The house edge, borrowed from a casino
Data on the S&P 500 going back to 1999 shows the index finished higher on roughly 54% of trading days and lower on the rest; Crestmont Research, tracking more than 50 years of daily results, arrives at a similar 53.7%. It’s not a coincidence that this echoes the house edge at a roulette table, created by the two green zero pockets that pay nobody. A casino doesn’t need to win every hand — it needs to win slightly more often than it loses, forever. The stock market’s edge works the same way: barely better than a coin flip on any single day, but persistent, and persistence compounds.
The trouble starts when that single number gets asked to do more work than it can. A reader responding to the original piece made the obvious objection: counting up days says nothing about how big those days were. A market could post more winning days than losing ones and still end a year in the red, if the losing days are simply larger. Separating what "54%" tells you from what it doesn’t is the first real lesson here.
| What the statistic shows | What it leaves out | Why the gap matters |
|---|---|---|
| The market rises more often than it falls, day to day | The average size of an up-day versus a down-day | A market can have a majority of green days and a weak year if red days are bigger |
| A small, persistent frequency edge, similar in spirit to a casino’s | Total return, which also depends on dividends reinvested, not just price moves | Price alone understates what long-term investors actually earn |
| Consistency across decades of data | Which specific years, or which investor’s holding period, will feel this "average" | Any single stretch of years can look nothing like the long-run figure |
| A reason to expect a long-run tailwind | A reason to expect comfort along the way | The edge and the discomfort coexist; one does not cancel the other |
None of that means the 54% figure is meaningless — it just means it’s a starting point for a longer argument, not the argument itself.
How a small edge becomes a large result
The reason a slight daily tilt matters at all is that it isn’t experienced once — it’s experienced, and reinvested, thousands of times over a working lifetime. That repetition is what separates an edge from a curiosity.
flowchart TD A[Small daily edge: ~54% up days] --> B[Gains and dividends reinvested] B --> C[Compounding over years and decades] C --> D[Occasional attempt to time an exit] D --> E[Edge partly or mostly lost if mistimed]
That last arrow is where the real damage happens. Research from Hartford Funds found that missing just the ten best trading days over a recent 30-year stretch would have cut returns roughly in half, and missing the 30 best days would have wiped out 84% of the gain. A separate Vanguard analysis of a 60/40 portfolio over nearly 30 years showed $100,000 growing to $865,000 for an investor who stayed fully invested, versus $659,000 for one who missed only the five best days, and $540,000 after missing ten. The unsettling detail is that the best days tend to arrive precisely when investors are most tempted to leave: about 76% of the market’s best days have occurred during a bear market or in the first two months of a new bull run — exactly the moment when confidence is lowest.
The Federer point
There’s a useful, if unlikely, parallel in tennis. In a 2024 commencement address at Dartmouth, Roger Federer noted that across a career in which he won nearly 80% of his singles matches, he won only 54% of the individual points he played. Losing nearly every other point, for 24 years, and still becoming one of the most dominant players in his sport’s history. His explanation wasn’t that he stopped losing points — it was that he learned not to dwell on the ones he lost. A double fault was just a point; you reset and played the next one.
That discipline, more than any statistic, is the actual investing skill on display. The market’s edge is real, but it only compounds for people who don’t quit after a bad week.
When volatility is normal, and when it isn’t
None of this means downturns are rare exceptions to be waited out casually. Over the full history of the S&P 500, declines of at least 20% — bear markets — have occurred a dozen times, among 73 qualifying drawdowns of various sizes; the typical decline has run about 8.2% and taken roughly a month to reach its low point, while the deepest on record, from 1929 to 1932, erased 86.2% of the index’s value and took nearly three years to bottom. What that history does not offer is a reliable timetable: recovery periods have varied enormously from one episode to the next, and no single past drawdown should be treated as a preview of the next one.
| Measure | Historical figure |
|---|---|
| Qualifying drawdowns recorded since 1928 | 73 |
| Drawdowns classified as bear markets (-20% or worse) | 12 |
| Median decline | -8.2% |
| Median time to trough | 34 days |
| Deepest decline on record | -86.2% (1929–1932) |
The lesson isn’t that a crash of that magnitude is likely soon — it’s that meaningful declines are a normal feature of long equity histories, not evidence that a portfolio is broken.
What actually justifies a change
The uncomfortable feeling of watching a balance dip is not, by itself, information about whether a plan is working. A change in the underlying facts of your life is different: a job loss, an unexpected large expense, or retirement arriving sooner than planned are reasons to revisit an asset mix. "The market is down and it’s hard to watch" is not. The edge behind long-term investing was never a promise of comfort — only of arithmetic, applied patiently, to people who don’t interrupt it at the wrong moment.
This article is educational and does not recommend any specific security, fund, or trade. Historical patterns, including recovery timelines, do not guarantee future results.
