High valuations – not rising bond yields – threaten American stocks

Overview

In The Market is obsessed with the wrong things (25 March), Chris Conway wrote: “markets obsess over war and oil, but interest rates drive outcomes. Understand (central banks’) policy, and you understand the path of markets.” That, he claimed, is because “interest rates are the gravitational force of investing. They anchor valuations, shape behaviour, and determine outcomes across asset classes. Everything else, no matter how dramatic, is secondary.”

Similarly, in Why rising bond yields are a tax on every asset you own – and what it means for the ASX right now (19 May), Carl Capolingua contended: “the bond market sets the price of risk-free money. And (this) price … is the foundation on which every other asset is built … The bond market doesn’t comment on what other markets are doing; it determines what other markets can do.”

These statements reflect abstract and tidy – and impeccably orthodox – theory. They also ignore concrete and messy history and current reality. Accordingly, and I assume unwittingly, they paint a grossly distorted picture.

They’re hardly alone. “Bond markets are viewed as an alarm bell for equity markets,” claimed an article in The Weekend Australian (23-24 May), “and (the) alarm has been growing louder … When (bond yields) come under strain, the effect on equities is outsized – and if left unchecked, strain can rapidly turn into a crunch.”

Most finance academics agree: the relationship between the yield of a “risk-free” bond and stocks’ “earnings yield” (E/P, which is the inverse of their price-to-earnings (P/E) ratio) is – or, rationally, should be – positive. That is, if the bond’s yield rises then stocks’ E/P does likewise. And because yields and returns vary inversely, if the bond’s yield rises then stocks’ return should fall, and vice versa.

I don’t question the theory’s logic: I reject its correspondence to reality.

Conway and Capolingua, and the mainstream more generally, consider yields in isolation from the major factors which influence equities’ long-term returns. Yet yields per se are at best minor factors; as a result, it’s easy to overstate their impact upon stocks’ returns. Conway and Capolingua barely mention stocks’ valuations, and when they do it’s solely as a consequence of yields.

They thereby overlook the crucial importance of equities’ current valuations to their long-term returns. Ironically, they obsess about the wrong things!

Above all, abstract and theoretical claims about rates of interest, bonds’ yields and stocks’ returns require concrete empirical tests. This article provides them: I review the relationship between bonds’ yields, stocks’ valuations and stocks’ long-term returns, conduct my own analyses with almost 150 years of valid and reliable data, and draw two key conclusions:

1. it’s clear that, as predictors of stocks’ long-term returns, valuations are much more important than bonds’ yields;
2. the combination of stocks’ high valuations and bonds’ “real” yields threaten American and growth stocks; Australian and value stocks abate this risk.

Let’s First Clarify Key Terms

Conway and Capolingua mention rates of interest much more than yields. Conversely, I’ll analyse yields and mostly ignore rates. They’re related but distinct. A bond’s rate of interest, also known as its “coupon rate,” is fixed: it’s the percentage (expressed on an annualised basis) of the bond’s face value which its issuer agrees to pay its owner. Current yield, on the other hand, fluctuates: it’s the cash (as opposed to the total) return you’ll earn from the bond, expressed in annualised percentage terms, based upon its current market price.

Suppose that you purchase a bond whose face value is $100 and which promises to pay annual interest of $5. This bond’s rate of interest rate is thus fixed at $5 ÷ $100 = 5.0%. Its current yield, on the other hand, varies: if you sell the bond for $105, the purchaser still receives $5 of interest per year, but his initial yield is $5 ÷ $105 ≈ 4.8%. If the bond’s price subsequently falls to $95, its rate of interest remains unchanged but its current yield rises to $5 ÷ $95 ≈ 5.3%.

Current yield, unlike yield to maturity, ignores capital gains and losses. It thereby quantifies only a portion of the return a bond’s owner receives.

The U.S. Treasury 10-year bond’s current yield (hereafter simply “yield”) is the global benchmark: it underpins consumers’ borrowing costs, and establishes a baseline for corporate and other governments’ debt, worldwide. According to financial orthodoxy, it’s the global economy’s thermostat: a higher yield cools (tamps) and a lower yield heats (stimulates) corporate and household borrowing, assets’ valuations and economies’ growth.

That’s because 10-year Treasury’s yield is a close relative of the global “risk-free” rate of interest.

The “risk-free rate” is the minimum rate of interest that an investor expects (or will accept) from a security which purportedly carries zero risk. The orthodoxy concedes that no real-world investment meets this criterion. Accordingly, a “risk-free” investment doesn’t actually exist; as its proxy, the financial world uses 10-year bonds issued by the governments of leading economies – particularly the U.S.

In this respect the orthodoxy spouts nonsense. In particular, it ignores the crucial difference between risk-free return and return-free risk.

