Oversold doesn’t mean cheap

Equity markets are especially vexing right now. In some ways, stocks might be easier to buy, but maybe not attractive to hold at current valuations. That is the tension in this week’s four charts. First, we can see that recent weakness has pushed a measure of S&P 500 technical conditions into oversold territory. At the same time, equities account for record levels — nearly half — of household and nonprofit financial assets, the earnings-yield comparison with Treasuries offers little obvious cushion, and investors have enjoyed an unusually rewarding decade.

On one hand, there is a plausible case for a rebound here. The sort of technicals we observe often point in that direction. But there is a less convincing case for assuming that the rebound, if it arrives, stems the longer-term concerns. For investors, the challenge is to keep those two judgments separate. The next several weeks may reward a willingness to buy weakness, particularly as investors try to deal with Federal Reserve decisions and guidance, which has some on edge. Looking out further, the next several years will depend much more on what we pay for earnings, how those earnings develop and how much risk portfolios want to bear.

The technical picture has improved.

MarketDesk’s S&P 500 technical composite ends at −1.4, within the chart’s oversold band. Over the two years shown, comparable readings have repeatedly been followed by recoveries in the indicator. History doesn’t always repeat itself, but this has been a reliable signal.

That is a reasonable starting point for a tactical argument. Selling pressure eventually exhausts itself. Investors who were reluctant to buy at higher prices reconsider, while even a modest improvement in the news can be enough to produce a substantial change in sentiment. But an oversold reading is an observation about market conditions, it’s not a weighing machine, as certain astute investors like to say.

It also deserves some circumspection. The chart includes deeper declines than the current one, and two years is a short period from which to infer a dependable trading rule. Moreover, the recovery of a technical indicator is not itself a record of the returns an investor would have earned after buying.

Therefore, we might say that the evidence supports the possibility of a bounce; it does not establish either its timing or its durability. For someone who has been waiting to restore an underweight equity allocation, that may be enough. But, for someone carrying more equity exposure than desired, it could be a concern.

Chart 1

The S&P 500 Has Entered Oversold Territory

S&P 500 Technical Composite, September 2024–September 2026

The latest reading falls within the indicator’s oversold band. Earlier episodes show that conditions can deteriorate further before recovering; the chart does not provide a backtest of subsequent equity returns. In the same report, MarketDesk’s summary table flags an oversold condition only at readings of −1.5 or below, so the S&P 500 sits just inside the shaded band rather than at an extreme. Chart reproduced from MarketDesk Research, Charts of the Week, 11 September 2026.

This matters (households have a great deal riding on the market.)

The Federal Reserve provides a firmer foundation for the second chart. Its measure of directly and indirectly held corporate equities reached 48.2% of household and nonprofit financial assets in the second quarter of 2026. That compares with 45.4% at the end of 2025. These are quarterly observations, rather than an estimate of today’s positioning.

A rising equity share does not necessarily mean households have been buying aggressively. The stocks they already own can appreciate faster than their other assets. The Federal Reserve explicitly distinguishes transactions from changes in the market value of existing holdings when explaining movements in household wealth.

Nor does this aggregate tell us that the typical family has half its financial assets in stocks. It describes an entire sector, including nonprofit organizations, certain institutions, etc., with very different balance sheets among its members.

What it does tell us is that a large portion of aggregate financial wealth is exposed to equity values. For individual investors, the practical question is whether appreciation has insidiously changed the portfolio they thought they had.

Consider a simple illustration. A portfolio begins with 60% in stocks and 40% in other assets. If the stocks double while the other assets remain unchanged, the equity weight rises to 75%. The investor has made no new purchase, but the portfolio now depends much more heavily on the stock market. Bonds have been weak, other investment assets have struggled. So, equities have become more meaningful by default.

Chart 2

Equities Represent Nearly Half of Household Financial Assets

Directly and indirectly held corporate equities as a percentage of household and nonprofit financial assets, Q4 1951–Q2 2026

The measure includes households and nonprofit organizations and captures direct and indirect equity holdings. Changes reflect asset-price movements as well as transactions. It is not a measure of the typical household’s allocation or of remaining buying power. The 48.2% reading for the second quarter of 2026 is the highest in the history of the series.

Comparing yields is a good rule of thumb, as long as you understand what they mean

The third chart compares the S&P 500’s earnings yield with the yield on a 10-year Treasury. The two now sit close together — an earnings yield of roughly 4.0% in June 2026 against a 10-year Treasury yield of about 4.5% — although the earnings yield has been the lower of the two since early 2024.

