If you’ve been following the news – which, two of my recent posts concluded, is a mistake! – for the past six months or more, you’ll be forgiven for thinking that the earnings of ASX-listed firms are recovering and therefore that the market’s remarkable gains since March 2020 are “sustainable.” The truth is diametrically different. Alan Kohler wrote in The Australian on 7 April 2020 that “the stockmarket has come nowhere near to correctly pricing for the earnings recession now under way.” A year later, the situation is far worse: Australian equities have entered an earnings depression – and market participants have mispriced it even more egregiously.
In this article, I established five crucial facts that the bullish mainstream has resolutely ignored:
- in CPI-adjusted terms, the All Ordinaries Index’s earnings are lower today than they were almost 50 years ago;
- since the GFC its earnings have been falling;
- over the past year they’ve been collapsing;
- during the most recent earnings season – which one “expert” labeled “extremely positive” – their disintegration continued;
- because its earnings have cratered and its level has risen, the All Ordinaries Index’s PE ratio has skyrocketed to a previously-inconceivable height – almost 80!
These facts’ implications are stark: if past is prologue, then unless they act very conservatively investors’ prospective returns are poor and speculators who think they’re investors are (presumably unwittingly) tempting fate. Why aren’t analysts and journalists acknowledging these facts and admitting these consequences?
What “Analysts,” Journalists and Fundies Apparently Believe
The earnings of ASX-listed companies are allegedly increasing. “Behind the bullish – some might say heroic – assumptions shared by leading analysts,” said The Australian (“Stocks Set for Bumper Year,” 21 December 2020), “is the core belief that company earnings will jump in 2021.” One “veteran analyst” stated: “as we move into 2021, for shares to continue moving higher, earnings recovery now needs to be delivered. Promising signs are already emerging in respect of consensus earnings per share (EPS) revisions.” The article concluded: “as leading global brokers pencil in earnings growth of least 10% next year, suggests that, all things considered, even 20% earnings growth is not unrealistic.”
“The reporting season has been better than expected,” ABC News reported on 3 March 2021. It cited one Chief Investment Officer (CIO): “this reporting season has been an extremely positive one. There been more beats than misses and in fact it’s been the best one we’ve seen in a while in terms of an upgrade to earnings expectations.” According to the founder and CIO of a prominent funds manager profiled in The Australian (26 March), “strong earnings will deliver market growth over the next five years.” “Expectations for earnings and dividends have been surging since September,” The Australian (“Options for Investors as Bourse Nears Record High,” 9 April) added. “At the pace the market is now rising, the pressure is on for companies to keep providing rapid improvements in earnings and dividends.”
Fact #1: the All Ordinaries Index’s Earnings Have Collapsed 75% since January 2020
These assertions share two key attributes: first, they contain soft, wishful and comforting words but no hard, bracing and reliable numbers; as such, and second, they’re rubbish. Figure 1 plots recent earnings of the All Ordinaries Index. On the eve of the Global Viral Crisis (GVC) in January 2020, its earnings were $352.53. (The Index closed that month at 7,121.2; its PE ratio was 20.05; hence its earnings were 7,121.2 ÷ 20.2 = 352.53.) During 2020’s full-year earnings season (August), earnings collapsed to $145.
Figure 1: Earnings, All Ordinaries Index, Monthly Observations, January 2020-March 2021
Since then the plunge has continued. By January of this year – as analysts, journalists, politicians and others celebrated the economy’s alleged recovery – earnings sagged to $124 – that’s 65% less than they were just one year previously! And in the wake of the latest earnings season in February, earnings sank to $91. That’s more than one-third less than in mid-2020 and – astoundingly – almost 75% lower than in January 2020! Why haven’t you been told? I can think of several reasons – and I’ll bet you can, too!
Fact #2: the Index’s Earnings Have Never Collapsed This Much
The All Ordinaries Index has existed only since December 1979. Figure 2 extrapolates monthly estimates of its earnings to January 1974 – and plots their 12-month percentage changes. The average change is 5.4% and these observations are reasonably normally-distributed. Given their standard deviation (17.6%), ca. 95% of the observations lie ± two standard deviations from the mean, i.e., within the range +29.8% to -40.6%. The most recent changes, in contrast, deviate almost four standard deviations from the mean (see also Table 1 below). Clearly, that’s unprecedented: during the GFC, earnings decreased just 30% – little more than during the recession of the early-1990s.