Return-free risk is a phrase coined by James Grant, publisher of Grant’s Interest Rate Observer, describing an “investment” which offers no meaningful prospect of profit yet exposes its owner to considerable risks. These include probable loss of (1) nominal capital through adverse mark-to-market movements, and/or (2) purchasing power through consumer price inflation.

The 10-year Treasury’s all-time low yield of 0.62% in July 2020 provides a startling example.

If you purchased it then and sold it May 2026, your net loss (taking into account the payments of interest you received) was 10.3%. On a CPI-adjusted basis, your total net loss ballooned to 33.5%. Under these conditions, each $1 of purchasing power which you invested in July 2020 collapsed to just $0.665 in May 2026. That’s a compound annual growth rate (CAGR) of -6.8% per year over ca. 5.8 years. So much for Treasuries’ “risk-free return!”

So don’t be fooled: theoretically, it’s reasonable to assume that U.S. Treasury securities’ rate of interest is “risk-free” in the sense that the probability of non-payment of interest is effectively zero; but in the real world it’s anything but sensible to suppose that their CPI-adjusted total returns are risk-free.

Two Preliminary Results

For long-term investors, a crucial question thus arises: to what extent has the 10-year Treasury’s current nominal (that is, unadjusted for CPI) yield, considered in isolation, influenced the S&P 500 Index’s CPI-adjusted total (that is, including dividends) 10-year prospective return? Do higher yields, like those occurring presently, crimp prospective long-term returns? Figure 1 answers this question.

Considered in isolation, since 1881 nominal current yields have told us little about the S&P 500’s prospective total return: its variation explains just 8% of the total variation of stocks’ prospective returns; other factors (including random variation) explain the other ca. 92%.

Figure 1: S&P 500’s 10-Year, CPI-Adjusted Total Prospective Return (CAGR) by 10-Year Nominal Yield, January 1881-May 2026

Variations, such as yields’ changes, etc., as well as a focus upon more recent decades, etc., produce much the same conclusion. And over intervals of less than ten years, Treasuries’ yields are even poorer predictors of equities’ returns.

For the orthodoxy, it gets worse: the best-fitting (polynomial) trend line in Figure 1 shows that, at nominal yields ranging from 2% to 7% (which prevailed during 82% of the months between January 1881 and May 2026), stocks’ long-term prospective returns hardly budge. Only when yields approach 10% – which is more than twice the current yield – do returns fall significantly.

The scatter of observations in Figure 1 demonstrates that each range of yields encompasses a wide range of returns. At yields between 3.5% and 4.5%, for example, returns have varied between -4.6% and 18.8% per year, and averaged 7.8%; given this mean and its standard deviation (4.8%), the probability of a negative 10-year CAGR has been 5.3%.

Presently (22 June), the 10-year Treasury’s nominal yield is 4.52%. At yields between 4.5% and 6.5%, stocks’ long-term prospective returns since 1881 have varied between 5.6% and 6.5% per year, and averaged 6.0%; given this mean and its standard deviation (3.6%), the probability of a negative 10-year CAGR has been just 0.2%.

On an historical basis extending over 150 years – and, it’s important to emphasise, considering the Treasury bond’s nominal yield in isolation – long-term equity investors have nothing to fear from yields which rise, say, 150 basis points from their current level.

What about the bond’s “real” (CPI-adjusted) yield? For each month from January 1881 to May 2016, I subtracted CPI’s 10-year CAGR from the Treasury’s nominal yield, and to these real yields I matched the S&P 500’s CPI-adjusted total return over the next 10 years (i.e., to May 2026). Figure 2 plots the results.

Figure 2: S&P 500’s 10-Year, CPI-Adjusted Total Prospective Return (CAGR) by “Real” 10-Year Treasury Bond Yield, January 1881-May 2026

The relationship is statistically significant: as the orthodoxy contends, an increase of “real” yields is associated with a decrease of stocks’ long-term prospective returns.

Substantively, however, it’s hard to regard this relationship as important: the current (May 2026) real yield of 0.9% has been associated with prospective long-term returns varying from -2% to more than 12% per year. If the real yield suddenly rose to its average since 1881 (2.2%), stocks’ average prospective long-term return would barely fall (i.e., from ca. 8% to 7.2% per year).

If nominal and “risk free” yields are minor influences upon stocks’ prospective long-term returns, what’s a major (or, at least, a much more important) influence? Stocks’ valuations are.

Figure 3: S&P 500’s Prospective 10-Year, CPI-Adjusted Total Return (CAGR) by CAPE Ratio, January 1881-May 2026

During a given month from January 1881 to May 2016, the higher has been the S&P 500’s Cyclically-Adjusted PE (CAPE) ratio (don’t worry if you don’t know what it is; I’ll describe, explain and justify it below) the lower, on average, has been the Index’s total return over the next ten years (Figure 3).

