The Behavior Gap Is Smaller Than You've Heard — and That's the Wrong Comfort

Last week, the main Korean stock index fell for seven sessions in a row — its longest losing streak since September 2024. On Thursday, October 15, it dropped 7.57% in a single day to close at 5,554.22, with circuit breakers halting trading on both of Korea's main markets. Over the week, Samsung Electronics fell 17.8% and SK Hynix 24.3%, their steepest weekly declines since the 2008 financial crisis.

Then, on Friday, the index rose 2.05%. On Monday it added another 1.76%.

I am not offering that as a forecast. Two up days say nothing about the next two weeks. I am offering it because of who was selling on Thursday. Between October 13 and 16, forced sales — positions liquidated because investors could not meet their brokers' margin requirements — topped 50 billion won in every session, 286.2 billion won in all. Cash waiting in brokerage accounts fell by more than 6 trillion won in the same four days.

Some of those people chose to leave. Some had the choice made for them. All of them were living, in real time, the most famous sentence in investing advice: buy high, sell low.

That sentence has a number attached to it, one you have probably seen. This year, that number is under serious challenge. Here is the judgment I keep returning to: the challenge is probably right about the average, and the average is exactly where the damage hides.

The Number Everyone Quotes

Every year, Morningstar publishes a study called Mind the Gap. It compares two returns for the same funds. One is the fund's own return — what you would have earned by investing at the start and never touching it. The other is the return the funds' actual investors earned, given when their money really arrived and when it really left.

The 2026 edition, published in August, covered the ten years through the end of 2025. Morningstar estimated that the average dollar invested in U.S. mutual funds and ETFs earned 8.7% a year, while the funds themselves returned 9.9% a year. The difference, 1.2 points a year, is the "investor return gap," and Morningstar attributes it to the timing and size of investors' purchases and sales. It works out to roughly 12% of the funds' total return over the decade.

In dollars, the study's estimate is large. The funds held about $13.6 trillion at the start of 2016. Left untouched, Morningstar calculates, that money would have grown to nearly $35 trillion. The funds actually held $29.7 trillion at the end of 2025, a shortfall the study largely attributes to about $3.8 trillion in timing-related effects.

The same report contains a detail that gets quoted far less. For U.S. stock funds specifically, the gap was small — investors earned 12.8% a year against the funds' 13.3%. Morningstar called the decade arguably the largest haul in fund history for that group.

The Challenge

In May, the Financial Analysts Journal published a paper with an unusually direct title: "Bad Timing Does Not Cost Investors 15% of Their Funds' Returns." Its four authors — Jon Fulkerson, Bradford Jordan, Timothy Brandon Riley, and Qing Yan — examined Morningstar's previous edition, which had reported a 1.2% gap for 2015 through 2024.

Their conclusion, using the same sample: poor timing by mutual fund investors costs them only about 0.10% a year.

That is a twelvefold difference between two careful readings of the same data. I do not have the standing to settle it, and I would be suspicious of anyone who claims to from the abstracts alone. But the size of the disagreement tells you something useful on its own. The "investor return gap" is not a direct measurement of panic. It is a calculation, and calculations of this kind depend heavily on what you decide counts as timing.

Here is one reason the two approaches can diverge, and it does not require any study to see. A dollar-weighted return — the kind that measures what investors actually earned — is pulled down whenever a lot of money arrives just before a weak stretch. But money does not only arrive because people chase performance. It arrives because paychecks arrive. A worker who adds to a retirement account every two weeks, without ever looking at the market, will still put more dollars to work as the account grows and as markets rise, because that is when there is more to invest. Some of what looks like bad timing in the aggregate can simply be the shape of ordinary saving.

So the honest summary is this. For the average investor, the cost of buying high and selling low may be much smaller than the famous number suggests. That should be good news. It mostly is.

Why the Small Average Is the Wrong Comfort

The trouble is the word average.

Every panic sale has a buyer. When someone sells at the bottom, someone else buys at the bottom, and in a dataset that adds all investors together, the seller's loss and the buyer's gain largely cancel. An aggregate number can look calm while specific people inside it are badly hurt — the same way a lake can be still on average with one whirlpool in it.

Last week in Korea is a clear example of where that whirlpool sits. Forced sales of 286.2 billion won in four days are not spread evenly across every investor in the country. They land on particular accounts — the ones that borrowed, often the ones that bought late, and almost always on the worst days. I wrote about where that borrowing came from earlier this month, and about the anchors that make people hold on and then let go last week.

Neither study, however its numbers are settled, can tell you whether you were the one who sold on Thursday. That is the only gap that ever shows up in your account. A small average cost does not make it small for the person who paid it.

Why Korea Shows What the Averages Hide

Living in Korea gives me an unusual view of this, and it comes from a dull administrative fact.