Figure 2: Earnings, All Ordinaries Index, 12-Month Percentage Changes (Jan 1974-Mar 2021)
The most recent earnings season, the mainstream’s “consensus” affirmed, was very good, the best in a decade, etc. In particular, so-called “analysts” repeatedly babbled that earnings “beat expectations” – without mentioning that their expectations are designed to be beaten! A large body of academic literature has assessed this phenomenon rigorously: earnings virtually always exceed expectations – that’s how the game’s played! Experts seldom reference actual, hard data like those in Figures 1 and 2: if they did, it’d be obvious that earnings are (and the most recent earnings season was) dismal.
The two most recent earnings seasons were abysmal. Indeed, collectively they were the worst on record. Why haven’t “experts” told you?
Fact #3: the Index’s Earnings Were Lower in January 2020 than on the Eve of the GFC
It’s easy to anticipate the mainstream’s likely response to Figures 1 and 2. “COVID-19 is obviously to blame,” they’ll say. “But never mind: thanks to governments’ fiscal and monetary stimulus, the economy is recovering – indeed, GDP has already surpassed its pre-pandemic peak. As a result, and like the economy, earnings will also recover.” It’s easy to cast considerable doubt upon this expectation. In a previous post to Livewire, I showed why “stimulus” is actually poison – and Figure 3 corroborates this claim.
Figure 3: Earnings, All Ordinaries Index, Monthly Observations (August 2007-March 2021)
It plots the All Ords’ earnings in nominal and CPI-adjusted terms since their pre-GFC peak in August 2007. During that month, its CPI-adjusted earnings were $611. Then came the GFC – and huge amounts of so-called stimulus. Perhaps it lifted GDP, but it certainly didn’t boost the All Ords’ earnings: in March 2021, as we’ve already seen, they were $91 – that’s 85% less than they were more than 14 years ago!
Why haven’t “experts” told you? On the eve of the GVC, earnings hadn’t come remotely close to recovering from the GFC. So how is it plausible to assert and expect – as the mainstream clearly does – that earnings will quickly recover from the GVC?
Fact #4: CPI-Adjusted Earnings Have Slumped to a 50-Year Low
What else aren’t “experts” telling you? In January 1974, the All Ords’ CPI-adjusted earnings were $308 (Figure 4). As I’ve emphasised, in March 2021 they were $91. That’s 70% lower than 47 years previously! During the last quarter of the 20th century, earnings rose steadily in nominal terms (from $33 in January 1974 to ca. $150 in December 1999) and were roughly stable (despite plunges during the recession of the early-1980s and early-1990s) in CPI-adjusted terms (i.e., average of ca. $250 in 2021 dollars). From the trough of the Dot Com Bust to the crest before the GFC, on the other hand (which, more or less, encompassed the mining boom), earnings zoomed: in both nominal and CPI-adjusted terms, they effectively doubled.
Figure 4: Earnings, All Ordinaries Index, Monthly Observations (January 1974-March 2021)
The GFC crushed earnings: from peak (October 2008) to trough (December 2012), they plunged 41% (nominally) and 46% (CPI-adjusted). For the next seven years, they fluctuated without trend. The GVC slammed earnings further: during the 12 months to August of last year they plummeted 60%; by September they plumbed the nominal level they first reached in 1999; and since then, they’ve sagged further. Adjusted for CPI, they haven’t merely matched the depths they plumbed during the recession of the early-1990s: they’ve revisited all-time lows. Who knew? Who hasn’t told you? Why not?
I’m therefore very sceptical that earnings will soon recover from the GVC’s (never mind the GFC’s) hammerings. A key risk, it seems to me, is that when market participants finally acknowledge the “earnings depression” in Australia – and that it won’t lift soon – the All Ordinaries and other indexes will plunge. Moreover, if something remotely resembling an average PE reappears, the Index will collapse well below its low of March 2020.