CAPE’s variation explains ca. 25% of the total variation of stocks’ long-term returns; other factors (including random variation) explain the other ca. 75%. On that basis, today’s valuations are three times more important than “risk-free” long-term bond yields as influences of stocks’ future long-term returns.

Today’s bulls might reply: “most CAPEs, particularly those between 10 and 25, which comprise almost 85% of the months under consideration, encompass a wide dispersion of long-term returns.” That’s true, but it’s incomplete. Specifically, it omits three salient facts: firstly, the higher is CAPE, the lower is the returns’ dispersion; secondly, at CAPEs above 30 long-term returns are mostly negative.

Thirdly and most relevant to the present, for the first time since the Dot Com Bubble (when it attained its all-time high of 44.2 in January 2000) CAPE currently (May 2026) exceeds 40. This bodes very poorly for American stocks’ return over the decade to come (for details, see Australian versus American equities: past, present and future, 24 November 2025).

(As a brief aside, what underpins CAPE? I don’t know – this article’s conclusion nominates a candidate – but I do know that it’s not the 10-year Treasury’s “risk-free” yield: their scatterplot’s R² is just 0.01.)

It’s important to stress: in this section I’ve summarised preliminary results. I’ve considered in isolation the influence of “risk-free” bonds’ yields and stocks’ valuations upon stocks’ long-term returns. Clearly, what’s needed is an analysis which incorporates CPI, yields and valuations, and considers their joint effects.

To conduct such an analysis, it’s necessary first to review models and measures of the joint influence of (a) bonds’ yields, (b) consumer price inflation and (c) stocks’ current valuations upon (d) stocks’ long-term returns.

Robert Shiller’s CAPE and Excess CAPE Yield

In 2020, Robert Shiller, professor of economics at Yale University and winner in 2013 of the Bank of Sweden Prize in Economic Sciences in Memory of Alfred Nobel (which is invariably but erroneously called “the Nobel Prize in Economics”), devised the Excess CAPE Yield (ECY). As a measure of stocks’ valuation, the Cyclically-Adjusted P/E (“CAPE”) ratio, which Shiller and his colleague, John Campbell, devised in 1988, incorporates stocks’ earnings and CPI, but not “risk-free” bonds’ yields.

Unlike the Dividend Discount Model and the so-called “Fed model,” realistic assumptions underpin ECY. It’s therefore much less sensitive to its assumptions, and provides a more comprehensive measure of stocks’ valuation: it shows how “risk-free” bonds’ yields influence equities’ attractiveness or otherwise.

ECY = the inverse of the CAPE ratio minus the 10-year Treasury bond’s “real” yield.

The “real” yield is the nominal yield of the 10-year Treasury minus the rate of growth (expressed as a CAGR) of CPI over the preceding ten years. This “real” yield represents the “safe” or “risk-free” return that an investor could earn from these bonds by holding them to term and after taking into consideration the change of purchasing power over the preceding decade.

In effect, ECY measures the equity risk premium (ERP), i.e., the extra return which investors demand from the risk of holding stocks instead of “risk-free” securities like 10-year Treasury bonds. ERP measures the financial compensation required to bear the stock market’s additional volatility and risk of losses.

The higher ECY rises above 0%, the more attractive stocks become relative to bonds, and vice versa. Holding constant ERP, the lower is CAPE (and thus the higher is its inverse), the higher is ECY; holding constant CAPE’s inverse, the lower is ERP the higher is ECY.

Shiller devised ECY in order to tell us what, for example, the Fed model (given its denial, in effect, that an Equity Risk Premium exists) can’t: “at any given point in time, how much extra return can investors reasonably expect from ‘risky’ stocks compared to ‘safe’ bonds?”

The higher ECY rises, the stronger is the premium which stocks offer over bonds; under these conditions the stock market is relatively attractive. The lower ECY falls, on the other hand, the smaller is the premium stocks pay over bonds; under these conditions, stocks look less attractive (or unattractive and overvalued) relative to fixed-income alternatives.

It’s crucial to appreciate: ECY’s foundation (CAPE) isn’t just the empirically best yardstick of stocks’ valuation; logically and conveniently for value investors, it borrows heavily from Graham and Dodd.

They observed that the standard (one-year) P/E ratio often makes stocks and markets look artificially cheap during market peaks and unduly expensive during recessions. They therefore suggested that averaging the ratio’s “E” over 7-10 years would smooth temporary and cyclical distortions.

Shiller incorporated this insight into CAPE. It averages corporate earnings over ten years, and thus tamps the distortions of business and market cycles, etc. He also adjusted earnings for CPI to ensure an accurate, long-term comparison of historical value.

