Peter Lynch: a third gone in a week, and the myth about his investors
Peter Lynch ran Fidelity’s Magellan Fund for thirteen years and is usually remembered for one number: 29.2% a year. Two other numbers from the same record matter more to anyone holding a volatile asset. In one week of October 1987 the fund lost about a third of its value. And a story that half of its investors still lost money has been repeated for thirty years, although no one has ever produced the study behind it.
KEY TAKEAWAYS
- At 29.2% a year for thirteen years, $10,000 grows to about $279,500, roughly 28 times. Our arithmetic on the published average.
- A one-third fall needs about +50% to undo. Even at Lynch’s own pace, our calculator puts that at about 19 months.
- Lynch counted 53 stock-market declines of 10% or more between 1900 and 1995, and 15 of 25% or more. On Bitcoin’s daily closes since 2013 we count a 25% fall about once every 14 months; his figures work out to one every six years or so.
- “Half of Magellan’s investors lost money” is unverified. Fidelity denied it in 1996; the one firm that measured the fund found it lost money in 17% of twelve-month periods.
Who is Peter Lynch?
Peter Lynch was born in Massachusetts in 1944 and has spent almost his whole career at Fidelity Investments in Boston. He took a graduate degree from the Wharton School in 1968, joined Fidelity as an analyst, and in 1977 was given a small fund called Magellan. He ran it until May 1990, when he retired from day-to-day management at the age of 46. He stayed at Fidelity afterwards and wrote three books for ordinary investors, starting with One Up on Wall Street in 1989 and Beating the Street in 1993.
His public reputation rests on two ideas: that an ordinary person can understand a business well enough to own it, and that the people who do best in stocks are the ones who stay in through the falls. This profile is about the second idea, because it is the one with a price tag attached, and the record shows what that price looked like.
What did Magellan return under Lynch?
The figure usually quoted, including by Wharton’s own alumni magazine, is an average of 29.2% a year from 1977 to 1990, with the fund trailing the S&P 500 in only two of those years. Over the same stretch Magellan grew from a small fund into one of the largest in the United States; by October 1987 it held about $12 billion, according to The Christian Science Monitor.
Compounding makes that average hard to picture, so here it is in money. At 29.2% a year for thirteen years, $10,000 becomes about $279,500, almost 28 times the stake. That is our arithmetic on the published average, not an audited figure for any real investor, and it assumes someone bought on the first day and held to the last. Very few people did, and the next section shows why.
What happened to Magellan in October 1987?
In the four trading days ending on Monday 19 October 1987, the Dow fell more than 750 points. Lynch was on holiday in Ireland. Speaking to The Christian Science Monitor that December, he put the fund’s loss plainly:
We lost about a third and the market lost about a third.Peter Lynch, The Christian Science Monitor, 8 December 1987.
The paper reported that the fund’s assets fell by more than $4 billion from about $12 billion in less than a week, and that nearly all of the fall was in the value of its stocks; only about 2% came from investors pulling money out.
His answer to what he did about it was one sentence:
We’re fully invested, at the bottom and the top.Peter Lynch, The Christian Science Monitor, 8 December 1987.
To see what a one-third fall means, we put it into our drawdown calculator. For the monthly return we used 2.2%, which is what 29.2% a year works out to per month. In other words, we assumed the fund went straight back to Lynch’s own best long-run pace.
A $10,000 holding became $6,670. Getting back needed about +50% (the screen shows +49.9% because a third was entered as 33.3). And at one of the best long-run paces any public fund has published, the climb back takes about 19 months. At a more ordinary 10% a year, our arithmetic gives about 51 months.
Lynch did not treat 1987 as the worst moment of his career. In a 1996 interview with PBS Frontline he said the crash “wasn’t that scary because I concentrate on fundamentals”, and that 1990, with the big banks in trouble and a recession arriving, “was by far the scariest period”. The fall that frightened him was the one in which the businesses themselves were in danger, not the one in which only their prices were.
How often did Lynch say markets fall?
Often enough that he thought planning for each fall was a mistake. In a column for Worth magazine in September 1995, titled “Fear of Crashing”, he counted 53 declines of 10% or more in US stocks since the turn of the century, roughly one every two years, and 15 declines of 25% or more, roughly one every six. Then he wrote the line he is best known for:
Far more money has been lost by investors preparing for corrections or trying to anticipate corrections than has been lost in corrections themselves.Peter Lynch, “Fear of Crashing”, Worth, September 1995.
In the same column he argued that people who sell to wait for a crash tend to keep sitting in cash after the bottom, and miss the recovery that pays for the whole holding period.
Over a thirty-year investing life, his figures imply about 17 declines of 10% and about 5 of 25% (our arithmetic: 53 ÷ 95 years × 30, and 15 ÷ 95 × 30). A fall is not a rare event that might happen to you. It is the normal weather, and the only open question is whether your position is sized to stand in it.
We ran the same count on Bitcoin. Using daily closing prices from Bitstamp since 1 January 2013, we counted every fall from a running high to the lowest close before that high was beaten again.
