Have you ever noticed how the loudest warnings about market crashes seem to arrive right before nothing much happens? Capital spending looks excessive. Deficits keep widening. Oil prices jump. Each time the story ends the same way: someone claims investors are about to lose half their money. I have watched this cycle for years, and it still feels like a smoke detector that screams every time someone burns toast. What actually moves the market is quieter and far more consistent.
Earnings drive market outcomes. That simple idea sits at the center of more than a century of price history. When you dig into the numbers instead of the headlines, a clear pattern emerges. Mild dips in profits rarely hurt the index much. Large ones almost always precede the kind of declines that leave portfolios permanently smaller. Direction alone tells you almost nothing. Severity tells you nearly everything.
Why Popular Scare Stories Rarely Time the Market
Capital expenditures, government borrowing, and energy prices are real forces. None of them, by themselves, reprice equities with any reliability. A stock represents a claim on future cash flows. Its price is that claim divided by a discount rate. Those are the only two levers that can produce a serious decline. Everything else has to travel through one of those channels before it matters.
Consider the current wave of heavy technology spending. Companies can pour money into infrastructure for years without triggering a bear market, as long as revenue keeps validating the investment. The spending itself simply shifts cash into depreciation schedules. The market funds that trade happily while the growth story holds. The real risk appears the moment revenue stops confirming the bet. At that point the problem is no longer labeled capital spending. It becomes an earnings problem wearing a different costume.
Deficits work in a similar way. They can expand for a decade and still leave stock prices largely undisturbed until interest rates react. That is the discount-rate channel, not the earnings channel. Oil is more direct because energy costs squeeze margins. Even then, prices do not fall the day crude rises. They fall when the margin pressure shows up in guidance and reported results. I have found that focusing on the intermediate variables instead of the ultimate ones is one of the most common ways investors stay permanently early or late.
What 151 Years of Data Actually Show
Rather than accept someone else’s summary chart, it helps to rebuild the record from the longest available series of index prices, dividends, and trailing reported earnings. That exercise covers 151 complete calendar years. The measure is actual bottom-line profits, not adjusted operating numbers and certainly not optimistic forward estimates.
Two results stand out immediately. The market rises in roughly three-quarters of all years. That figure matches independent long-run compilations fairly closely. The higher percentages sometimes quoted only appear when the sample begins in the mid-1980s and quietly excludes the Depression, the 1970s, and both world wars. Sample choice changes the story more than most people admit.
The second finding is more useful. In years when earnings declined, the market still advanced about two-thirds of the time. In years when earnings rose, the advance rate climbed only modestly higher. Widen the window and the gap between the two branches shrinks further. A simple up-or-down tree therefore throws away most of the information an investor actually needs.
Severity Matters Far More Than Direction
Sorting the same years by the size of the earnings change reveals the real relationship. When reported profits fell by less than 10 percent, not a single year produced an index decline worse than 10 percent. The worst outcome in that entire group finished down roughly 9 percent. Mild earnings softness is largely noise for the broad market. That is why soft patches that dominate financial television so often fail to produce lasting damage.
The picture changes once the decline exceeds 25 percent. Roughly half of those years saw double-digit market losses. A quarter of them produced declines greater than 20 percent. The average return in that bucket turns negative. It is the only category across the full 151-year span where the average annual result sits below zero. Earnings drive market corrections through severity, not through the mere fact of a decline.
Whether the market finishes a given year higher or lower depends on sentiment, liquidity, and valuation. Whether it takes a genuine 20 percent beating is almost always an earnings question. The historical record answers that question with almost no exceptions.
The Eight Worst Years and What They Share
Only eight calendar years in the entire sample delivered total returns worse than minus 20 percent. That is a short enough list to examine one by one. Every single one of those years was accompanied by a double-digit drop in reported earnings, either in the same year or in the immediately surrounding period.
A few of them appear, at first glance, to break the pattern. In three cases earnings were still growing in the year the market fell hard. Those are not true exceptions. The large earnings collapse simply arrived the following year, or the market had already begun pricing the damage in advance. The sequence is consistent: the market anticipates the profit cycle rather than waiting for the final numbers to appear in print.
In each apparent exception the market did not ignore earnings. It simply got there first.
That timing difference is the practical heart of the story. Anyone who waits for the profit decline to show up in trailing data before reducing risk is reading a rear-view mirror and treating it like a windshield. Markets bottom before earnings bottom. The reverse is also true: they often peak before the peak in reported profits becomes obvious.
The Strongest Counter-Argument and Its Limits
One objection deserves careful attention. In 2022 the market suffered a decline of roughly 25 percent while many forward operating estimates continued to rise. On that basis the year looks like a pure multiple-compression event driven by rising rates. That description is half correct.
On trailing reported earnings the same year shows a clear double-digit decline. Both statements can be true at once. The gap between operating numbers and GAAP results is itself informative. When those two series diverge sharply, quality of earnings is usually under pressure. Rates are an independent channel. A genuine discount-rate shock can produce a serious decline on its own. Several of the historical 20 percent-plus drops carried heavy valuation compression alongside their earnings problems.
The honest formulation is therefore not that earnings are the only variable that matters. It is that earnings are the variable that separates a routine 10 percent air pocket from a portfolio-altering event. Rates determine how much valuation cushion exists when the earnings news finally arrives. Watch both. Weight the profit cycle more heavily.
When Big Earnings Drops Fail to Produce Big Market Losses
The mirror image of the bear-market pattern is equally important and costs investors more money in practice. Earnings collapsed by more than 25 percent in a dozen separate years. In half of those years the market still finished higher. Sometimes the gains were substantial.