The higher CAPE rises at a given point in time, the lower have been stocks’ CPI-adjusted returns over the next 5-10 years. Equally, the lower CAPE falls at a given point, the higher stocks’ CPI-adjusted returns will tend to rise subsequently.

CAPE isn’t merely logically and empirically rigorous; it also reflects the assumptions of the founders of value investing – and thus of long-term investors. As such, it ISN’T a market-timing tool; perhaps for that reason, speculators have generally ignored and dismissed it.

CAPE has been subjected to a considerable amount of uninformed criticism – seemingly from people who simply don’t like its implications. In “The Many Colours of CAPE” (Yale ICF Working Paper No. 2018-22), Shiller and Farouk Jivraj of Imperial College London investigated “the efficacy and validity of CAPE from several different perspectives.” Their analysis found that “CAPE consistently displays economic and statistical significance far better than any of its peers.” They also explored “alternative constructions of CAPE ... (and found) that original … is still best when comprehensively and fairly reviewing the other proxies ...”

Ignore its uninformed critics: CAPE is, logically and empirically, superior to any alternative method of valuation – including those advocated by informed critics (uninformed ones offer no credible alternative).

Other studies (I won’t bother to cite them; interested and energetic investors can easily locate them) have corroborated this conclusion. I’m unaware of any dispassionate and rigorous analysis – as opposed to idle opinion from people unfamiliar with this literature and with an axe to grind – which challenges (never mind overturns) it.

CAPE and ECY are most compatible with Graham’s and Dodd’s assumptions and approach; in contrast, the key assumption of the Fed model (namely that the equity risk premium is zero), is clearly untenable.

Results

High CAPEs at a one point in time generally foreshadow low(er) stock returns over the following decade (recall Figure 3). Stocks’ long-term returns, in other words, reliably regress to their very long-term means.

The 10-year Treasury’s nominal (that is, unadjusted for CPI) yield, on the other hand, doesn’t. Instead, it follows cycles which often last decades (Figure 4).

From the 1880s until 1920, these yields rose gently; and from 1920 to 1941, they fell mildly. For the next 40 years, at first moderately and eventually violently, they increased; and in September 1982, they scaled their all-time high (15.3%). Then for almost 40 years they fell almost continuously and cumulatively drastically: as already mentioned, in July 2020 they plumbed their all-time low of 0.62%.

Figure 4: Nominal 10-Treasury Bond Yield, January 1881-May 2026

Finally, over the past six years these yields have mostly risen sharply: in January 2026, they briefly touched 4.63% – the first time since the eve of the GFC that they exceeded their mean since 1881 of 4.46%. The recent upward spurt is the sharpest since the late-1970s and early-1980s.

The crucial questions – for owners of stocks as well as bonds – thus are:

1. have the 10-year Treasury’s yields, and interest rates more generally, commenced a decades-long climb?
2. if so, how high will rates eventually rise?
3. what does the answer to question #2 imply for equities’ long-term returns?

If rates have commenced a long upward cycle, and in the context of equities’ presently very high (both by historical and recent standards) valuations, the implications are profound – and potentially lethal.

Figure 5 plots the S&P 500 Index’s CAPE ratio since 1881. Until ca. 1996 it was mean-regressing. So has it been since then – but at a significantly higher mean. Only at the depths of the GFC did it return to its pre-1996 average; as a result, its mean since 1996 (28.6) is, statistically and substantively, significantly higher than its mean to 1995 (14.8). Bearing in mind that it’s a valuation and not a market-timing tool, CAPE’s mean since 1996 is higher than any CAPE before 1996 – and is comparable to its peak on the eve of the Great Crash of 1929.

Figure 5: CAPE Ratio, S&P 500 Index, January 1881-May 2026

Indeed, CAPE has presently scaled heights (40.2 in May 2026) unseen since the Dot Com Bubble of the late-1990s and early 2000s.

Figure 6 plots ECY – and thus quantifies the value of equities relative to “risk-free” bonds since 1881. First note that ECY – and, by implication, the equity risk premium averages 4.6% and has virtually always exceeded zero. (Consequently, and as a brief aside, the Fed model rests upon a false premise.)

Figure 6: Excess CAPE Yield, S&P 500 Index, January 1881-May 2026

Also note that, unlike CAPE, ECY regresses to its mean; in other words, it’s trendless. In that respect two recent (that is, since ca. 2000) sets of results are particularly germane to the present. Firstly, from the apex of the Dot Com Boom in 2000 to the nadir of the GFC in 2009, CAPE collapsed from 44 to 12.5; hence its inverse rose from 1 ÷ 44 = 2.2% to 1 ÷ 12.5 = 8.0%. Also at this time, the Treasury bond’s CPI-adjusted yield fell from 3.8% to 0.2%.