Lynch’s stocks fell 25% or more about 1.6 times a decade. Bitcoin has done it about 8.7 times a decade, and six of those twelve falls went past 50%. If Lynch’s rule was to expect a 10% fall every couple of years, the crypto version is to expect a 25% fall most years and a halving every two or three.
PRACTICE CORNER
Lynch’s point was that the decision about a fall should be made before it starts, not during it. Write down three numbers for each holding you keep for the long run: what you will do if it falls 10%, 25% and 50% from today. Then check the size: if a 50% fall in that holding would force you to sell to pay a bill or to calm down, the position is too big for the plan, whatever the plan says. Most exchanges let you set price alerts at those three levels, so the moment arrives with the decision already written.
These are affiliate links: we earn a commission if you sign up, it costs you nothing, and it does not change what we write. See our disclosure and the full exchange comparison.
Where the legend and the record disagree
So the famous number is unsourced. What survives is smaller and still useful. Lynch did tell Frontline that in his first three years running Magellan, about a third of the shares were redeemed: many early investors left long before the record was made. And Micropal’s figure means that someone who held the best-known fund of the decade for any single year had roughly a one-in-six chance of ending that year with less than they started with.
We repeat the correction here because this section of the site promises to print the record, not the legend. The lesson people draw from the myth, that investors hurt themselves by selling in falls, may well be true. It is just not proven by a Magellan study that no one can find.
What does Lynch’s record mean for a crypto holder?
The crypto version of October 1987 is not hypothetical. Here is Bitcoin’s most recent large fall on the TradingPrimer live chart, captured on a phone:
That is a bigger hole than Magellan’s. A 53% fall needs about +112.6% to undo, and at the time of capture Bitcoin was trading near $82,600, still about a third below the high. Here is how recovery time scales with the size of the hole, at Lynch’s pace and at a more ordinary one. These are our numbers.
| Fall from the high | Gain needed to get back | Months at 29.2% a year | Months at 10% a year |
|---|---|---|---|
| 10% (a Lynch “correction”) | +11.1% | 4.9 | 13.3 |
| 25% | +33.3% | 13.5 | 36.2 |
| 33.3% (Magellan, Oct 1987) | +50.0% | 19.0 | 51.0 |
| 50% | +100.0% | 32.5 | 87.3 |
| 80% | +400.0% | 75.4 | 202.6 |
The table assumes the money keeps earning a steady pace from the bottom, which no market does. Its real use is the shape: a fall twice as deep takes far more than twice as long to repair. Lynch could stay fully invested through a third because his holdings were businesses with earnings and his fund carried no borrowed money. The bigger the hole an asset can dig, the smaller the share of your account it can be if you mean to sit through it. The emotional side of that, why people sell near the bottom even when they planned not to, is the subject of the emotional cycle (Stage 10, Lesson 48).
Where this lesson stops working
He owned businesses, not just prices. Lynch said 1990 scared him more than 1987 because the companies themselves were in trouble. Staying in through a fall assumes something underneath will recover. Many individual coins have no earnings and no guarantee of coming back; some never revisit their old highs. Sitting through a fall in an asset that is broken is not patience.
Fully invested is not leveraged. Magellan fell a third and survived because no lender could force it to sell. A one-third fall with 3× leverage is a wiped-out account before the bottom arrives. Lynch’s advice assumes you own what you hold outright.
It is advice for long-term holdings, not trades. If you trade a setup with a stop, the stop is your plan; “stay invested” does not mean cancelling it because a famous investor sat through 1987. And money you need within a year or two does not belong in an asset that can halve in a year. A long-horizon approach to deep falls, buying in planned lots, is laid out in cycle investing.
Common mistakes when learning from Lynch
Quoting the 29.2% without the third. The average includes the week it lost a third, and you only earn it if you are still there afterwards. Repeating the “half lost money” story as fact. It is a useful warning with no study behind it. Reading “stay invested” as “never sell”. Lynch sold companies whose story had changed; what he did not do was sell everything because the market fell. Using his stock-market frequencies for crypto. On our count, Bitcoin’s 25% falls have come about five times as often. Deciding during the fall. His whole argument is that the decision has to be made before it, when the stomach is calm.
FAQ
What return did Peter Lynch get at Magellan?
Under Peter Lynch, from 1977 to 1990, Fidelity's Magellan Fund averaged 29.2% a year, trailing the S&P 500 in only two years. By our arithmetic, $10,000 compounding at that rate for thirteen years grows to about $279,500.
Did half of Magellan's investors lose money?
There is no study showing it. In 1996 a Fidelity spokesman told The Globe and Mail that Lynch never said it and that Fidelity had never studied its investors that way. The one firm that measured the fund, Micropal, found it lost money in 17% of twelve-month periods during Lynch's tenure.
How much did Magellan lose in the 1987 crash?
About a third of its value in less than a week, according to Lynch speaking to The Christian Science Monitor in December 1987. Assets fell by more than $4 billion from about $12 billion, almost all of it from falling stock prices, and Lynch kept the fund fully invested.
What did Peter Lynch say about market corrections?
In Worth magazine in September 1995 he counted 53 declines of 10% or more in US stocks since 1900 and 15 of 25% or more, and wrote that far more money has been lost by investors preparing for corrections or trying to anticipate them than in the corrections themselves.