The Korea Financial Investment Association publishes, every trading day, how much stock was sold because investors could not meet their settlement obligations, how much cash is sitting in brokerage accounts, and how much has been borrowed to buy shares. Korean financial newspapers report those figures within a day or two. During a week like the last one, you can watch forced selling arrive in near real time: 59.3 billion won on Tuesday, 89.3 billion on Wednesday, 80.5 billion on Thursday, 57.2 billion on Friday.

I have not found an American equivalent published daily. The closest public figure in the United States that I know of is the monthly total of margin debt that FINRA reports. A monthly total tells you how much was borrowed. It does not tell you how much was sold out from under people on a particular afternoon.

That difference matters for how the behavior gap gets discussed. In the U.S., the debate happens almost entirely at the level of decade-long averages, because those are the numbers that exist. In Korea, the individual week is visible. And the individual week is where selling at the bottom actually happens.

What Selling at the Bottom Actually Buys

It helps to be precise about why people sell at the bottom, because the explanation is usually told as a story about foolishness, and it is not one.

At the bottom, the pain is certain and the recovery is not. A falling price is felt every hour; a recovery exists only as a possibility. Selling does not buy safety from further losses so much as it buys an end to the feeling — the relief of no longer having to watch. That relief is real, and it is immediate, which is why it is so hard to refuse.

The cost comes afterward and quietly. The relief was purchased at the price of being absent for whatever came next, and the days that come next after a crash are often the strongest of the whole period. I wrote in July about how the market's best days tend to sit right next to its worst. Thursday's 7.57% drop and Friday's 2.05% rebound are one more pair of neighbors. That proves nothing about the future. It is simply what the pattern looks like up close.

You may already know which kind of week makes you want to sell. Most people do. The useful question is not whether you will feel it, but whether anything is set up to keep the feeling from making the decision.

Where This Stops

This is not a view on Korean chip stocks, on the KOSPI, or on whether last week was a bottom. Analysts in Seoul are openly divided on that, and I have no prediction worth publishing.

And it is not an argument that selling is always wrong. Selling because your circumstances changed, because a position grew too large, or because you need the money is a decision. Selling because a week was unbearable is a reaction. The two can produce the same trade on the same day, and only one of them was made in advance.

Questions to Answer Before the Next Bad Week

None of these require predicting anything. All of them are easier on a calm afternoon than on a Thursday like last week's.

Is any of this money borrowed? Borrowed money is the only kind that can be sold for you. If the answer is yes, the bad week is not fully yours to wait out.

How much of a drop could you sit through without acting? Not in theory — in a week like the one Korea just had. The Sunday Night Test is a quick way to answer it honestly.

What, specifically, would make you sell? If you can name the reason now, you will recognize it later. If the only reason you can imagine is "it got too painful," that is worth knowing too.

What will you do with the feeling instead? The urge to act is strongest when acting is most expensive. Having something else to do with it — a written rule to reread, a person to call, a fixed date to revisit — is often what separates a reaction from a decision. A guide to the year your portfolio is down 30% walks through what that looks like over a longer stretch.

Lines to Watch

Not predictions — the figures that would tell you whether the selling is still being forced.

Daily forced-sale totals published by the Korea Financial Investment Association, which stayed above 50 billion won for four straight sessions last week.
Investor deposits, which fell from about 100.4 trillion won on October 12 to about 94.4 trillion won on October 16.
The margin loan balance, at about 32.25 trillion won on October 16 — the lowest since September 10.
The next edition of Mind the Gap, and any reply to the Financial Analysts Journal paper. How that argument is settled will change the number everyone quotes, not the experience of the person who sold.

Averages Don't Sell. People Do.

There is a version of this debate that ends comfortably: investors are not nearly as bad at timing as the famous figure suggests, so there is less to worry about.

For the average dollar, that may well be true. But nobody owns the average dollar. Each person owns one account, and the week that matters for it is not the decade that a study measures. It is the specific Thursday when the screen was red, the margin call came, and selling felt like the only way to make it stop.

That week does not show up in the aggregate. It shows up in one place only, and it is the place you can actually prepare.

Data as of October 19, 2026. Sources: Morningstar, Jeffrey Ptak et al., "Mind the Gap 2026" (August 6, 2026); Jon A. Fulkerson, Bradford Jordan, Timothy Brandon Riley, and Qing Yan, "Bad Timing Does Not Cost Investors 15% of Their Funds' Returns: An Examination of Morningstar's 'Mind the Gap' Study," Financial Analysts Journal, Vol. 82, Issue 3 (May 12, 2026); Seoul Economic Daily, "Samsung, SK hynix Post Worst Weekly Drop Since 2008 Crisis" and "Forced Stock Sales Top 50 Billion Won for Fourth Straight Session" (October 19, 2026), citing the Korea Exchange and the Korea Financial Investment Association.

Related Posts:

Visuals on this post are AI-generated. The author works with AI as a research and drafting assistant; topics, judgments, and final edits are the author's own. This post is observation, not investment advice. See full Disclaimer for details.

Search This Blog

IJin Insight