Consequently, ECY – and hence stocks’ attractiveness relative to bonds – increased from 2.2% - 3.8% = -1.6% in January 2000 to 7.5% - 0.2% = 7.3% in May 2009. In 2000, stocks hadn’t been so dear relative to bonds since the 1920s; by 2009, they hadn’t been so cheap since the 1980s.

Secondly, since the GFC stocks’ attractiveness relative to 10-year Treasuries has evaporated. Since March 2009, CAPE has zoomed from 13.3 to ca. 40 (May 2026); hence its inverse has collapsed from 8% to 2.5%. Over the same interval, the 10-year Treasury’s CPI-adjusted yield has risen from 0.2% to 1.0%; accordingly, ECY is presently (May 2026) 2.5% -1.0% = 1.5%. That’s very low compared to the average since 1881 (4.6%); it’s also low relative to its mean since the GFC (3.4%); indeed, it’s the lowest since the height of the Dot Com bubble.

Given CAPE is now hovering near historic (even by post-1996 standards) highs, and inflation and interest rates at their present levels, today’s ECY indicates that stocks presently offer a thin (by historical standards) margin over Treasuries.

CAPE is telling us that since the GFC stocks have become increasingly expensive relative to stocks on an historical basis; ECY adds that they’re also increasingly unattractive relative to bonds.

How unattractive? One logically possible answer to this question, an instantaneous reversion of ECY to its mean, is empirically highly unlikely but nonetheless instructive. In order instantly to restore ECY to its historical mean of 4.6%, either the S&P must collapse to 3,250 (i.e., plunge ca. 55%) or the Treasury’s yield must collapse to 1.2% (i.e., by ca. 70% from the current 4.56%).

I’m NOT saying that stocks are poised to crash; I AM saying that low ECYs, like high CAPEs, presage low long-term returns.

Figure 7: S&P 500’s Prospective 10-Year, CPI-Adjusted Total Return (CAGR) by ECY, January 1881-May 2026

For more than 100 years, ECY has been a reliable predictor of equities’ long-term returns (Figure 7). Low ECYs precede low long-term returns. The trendline tells us that today’s ECY of 1.5% implies that the S&P 500’s CPI-adjusted total return over the next ten years will average 4.1% per year.

Like CAPE, ECY isn’t a market-timing mechanism; accordingly, it can’t predict sudden market crashes. However, and like high CAPE ratios, low ECYs indicate that stocks’ long-term future returns will be modest compared to their recent actual returns.

Notice that the R² in Figure 7 (0.33) is higher than the one in Figure 3 (0.25), and much higher than in Figure 1 (0.08). In isolation, bonds’ nominal yields explain just 8% of the variation of stocks’ prospective returns; on their own, valuations measured by CAPE explain 25%. ECY, which incorporates valuations and bond yields, explains 33%. One the one hand, that leaves two-thirds to other factors and random variation; one the other hand, by real-world standards over the past ca. 150 years, explaining one-third of equities’ prospective long-term returns is a gold-medal achievement.

How will ECY evolve over the next decade? Nobody can know; everybody, however, can and should know that over ten-year intervals since 1881 it’s reliably regressed to its mean – and that there’s no compelling reason to doubt that it will continue to do so.

For each month since January 1881, I’ve recorded the S&P 500’s present ECY, its ECY ten years hence, its CPI-adjusted total return (expressed as a CAGR) over the past ten years, and its return over the next ten years. I then ranked these data by current ECY, divided the dataset into five segments (quintiles) with equal (net of rounding) numbers of observations, and computed key statistics for each quintile.

Table 1 summarises the results. How will ECY evolve over the next decade? Its current level (1.5%) is below-median, i.e., falls within Quintile #2. If past is prologue, ECYs within this quintile tend subsequently to rise: on that basis, over the next decade we can expect that today’s ECY will increase to ca. 4.8%.

Table 1: CPI-Adjusted, Total 10-Year Return (CAGR), S&P 500 Index, by Quintile of ECY, January 1881-May 2026

How will this rise of ECY impact stocks’ returns? On average within Quintile #2, the S&P 500’s retrospective return averages 9.0% per year (the Index’s CPI-adjusted total return to May 2026 is 11.2%), and its prospective return averages 4.4% per year. That’s hardly a crash; it is, however, a sharp deceleration – and, particularly for bulls, potentially a big disappointment.

The analysis of ca. 150 years of valid and reliable data tells that, over the next decade, ECY will likely increase. If so, logic tells us that this rise will either be sharp or gradual. For owners of stocks, a sudden lurch of ECY will likely generate shattering losses; a gradual rise implies general disappointment.

In summary, the combination of two factors (the major one is CAPE’s decrease; the minor one is rising “real” bond yields) implies that the S&P 500’s returns over the next decade will be sharply lower than it’s been over the past decade.

What about Australia?