Fact #5: the Index’s PE Ratio Has Skyrocketed to Almost 80!
The All Ordinaries Index’s price-to-earnings (PE) ratio is the primary gauge of its valuation. In January 1974, the Index was 328.5; its earnings were $33; hence its PE ratio was 328.5 ÷ 33.0 = 10.0 (Figure 5). In that month, buyers of stocks paid $10 per $1 of the Index’s earnings. Since January 1974, the ratio’s mean has been 14.2. Extremely high ratios (by pre-2020 standards) occurred in August 1987 (21.1, shortly before the Crash of 1987), January 1994 (22.8) and November 1999 (23.9, not long before the Dot Com Bust).
Figure 5: PE Ratio, All Ordinaries Index, January 1974-March 2021
But that’s nothing compared to the past year and today! On the eve of the GVC in January 2020, the Index’s ratio was higher (20) than at any time since the Dot Com Bubble. In August, when most of the companies comprising the Index reported their first pandemic-blighted results, it skyrocketed to a new (by a huge margin) all-time high (40); by November, it exceeded 55; presently, it’s 77!
The All Ords’ PE has soared to an unmatched – and hitherto-unimaginable – high. This isn’t merely because its CPI-adjusted earnings have collapsed to an all-time low; it’s also because, as Alan Kohler foreshadowed last year, “the stockmarket has come nowhere near to correctly pricing for the earnings recession now under way.”
Recent PEs are beyond astounding, and it’s hard to express how improbable they are by historical standards. (Since August, I’ve often wondered whether there’s been some mistake. But no: Standard & Poor’s computes them every month. To confirm the market’s PE of 77 in March, click here, download the Factsheet and scroll to the “Fundamentals” on p. 3). Buyers of a hypothetical portfolio that mimics the All Ords are now paying $77 per $1 of the Index’s earnings – that’s almost eight times more than they paid in January 1974! The PE’s standard deviation is 3.8. March’s PE is therefore a mind-numbing 16.5 standard deviations above its long-term mean, i.e., (77 – 14.2) ÷ 3.8 = 16.5.
Table 1: Standard Deviations and Associated Probabilities
What does that mean in plain English? Table 1 provides some context. If, before the outbreak of COVID-19, you sought to estimate the probability that the All Ords’ PE ratio would lie 1-2 standard deviations above its mean, you would have concluded that such a thing was probable (1 s.d.) or possible (2 s.d.). A PE of 3 standard deviations above its mean was very unlikely but still possible, and one 4 s.d. was very unlikely but hardly impossible. A ratio more than 5 s.d. above its mean is getting close to most people’s conception of “practically impossible” – and more standard deviations require a calculator with many more decimal points than mine!
Historically speaking, we’re sailing – many “experts” are seemingly doing so blissfully unaware of the reefs and darkening clouds – not just in utterly unchartered but frankly unfathomable waters. Yet to my knowledge no “analyst” or journalist has uttered a peep. Are they astonishingly incurious or wilfully blind?
Anticipating Three Criticisms
#1: Trailing versus Projected Earnings and PEs
Bulls will likely reject Figures 1-5 and their implications on the basis that they plot “trailing” earnings (or PEs based upon them). “The Index’s projected PE ratio was 18.5 in March,” they’ll say. “That’s little above its long-term average. So the market’s OK.” Trailing earnings – and hence the trailing PE’s denominator and the ratio overall – are (1) actual (i.e., reported to the ASX) earnings. As a result, they’re (2) consistently and (3) strictly defined. In short, they conform to Generally Accepted Accounting Principles. Colloquially, GAAP earnings include “all the bad stuff.” Conversely, “analysts” usually use “forward” (also known as “projected”) earnings. Crucially, these aren’t – because they can’t be – actual earnings as reported to the ASX: instead, they are estimates (frankly, “absurdly optimistic guesses” is a more apt phrase) of earnings during the next twelve months. Moreover, “analysts” don’t attempt to guess GAAP earnings: their estimates routinely incorporate “good stuff” (such as one-off gains) and exclude “bad stuff” (such as allegedly temporary losses, write-downs and write-offs of assets, etc.). As a result,
“Forward estimates” are subjective, inconsistent (over time) and relatively lax (at all times) estimates. Consequently, they’re rubbish. There’s a large and rigorous academic literature that corroborates this point: the forward earnings estimates of “analysts” are almost always inaccurate, sometimes wildly so; but they almost invariably err on the side of absurd optimism. Hence prospective earnings usually exceed – sometimes greatly – trailing earnings. Accordingly, forward/non-GAAP/subjective PEs are usually much smaller than their trailing/GAAP/objective counterparts.
