September 16, 2026 to a 3.75-4.00 percent range, with new chair Warsh presenting the unanimous decision as inflation-fighting resolve. In money manager Ken Fisher's phrase, front-page news is priced news; expecting surprise from the world's most watched institution is therefore futile.

The artificial intelligence question gets an answer both witty and uneasy: if it walks like a duck and quacks like a duck, it is probably a duck — though not everything that quacks is one. The speaker prices an AI bubble at 20 to 30 percent probability and recalls that the only bubble-like episode that failed to end in tears was Britain's mid-19th-century railways. The odd Vanguard revision feeds the suspicion: the firm's ten-year return forecast flipped from 1 to 6 percent in six months on AI buildout alone. Vanguard's own July 2026 outlook has since lowered expected US equity returns to a 4.2-6.2 percent band, with value stocks the most attractive slice; the gap between an expectation and a forecast — the enormous confidence interval around the central scenario — must not be forgotten.

Do giant IPOs shake the index?

Should index investors fear the coming giants — SpaceX, which raised some 75 billion dollars on June 12, 2026 at 135 dollars for 555.6 million shares, plus Anthropic and OpenAI waiting in the wings? According to a MarketBeat analysis, no: despite a 1.77 trillion dollar valuation, the free float sits at 4-5 percent, so the index footprint stays modest. Even the Nasdaq-100 weight doubling from 1.28 to 2.82 percent at the September rebalance leaves a top-seven company with a muted passive share. The free float rule means indices track the tradable slice, not the headline valuation; privately held shares never reach the funds.

The gap between dynamic allocation and strategic allocation is the most human passage of the talk. On paper the rule is simple: hold something like 60 percent stocks and 40 percent bonds, sell what rises, buy what falls, tilt the band when valuations stretch. But buying stocks a third time into the 2007-2009 slide, cash running dry, feels nothing like pressing a key on a spreadsheet. The pilot analogy sticks: crashing in a simulator and hitting turbulence in a real plane are different experiences; discipline is cheap in theory and dear in practice.

Winning the game and the real risk

The famous maxim — if you have won the game, stop playing — is a metaphor with concrete math: someone collecting 30 thousand dollars from Social Security and spending 70 thousand faces a 40 thousand gap, times 25, equals one million dollars to be secured in a TIPS ladder . The build site is tipsladder.com, where Kevin Esler's tool buys one rung per year on the secondary market and fills the 2036-2039 gap years with duration-matched pairs. Summitward's detailed guide shows how the ladder converts retirement spending into scheduled inflation-linked cash flows. Two traps matter: a 1,000-dollar par bond selling at 90 cents combines with a 1.6 inflation factor to cost about 1,500 dollars out of pocket; and the yield is taxed federally though exempt at state level.

On tax and liquidity the speaker is merciless: money you might need tomorrow does not belong in municipal bonds, which traded at 10-20 percent discounts in 2008. Emergency funds belong in short Treasury bills ; Fidelity's auto-roll convenience and Vanguard's 6-basis-point short-term fund solve the job nearly free. Factor investing — the long-run edge of value and small-cap stocks — is framed as a 55-to-45 bet; anyone unwilling to endure 15 years of trailing the market should simply buy it whole. Momentum's turnover devours its edge, commodity futures carry a roll return near minus 5 percent a year; those wanting commodity exposure should buy producer equities, and for 99 percent of savers a total-market fund suffices.

The generational fairness question is the most political moment: will Social Security survive for the young? The speaker opens by confessing that luck beat brains — his cohort entered cheap markets in 1980. The system will not vanish, he argues, but it will be means-tested ; low earners may still recover 90 percent of pre-retirement pay while doctors settle for 15-20 cents on the dollar. And the closing definition of risk ties everything: real risk is not losing 20 percent next month but living under a bridge at seventy. Mean-variance charts measure short-term wobble, not long-run shortfall; LTCM, sunk in 1998 with Nobel laureates aboard, is the proof. As Warren Buffett says, this game is won not by IQ 160 but by the most disciplined nervous system; the prescription for doctors is four courses — theory, history, psychology, and market microstructure.