To what extent do the foregoing results apply to Australia? To answer this question, let’s first ask: to what extent does the yield of the 10-year Australian Commonwealth Bond influence the All Ordinaries Index’s CPI-adjusted total (that is, including dividends) 10-year prospective return? Do higher yields crimp prospective returns?

Figure 8: All Ordinaries Index’s 10-Year, CPI-Adjusted Total Prospective Return (CAGR) by Commonwealth 10-Year Bond’s Nominal Yield, January 1985-May 2026

Figure 8, which is the Australian counterpart of Figure 1, tells us: considered in isolation, since January 1985 the bond’s current nominal yield hasn’t strongly presaged the All Ordinaries Index’s total, CPI-adjusted return over the next decade – and to the extent that it does, higher nominal yields beget higher total returns. (January 1974 is the earliest month for which I’ve been able to locate reliable data for its dividends; January 1985 is thus the earliest month for which its term-year total return is available.). The yield’s variation explains ca. 15% (versus little more than 8% in the U.S.) of the total variation of stocks’ prospective returns; other factors (including random variation) therefore explain the other ca. 85%.

Considering the bond’s current nominal yield (ca. 5.0% during most of May) and the relationship’s trendline, long-term stock market investors in Australia have nothing to fear from yields which rise from their current level.

What about the bond’s “real” (CPI-adjusted) yield? Figure 9 is the Australian counterpart of Figure 2. Statistically and substantively the relationship isn’t just insignificant: it’s non-existent. An increase of “real” yields doesn’t affect stocks’ long-term prospective returns.

Figure 9: 10-Year, CPI-Adjusted Total Prospective Return (CAGR), All Ordinaries Index, by “Real” Australian 10-Year Bond Yield, January 1984-May 2026

If nominal and “risk free” yields are a relatively minor influence upon stocks’ prospective long-term returns, what’s a major (or, at least, a more important) influence? As in the U.S., so too in Australia: stocks’ valuations are. The higher has been the All Ords’ CAPE ratio, the lower, on average, has been its (CAGR) return over the next ten years (Figure 10, which is the Australian counterpart of Figure 3).

Figure 10: All Ords’ Prospective 10-Year, CPI-Adjusted Total Return (CAGR) by CAPE Ratio, January 1984-May 2026

CAPE’s variation explains more than one-half (55%) of the total variation of Australian stocks’ returns since 1984 (versus one-quarter of the S&P 500’s total return since 1881); other factors (including random and thus unpredictable variation) explain the other ca. 45%.

On that basis, today’s valuations are crucial – and “risk-free” bonds’ real yields are relatively unimportant – influences of Australian stocks’ long-term returns.

Figure 11 plots the All Ords’ CAPE ratio since 1947. As in the U.S., so in Australia: until the mid-1990s it regressed to its mean (12.7). It’s also done so since 1996, albeit to a significantly higher mean (18.2). Bearing in mind that it’s tool for valuation and not a market-timing, today’s CAPE (23.0) is higher than virtually any time before 1996.

But it’s much lower than its American counterpart. Comparison of Figure 5 and Figure 11 reveals that, from the 1940s to the GFC, CAPEs in the two countries were roughly comparable; since then, however, the S&P 500’s CAPE has risen far more than the All Ords’. Presently (May 2026), the former exceeds 40 whereas the latter’s is just 23.

Figure 11: CAPE Ratio, All Ordinaries Index, January 1947-May 2026

By this metric, on an historical basis Australian stocks aren’t cheap. Relative to the S&P 500, however, they are. Accordingly, the All Ords’ prospective long-term returns are higher than the S&P 500’s.

Figure 12 complements Figure 6. It plots the All Ordinaries’ ECY – and thus quantifies the value of equities relative to “risk-free” bonds since May 1969 (which is the earliest date for which the RBA records the 10-year bond’s yield). Like CAPE, ECY regresses to its mean; unlike CAPE, since 1969 ECY has regressed to the same mean; in other words, it’s been trendless.

Figure 12: Excess CAPE Yield, All Ordinaries Index, May 1969-May 2026

Table 2, the Australian counterpart of Table 1, summarises Figure 12’s implications. How will the All Ords’ ECY evolve over the next decade? Its current level (2.5%, versus the S&P 500’s 1.5%) is below-median, i.e., falls within Quintile #2. ECYs within this quintile tend subsequently to rise: on that basis, over the next decade we can expect that today’s ECY will increase to ca. 3.5% (versus 4.8% for the S&P 500).

Table 2: CPI-Adjusted, Total 10-Year Return (CAGR), All Ordinaries Index, by Quintile of ECY, January 1984-May 2026

How will this rise of ECY impact Australian stocks’ returns? On average within Quintile #2, the Ords’ retrospective return averages 7.6% per year (its CPI-adjusted, 10-year total return to May 2026 is 7.8%), and its prospective return averages 7.2% per year. That’s hardly any change at all.