The Index’s current prospective PE is highly revealing. What does it imply? Assume that the All Ords remains at the level it reached at the end of March (that is, 7,017). If P ÷ E = 18.5 and P = 7,017, then E = $379 – which, interestingly, approximates the All Ords’ average earnings ($393) during 2019. Tacitly or explicitly, bulls are assuming that earnings will soon revisit their pre-pandemic level – but are ignoring the key fact that, before COVID-19, they’d long been falling. Only if we assume that earnings will quickly zoom more than 3-fold from their current level will a “trailing” PE ratio of 18.5 reappear. This has never occurred before (recall Figure 2); on the other hand, in 2019 I would have scoffed that a market PE of almost 80 is impossible (Table 1)!
The colossal disparity between the latest trailing (77.1) and prospective (18.5) PEs substantiates what I’ve long believed: because they contain highly-inflated Es, prospective PEs rationalise paying too much for stocks – and thus, as we’ll see below, encourage speculators who think they’re investors to receive poor returns. “Trailing” (GAAP) earnings, in short, are much more conservative – and hence valid and reliable – than their “forward” counterpart. Why don’t “analysts” and journalists admit this?
#2: Can You Really Afford to Ignore or Deny History?
Many of today’s bulls, if they respond to Figures 1-5, will simply shrug and say “so what? Things are different this time.” In an interview in The Chicago Tribune in 1916, Henry Ford famously stated that “history is more or less bunk.” History, he reckoned, was past tense. We – or at least he – in contrast, lived in the present and could create his own future. The past, Ford reckoned, thus has no bearing upon the future. He concluded that it teaches us no lessons – and therefore need neither be studied nor considered. Today and also at most points in the past, participants in financial markets emphatically agree. “There can be few fields of human endeavour,” wrote John Kenneth Galbraith in A Short History of Financial Euphoria, “in which history counts for so little as in the world of finance. Past experience, to the extent that it is part of memory at all, is dismissed as the primitive refuge of those who do not have the insight to appreciate the incredible wonders of the present.”
Benjamin Graham begged to differ. “The combination of a record high level for bonds,” he wrote in the 1940 edition, of Security Analysis, “with a history of two catastrophic price collapses in the preceding 20 years and a major war in progress, is not one to justify airy confidence in the future.” Today, bonds are again sky-high (by historical standards) and on three occasions since 2000 the prices of stocks have collapsed. The study of financial history helps us to develop a better understanding of the financial world – as well as ourselves and others. What, in short, are implausibly high, sensible and attractively low prices?
History, it seems to me, is a better guide than today’s “experts” and their models. It forcefully reminds us that change of some kind always occurs, often unexpectedly, and the study of history sometimes suggests why some changes happen. Knowledge of and respect for financial history thus makes you a better decision-maker – and a more successful investor. Overweighting the present, in other words, discounts or even ignores the past – and thereby increases risk of error and failure. In this key sense the Spanish philosopher, George Santayana, was much wiser than Henry Ford: if we don’t learn from our past mistakes, he cautioned, we’re bound to repeat them.
#3: The All Ords’ Annual Rebalance Hasn’t Depressed Its Earnings and Inflated Its PE
It’s easy to cast doubt upon a final criticism – which, for want of a better name, might be dubbed the alleged “Afterpay effect.” In response to the analysis in Figures 1-5 bulls might say: “trailing earnings have fallen largely as a result of the rebalancing, which occurs annually in March, of the All Ordinaries Index.” Each March, some firms leave the Index and others replace them. But these firms differ in one key respect. Those which leave the Index (because a decrease of their market capitalisation has pushed them out of the top 500) tend to be (1) “old economy” firms that (2) earn profits. Firms which enter the Index, on the other hand, tend to be (1) “new economy” (i.e., technology) companies that (2) earn few or no profits (or, indeed, generate losses).