Earnings yield is earnings divided by price — the inverse of the price-to-earnings ratio. It gives investors a way to ask how much current corporate profit they are purchasing relative to the income available from a government bond. That isn’t the same as the equity risk premium, however. The return an equity investor expects includes future growth and the eventual value of the investment, not simply one year’s earnings divided by today’s price.

Damodaran’s work on implied equity risk premiums helps to define this better: the calculation estimates the return implied by market prices and expected future cash flows, then compares it with a risk-free rate. His data show that the S&P 500 earnings yield was below the Treasury bond yield in 1999, while his cash-flow-based implied equity premium remained positive. Sometimes, the two measures answer different questions.

When earnings yields are low relative to bond yields, the case for equities relies more heavily on future earnings growth and the value investors will assign to those earnings. That growth may arrive. But we are then paying for expected growth, not historical proceeds.

Chart 3

The Earnings-Yield Advantage Has Narrowed

S&P 500 earnings yield, on trailing twelve-month reported earnings, compared with the 10-year U.S. Treasury yield, 1960–June 2026

An earnings-yield spread is a relative-valuation measure. It does not incorporate all the growth and cash-flow assumptions required to estimate the forward-looking equity risk premium. Earnings yield here is trailing twelve-month reported earnings divided by the month-average index price, and the series ends in June 2026, the last month for which reported earnings are available; the 10-year yield had reached roughly 4.8% by September.

A good decade for stocks (and a bad one for bonds)

The final chart shows the S&P 500’s trailing 10-year annualized return at approximately 15% through August 2026. That is an extraordinary rate of compounding. At 15% annually, an investment grows to roughly four times its starting value over ten years, before taxes and costs. It is easy to understand why investors who have lived through such a period become reluctant to reduce their exposure. But that’s when complacency can become dangerous. A return that once would have seemed exceptional begins to feel ordinary. Financial plans become more ambitious. Diversification becomes harder to defend and out of favor assets are penalized further.

Strong trailing returns do not require an imminent reversal. Nor is the opposite true — although we observed pretty much the opposite results from the bond market in our previous article. In any case, a rolling return also changes as older observations leave the calculation; frothy returns can ease off. But, the more useful question is what would sustain another decade of comparable results from today’s starting point.

Earnings growth can carry markets a long way. Rising valuations can add to those returns, but investors cannot assume they will provide the same assistance indefinitely. Vanguard’s research makes a similar distinction: earnings growth and momentum can sustain markets in the near term, while valuation becomes more consequential over longer horizons.

Chart 4

Stocks Have Delivered an Exceptional Decade

Trailing 10-year annualized total return with dividends reinvested, S&P Composite, 1936–August 2026

Trailing returns describe the experience of an investor over the preceding decade. They do not establish the return available from today’s price or the timing of the next market reversal. Returns are nominal and assume monthly reinvestment of dividends; the index is the S&P Composite, which becomes the S&P 500 in 1957, so every ten-year window shown begins in 1926 or later. The August 2026 reading is 15.3%; on price alone, excluding dividends, it is 13.5%.

Compare and contrast

There is another reason to avoid becoming too confident. Three of these observations partly reflect the same market history. Rising stock prices lift trailing returns. They can increase equities’ share of household assets. And, when prices rise faster than earnings, they reduce the earnings yield.

The combination matters, but it is not a triptych shouting that a decline is approaching. It is partly three views of what a successful equity market has already done. The technical chart adds nuance. Recent weakness can create the conditions for a recovery even while the longer-term effects of that success remain in place. That is why the most defensible takeaway here is a responsible portfolio decision rather than a market forecast.

Don’t forget to rebalance

The most confident conclusion from these overlapping messages is the most modest. For investors whose equity exposure has grown beyond a considered target, a recovery would offer an opportunity to reduce that excess. It need not become a judgment that the entire market is uninvestable.

For investors below their target, the same weakness may provide a reasonable occasion to add. Starting allocations, spending needs and investment horizons matter a lot. Conversely, the destination of any proceeds also requires care. Reducing equities does not automatically imply buying long-duration bonds. Cash, shorter-maturity bonds and longer-duration securities serve different purposes and introduce different risks.

Our inclination is to treat the oversold reading as useful information about the near term, while allowing valuation and portfolio concentration to carry more weight in the long-term decision. A bounce would not persuade us, by itself, to raise a strategic equity allocation.

The market may continue to see strength. Earnings may continue to exceed expectations. But investors can participate in those outcomes without requiring their portfolios to depend increasingly on them.

Important disclosures