In contrast to the U.S., the combination of two factors (CAPE is the major one; the minor one is “real” bond yields) implies that Australian stocks’ returns over the next decade will be much the same as they’ve been over the past decade.

What, ultimately, are Table 1 and Table 2 telling you? If past is prologue and considering bonds’ yields and the likely course of ECY, over the next decade the All Ordinaries is likely to outperform the S&P 500 (see also Australian versus American equities: past, present and future, 24 November 2025).

How, in short, to mitigate the relatively minor (compared to valuations) risk of rising “real” bond yields? One way is to avoid overpriced American stocks and to accumulate more moderately-priced Australian stocks.

What about “Growth” and Value Stocks?

A second way is to buy and hold value stocks, and to shun and sell so-called “growth” stocks. The latter are more sensitive than the former to rising bond yields because their valuations rely not upon today’s earnings but upon those which they expect to generate in the future. They typically invest heavily in order to expand; as a result, they usually generate little or no cash. 

Investors who buy “growth” stocks are paying a hefty price now for the hope of large profits in the years to come.

Additionally, “growth” companies often rely upon debt to finance their expansion because they don’t generate enough cash to cover their operations. Higher interest rates increase the cost of this borrowing, which can decelerate or halt their growth trajectory, and thus crimp future profits.

As rates of interest rise, yields on assets like government bonds usually increase. Under these conditions, investors to demand higher yields from all assets; in particular, it prompts them to lower the prices they’re willing to pay for growth stocks.

Value stocks, on the other hand, are usually established companies; as such, they typically generate consistent, reliable cash flows and pay dividends. Their valuations rely much more upon today’s earnings, and those in the near future, than upon those they expect to generate in the relatively distant future.

Because investors receive their returns sooner, value stocks are less exposed to the punitive effects of the future discounting of cash flows and dividends.

Figure 13: U.S. Growth and Value Stocks’ 10-Year, CPI-Adjusted Total Prospective Return (CAGR) by 10-Year Treasury’s Nominal Yield, July 1926-May 2026

Using data compiled by Robert French and his colleagues (for details, see Want to shrink your returns? Buy “growth” stocks! 23 March), Figure 13 plots growth and value stocks’ total prospective returns (10-year, CPI-adjusted CAGRs) by the Treasury bond’s nominal yield. Yields don’t significantly affect prospective returns; the relationships are trendless. Equally, at any given nominal yield value stocks’ trendline is significantly higher (that is, their average return is greater) than growth stocks’.

What about “real” yields? In Figure 14 the trendlines are significantly negative: larger CPI-adjusted bond yields tend to decrease value stocks’ as well as growth stocks’ long-term prospective returns. As in Figure 13, however, at any given real yield value stocks’ trendline is higher (that is, their average return is greater) than growth stocks’.

Figure 14: U.S. Growth and Value Stocks’ 10-Year, CPI-Adjusted Total Prospective Return (CAGR) by 10-Year Treasury’s “Real” Yield, July 1926-May 2026

Conclusions and Implications

Cold Calculation versus “Animal Spirits”

In The Market is obsessed with the wrong things (25 March), Chris Conway wrote: “the variable that matters most, the one that sits beneath every valuation, … and every portfolio outcome, is interest rates. That has always been true, and it remains true today.”

That’s true in the tidy world of theory. In the messy world of reality, however, it’s never been true – and it’s not true now.

In Why rising bond yields are a tax on every asset you own – and what it means for the ASX right now (19 May), Carl Capolingua added: “the selloff of bond markets around the world … (is) a global repricing of the price of risk-free money. And when the price of risk-free money rises, the price of everything else adjusts accordingly … The significance of this global nature of this repricing cannot be overstated.”

Quite the contrary: it’s easy to overstate it.

If most actors in markets usually behaved like Spock in Star Trek, then claims like Conway’s, Capolingua’s and Montgomery’s would apply to the real world. Most of the time, however, market participants behave much more like Ben Graham’s manic-depressive creation, Mr Market, than Mr Spock. Most speculators are more Homer Simpson than homo economicus. As a result, these claims – and those of the orthodoxy they reflect – are theoretically elegant but empirically deficient.

What best explains stocks’ prospective long-term returns? The results of my analysis are unequivocal: what matters much more than bonds’ current yields are stocks’ present valuations.

“At its core,” Conway adds, “investing is an exercise in discounting the future. Every asset, whether it is a government bond, a growth stock, or a piece of infrastructure, is valued based on the present value of future cash flows. And the rate used to discount those cash flows is derived, directly or indirectly, from the risk-free rate.”

Value investors ascertain margin of safety – which, in the real world, entails the cautious assessment of price versus value. To that end, the “risk-free” yield is at best of limited relevance; often, it’s simply irrelevant.