Why? By the time the earnings collapse becomes measurable in trailing data, the market has already shifted to pricing the recovery. Selling into a confirmed earnings recession has frequently been the worst available trade. Markets bottom before earnings bottom without exception in the long-run record. That single fact should change how most people respond to a visible profit downturn.
Watch the Revisions, Not the Final Reports
If earnings drive the large moves and the market front-runs reported results, the practical question becomes which number carries usable information. The answer is not the final company report. It is the path of analyst revisions.
Analysts are not especially accurate at forecasting the level of growth. Long studies show that initial expectations consistently run well above the actual long-run rate, which itself tracks nominal economic growth fairly closely. The bias does not make the estimates useless. It makes the absolute level almost irrelevant and the direction highly relevant. Nobody should care that the consensus is too optimistic. The consensus is almost always too optimistic. What matters is the second derivative: the rate and breadth at which estimates are being cut.
Breadth often matters more than the headline magnitude. When a handful of very large companies carry the aggregate figure, the index-level estimate can still rise while the median company deteriorates. That divergence is one of the most common ways a weakening profit cycle stays invisible for an extra quarter or two. I have watched it repeatedly. The tell is rarely the big number. It is the quiet increase in the percentage of companies seeing downward revisions.
Practical Signals Worth Monitoring
None of this analysis helps unless it translates into a process. You cannot reliably predict the exact timing of the next large decline. You can decide in advance which signals will change positioning and by how much, so the decision is not made while red numbers are flashing and emotions are running high.
Credit markets often whisper what equity markets later shout. Bond investors get paid to worry about survival first. They tend to reprice deteriorating fundamentals ahead of equity holders, who spend more time pricing growth and less time reading the balance sheet. Changes in credit spreads have a long track record of leading both economic activity and equity prices. The level of spreads is less important than the rate of change. Spreads can sit at complacent readings for long stretches while equities keep compounding. Investors who de-risk the moment spreads look tight often give up substantial returns simply for the privilege of being early.
A few practical habits improve the odds:
- Track the breadth of earnings revisions more carefully than the aggregate growth number.
- Watch the direction and speed of credit-spread moves rather than absolute levels.
- Treat a mild earnings dip as noise unless it is accompanied by widening revision weakness and rising credit stress.
- Remember that confirmation across several signals matters more than any single reading.
- Respond to a deteriorating dashboard with a smaller position, not necessarily zero exposure.
These are monitoring tools, not mechanical triggers. The goal is preparation, not prediction.
Common Questions Investors Keep Asking
Do earnings declines always cause market corrections? No. That is the most frequent misunderstanding. Across the full historical sample the market still rose in roughly two-thirds of the years when reported earnings fell. Small declines are routine. The data show that large earnings declines are the near-necessary condition for large market declines.
How large does an earnings drop need to be before it starts to matter? Roughly 10 percent appears to be the practical threshold. Below that level the worst market outcomes stayed mild. Once the drop exceeds 25 percent, the odds of a double-digit market decline rise sharply and the average return turns negative.
Why did the market fall hard in a year when many people claimed earnings never declined? It depends on which series you examine. Forward operating estimates can rise while trailing reported earnings fall. The divergence between those two measures is often the real story. Valuation resets driven by rates can and do occur, yet the historical record still places the largest declines next to genuine profit problems.
Should you sell the moment earnings start falling? Usually the opposite is true once the decline is already visible in reported data. Markets bottom before earnings bottom. Several of the largest profit collapses in history occurred in years when the index still delivered double-digit gains. The useful signals appear earlier, in revision trends and credit markets.
Are capital spending and deficits irrelevant? Not irrelevant, but indirect. They affect equity prices only by working through expected cash flows or the discount rate. Watching them without reference to earnings and rates is watching the symptom rather than the disease.
What the Pattern Implies for the Years Ahead
Earnings drive the serious market corrections. That is the central finding from more than a century and a half of data. The next portfolio-altering decline is unlikely to arrive wearing a headline about capital spending or the size of the deficit. It will begin where the previous ones began: in the profit cycle. The early signs will surface in credit spreads and the breadth of estimate revisions long before they reach any earnings report most investors actually read.
The investors who get hurt are rarely the ones who missed the story entirely. They are the ones watching a different story, waiting for confirmation that always arrives after the damage is already priced. I have seen that pattern enough times to treat it as the default. The market is not required to make the process comfortable. It is only required to reflect, eventually, the cash flows that companies actually generate.
Preparation therefore means deciding in advance how much weight to give the profit cycle relative to other variables, how to interpret a sudden rise in revision weakness, and how large a position change is appropriate when several signals deteriorate together. None of that requires forecasting the exact month of the next large decline. It only requires respecting the variable that has separated routine setbacks from lasting damage for 151 years.
The noise will continue. New reasons to sell will keep appearing. Most of them will travel through the same two channels that have always mattered, and most of them will never complete the trip with enough force to produce a true bear market. The ones that do will almost always leave a clear trail in the earnings data, either in real time or with a short lag that the market itself has already anticipated. That is the record. It is not complicated. It is simply easier to ignore when the latest scare story feels more urgent.
Looking at the full span of history rather than the last few cycles changes the conversation. Mild earnings weakness is background noise. Severe weakness is the precondition for the declines that matter. The market prices expectations, so the useful information arrives in the path of revisions and in the credit market before it appears in the final reported numbers. Investors who treat those signals as early warnings rather than after-the-fact confirmation tend to navigate the large moves with less permanent damage. That approach will not eliminate every loss. It does improve the odds of remaining solvent and invested when the recovery eventually begins.
In the end the relationship is straightforward. Earnings drive both the durable advances and the serious setbacks. Everything else is secondary. Keep the primary variable in view and the secondary stories become easier to filter. That is the practical takeaway from a century and a half of market history, and it remains as relevant today as it was in any of the eight worst years on record.