Lacking the time to compile and analyse the required (i.e., company-level) data, I haven’t tested this conjecture. I have, however, tested its key index-level consequence. If this conjecture were true, then at each rebalancing over the years, the All Ords’ earnings should, on average, fall. And over the years, earnings in the “rebalance” month should be significantly lower than in the other (“non-rebalance”) months.
Figure 6 compares the percentage change of the All Ords’ earnings in the rebalance (February to March) month to the change in all other months since the GFC. It also compares rebalance and non-rebalance periods over the next six-months. Its results don’t support the conjecture; indeed, if anything they’re the opposite of what it leads us to expect. In the rebalance month, earnings rise on a one-month basis (albeit by a statistically-insignificant average of 0.5%) and fall in the non-rebalance months. It’s true that six months after the rebalance they tend to fall (by an insignificant -0.4%), but they do in the non-rebalance months, too.
Figure 6: Earnings Growth, All Ordinaries Index, “Rebalance” vs. “Non-Rebalance” Months since the GFC
Implications and Conclusions
Why is historical analysis (that is, using all valid and reliable data) of the All Ordinaries Index’s PE ratio – and the earnings that underpin it – so important? Why do bulls ignore or reject such analysis? First, the ratio is mean-regressing; second, at a given point in time it strongly influences the Index’s subsequent rate of return. Table 2 updates and elaborates figures I presented in a previous post (for background and explanation, see Speculators are playing with fire; investors, don’t get burnt!).
Table 2: All Ordinaries’ PE Ratio and Subsequent Annualised Returns, Jan 1974-Mar 2021
The ratio is mean-regressing in two senses. The lower its level at any given point in time (column headed “Range”), the more it (1) tended to decrease during the previous 12 months and (2) will tend to rise during the next 12 months. The implication is clear: if the past since 1974 is prologue, then today’s astonishingly high ratio (77.1) will subsequently fall.
The second conclusion should give pause to investors worthy of the name. The higher is the Index’s current PE, the lower will tend to be its subsequent average short-term (12 months hence) and medium-term (five years hence) return. And PEs above 20 have tended, on average since 1974, to produce comparatively poor results. Today’s PE thus bodes poorly for tomorrow’s returns. Why don’t “analysts,” “experts” and journalists frankly acknowledge this?
Table 2 doesn’t imply – and no analysis of financial history can reliably predict – that a bear market, crash, etc., is imminent. From my analysis you shouldn’t panic, sell your shares and head for the hills: even if an index comprising 500 corporations is grossly overvalued, it’s likely that some of those corporations’ share prices will still be sensible. Regardless of whether you accept or reject my analysis, you should ignore the mainstream, conduct your own analyses and draw your own conclusions – or find an investment firm (I can think of one!) which has a successful long-term track record of doing so.
Ignoring financial history, as the mainstream does, is foolhardy; but studying it doesn’t – to put it mildly! – provide a clear and unerring view of the future. Arm yourself with valid logic and reliable historical data, but above all remain modest. “The only thing you can be confident of while forecasting future stock returns,” Jason Zweig sagely wrote in his commentary on Chap. 3 of Benjamin Graham’s The Intelligent Investor, is that
you will probably turn out to be wrong. The only indisputable truth that the past teaches us is that the future will always surprise us—always! And the corollary to that law of financial history is that the markets will most brutally surprise the very people who are most certain that their views about the future are right. Staying humble about your forecasting powers, as Graham did, will keep you from risking too much on a view of the future that may well turn out to be wrong.
In my next article, I’ll elaborate a major corollary of today’s earnings depression: without a prompt and substantial increase of earnings (which, in an economy gorging on immense quantities of so-called “stimulus,” is implausible), the All Ords’ current level of dividends rests upon shaky ground. Income investors, take note!