Whence derives this “risk-free” yield? From expectations regarding consumer price inflation and economic growth – and, most fundamentally, market participants’ appetite for risk (which John Maynard Keynes famously dubbed “animal spirits”).

The mainstream obscures market participants’ passion with the veil of mathematics. But veiled emotion is no less powerful than overt passion.

Focusing upon theory and discounting actual data, the mainstream implies, and sometimes explicitly asserts, that cold reason and bloodless calculation drive financial markets. Analysing evidence, my results imply that for most people most of the time, psychological and emotional factors such as confidence and intuition, and fear and greed are much more important than rational assessment and dispassionate calculation.

Keynes’ economics is mostly absurd, but parts of his analysis of investment and financial markets are astute.

In The General Theory of Employment, Interest and Money (1936), he coined the phrase “animal spirits.” It describes the psychological and emotional factors which spur many financial decisions. Animal spirits – the instinctive and spontaneous urge to act – rather than rational assessments spur decisions to hire staff, purchase inventory, build factories, etc. The innate inclination to do something rather than nothing, Keynes contended, plus a bias towards overconfidence, prompts people to act in financial markets. In his words, a “delicate balance of spontaneous optimism” rather than strict calculation of risk underlies their actions.

The alternating strength and weakness of animal spirits, said Keynes, creates cycles of boom and bust.

When spirits are high, giddy optimism reigns. Conversely, when pessimism and fear prevail, spirits shrink or even disappear. Keynes, in short, emphasised the primacy of volatile emotions over consistent logic: for most people and most of the time in financial markets, volatile herd mentality overrides steady independent thinking.

Keynes helped to lay the groundwork of modern behavioural economics.

Enter Robert Shiller: something akin to animal spirits explains why stocks’ long-term returns are driven primarily by their current valuations – and only secondarily (if at all) by bonds’ yields. CAPE doesn’t measure animal spirits; it does, however, reflect them. That’s why CAPE is a much better gauge than the “risk-free” yield of stocks’ prospective returns.

The orthodoxy emphasises abstract and tidy theory, and tacitly supposes that reason trumps emotion. The real world, however, is highly complex and thus very messy: above all, in this world emotions regularly run riot. 

To disciplined value investors, real-world investing is an exercise in assessing value and weighing risks – and profiting from – but never following – the crowd’s follies.

Overreaction Creates the Herd Mentality

Theory posits that bonds’ rising yields will crimp stocks’ returns. The research of Shiller and others, on the other hand, demonstrates that the bond-stock relationship has never been simple, stable or especially important. In general, and above all in the short term, speculators overreact; in particular, they overreact to “news” – such as short-term changes of bonds’ yields.

In multiple previous articles, I’ve demonstrated that speculators routinely overreact: they systematically overweight the present and extrapolate short-term “trends” (which is often mere random fluctuation) into the indefinite future; consequently, they’re not just fooled by the random nature of stocks’ short-term returns: they overlook the cyclical nature of their medium-term and long-term returns.

Speculators routinely overreact to the latest “news;” they also treat current market conditions as if they’ll last forever. They repeatedly underweight the past, and consistently fail to appreciate that stocks’ returns eventually regress to their historical means.

This herd mentality causes speculators to become excessively optimistic on the upside and unreasonably despondent on the downside. They overconfidently buy assets whose market prices have recently risen, and despondently sell those whose prices have recently fallen. In particular, and as a consequence of overconfidence, they exuberantly overpay for “growth” stocks during a boom; then, as a result of emotional overreaction, they panic-sell them at rock-bottom prices during a bust.

Above all, the crowd heeds the “mispredictions” of so-called “experts.” The herd and its leaders routinely overreact to mainstream news, consensus expectations and analyst forward earnings – and overlook or ignore the obvious fact that the consensus and its news are systematically biased.

More of the Same Eventually Means Opportunities for Some

The herd’s systematic overreactions explain why strict value investing works. Speculators occasionally create massive gaps between price and value; patient value investors defy the crowd, buy while speculators are panicking and sell when they’re stampeding.

Today’s growing concern about rising bond yields isn’t totally misconceived; much more importantly, however, it reflects speculators’ innate inclination to overreact. Most people in financial markets, amateur and professional, are now doing what they’ve usually done in the past: they’re obsessing about the wrong things. Above all, they’ve repeatedly overlooked, ignored or denied generally unattractive valuations. Rising yields are usually unimportant. Valuations, however, are always crucial.

As has occurred in the past, so it’s reasonable to expect that it’ll remain: in the short term, others’ preoccupations and overreactions will provide attractive long-term opportunities to disciplined value investors.

Click LIKE so that Livewire knows that you want more of this type of content. Click FOLLOW for notification when my next wire appears.