When the Market Drops, I Do Nothing. Not Because I'm Disciplined — Because I Deleted the Decision.

In the first ten days of September the S&P 500 slid from around 7,760 to about 7,600, and the semiconductor index had quietly given back more than a fifth of its value since June.

I know all of that because I looked it up while writing this. On the mornings it was actually happening I was warming a bottle at five, and I checked nothing.

That is the whole article, really. But the interesting part is what it is not. It is not discipline. When the market drops I don't sit on my hands heroically, breathing through the urge to act. There is no urge, because there is no decision left to make. I deleted it years ago, on purpose, and what's left in its place is a kind of boredom that took me an embarrassingly long time to recognise as the product working.

The decision I deleted years ago

Our money leaves the account on payday. Not at the end of the month, not when I've decided how I feel about the news. On payday, automatically, into one accumulating world ETF through Interactive Brokers.

Before our daughter arrived we were putting away around 60% of household income. Since she arrived it's closer to 50%, which we modelled in advance rather than discovered in a panic, and it cost us about six extra months on the date. I wrote about what the baby actually did to our FIRE date at the time.

Here is what that setup does to a falling market. The purchase happens anyway. So a red month means this month's shares were cheaper than last month's. That is the entire emotional content of a crash for someone in my position, and I want to be honest that "my position" is doing a lot of work in that sentence. I'm still accumulating, roughly six to eight years from a coast number around 40. Nobody is sending me a withdrawal. A drawdown is a discount for me. For someone two years into drawing an income off a portfolio it's a genuinely different problem, and I'd be a fool to pretend my calm generalises to them — sequence-of-returns risk is a real thing and it isn't mine yet.

"Buying the dip" is panic selling with better manners

This is the part where I lose people.

Panic selling is when a price move tells you to act, and you obey. Buying the dip is when a price move tells you to act, and you obey. Same reflex, different direction, and only one of them gets applause in the comments.

And the dip-buying version is arguably the harder trick, because you have to be right twice. Right that this fall is the buying moment rather than the first third of a longer one, and right about when to stop waiting for a better price. Get the first call right and the second one wrong and you're the person who correctly identified March, then sat in cash until June.

AQR ran the numbers on exactly this, and it's the most useful thing I've read on the subject because it isn't rhetorical. They built 196 systematic buy-the-dip strategies — different dip definitions, different holding rules, no hindsight — and ran them on US equities from January 1965 to September 2025. More than 60% of the variants produced worse risk-adjusted returns than simply holding. The average Sharpe ratio came out 0.04 lower over the full period and 0.27 lower over 1989 to 2025. Only 8% of the 196 showed statistically significant outperformance. That is roughly what you'd expect from noise.

Nick Maggiulli ran the friendlier version of the same test: a timer with actual divine powers, buying at the exact monthly low every month for decades. Even God underperformed plain monthly buying in more than 70% of the rolling 40-year periods. Not because the timing was bad. Because the money spent all that time waiting instead of compounding.

Vanguard's research points the same way from the other side: investing a lump sum immediately beat spreading it out between 61.6% and 73.7% of the time, depending on the market, across rolling one-year periods from 1976 to 2022. Which is the opposite of what dip-buyers quote it for. It says get money in sooner. It says nothing about waiting for red.

So my position, stated plainly so someone can argue with it: most advice about market drops is written for a person who has a decision to make that day. If you built the thing properly, you don't.

The last time I trusted that muscle, it sent me an invoice

I've told this story properly elsewhere, so here's the short version: I got lucky on a single stock in my early twenties, concluded I could read markets, and went looking for the next one. It was Cardano. About a third of what I had at the time, which was almost nothing by today's standards and was very much everything by mine.

It doubled. Then it fell apart in days and never came back.

The expensive part wasn't the money. It was that I held on through the collapse because I was certain it was coming back — and I was certain because I'd been right once before. That is the dip-buyer's engine, exactly. A fall showed up, and I read it as information about my own judgement.

I still own that same position. It's a rounding error now and I have no intention of selling it, because it appears in my portfolio breakdown every single month and reminds me what my ego costs when it gets access to the buy button. Cheapest subscription I pay for.

When the market drops, nothing is actually happening

Two charts, because this is the one place where data beats my opinion.

Data: S&P 500 intra-year maximum declines and calendar-year returns, 1980-2024, via One Day In July's reproduction of the standard J.P. Morgan Guide to the Markets chart. Past performance is not a promise of future returns.

The average year in that chart contains a 14% fall at some point. And 34 of those 45 years still finished positive. Every year has a scary drop in it. Most of the scary drops are, in hindsight, the price of admission rather than an event.

Data: Capital Group, "How to handle market declines," S&P 500, 1954-2025. Past performance is not a promise of future returns.

A 5% dip shows up about twice a year and lasts about 46 days. A 10% correction turns up every year and a half. A proper 20% bear market arrives about every six years and takes a bit over a year to play out. Capital Group also point out that every decline of 15% or more since 1929 has eventually been followed by a recovery, with an average gain of 52% in the first year after the bottom.

None of that is a forecast. It's a description of the weather in a place I've decided to live.

The number I watch is a date, not a balance

This is where the tool I build actually earns its keep, and where it also gets used less than you'd think.

During a fall I don't open it more. I open it less. The only figure that changes my behaviour is the projected date, and a drawdown moves that date far less than a change in our savings rate does — which is the quiet reason I care much more about what we spend in a bad month than about what the market did.

My wife and I have been through a couple of 10% drops now, and those are plenty visible in a portfolio of any size. We talked about it once, for maybe two minutes, over one of our monthly coffee-and-bank-statement sessions, and haven't needed to since. I'd like to take credit for that alignment. Honestly, a plan neither of us has to revisit during bad news is just harder to argue about than one of us would prefer.

The one exception I allow myself

I'm not a monk about this, and pretending otherwise would be dishonest.

When a month leaves us with more cash than the plan needed, that cash gets invested sooner rather than sitting around waiting for a date. That's it. That is the whole exception, and notice what it isn't: it isn't keeping powder dry for a fall, and it isn't sized by how red the screen is. It's the Vanguard finding applied to money I already have.

What I won't do is touch the emergency fund. Six to twelve months of essential needs sits in a plain high-yield savings account, plus more than a year of general runway on top, and yes, I know that money is losing a slow argument with inflation. I keep it oversized deliberately and I take that loss on purpose, because its job is to let me be honest in a meeting, not to be ammunition. The fastest way to turn a market drop into an actual disaster in our house would be to spend the thing that makes drops survivable.

What would actually make me act when the market drops

Not the size of the fall. I want to be specific here, because "I'd never change anything" is a pose and I don't believe it.

What I watch is exposure, not price. I can see our geographic and sector breakdown, and that is what a decision would be based on: a region genuinely breaking — war, or a slide from democracy into something else — that changes the long-run case for owning it. Then I'd rebalance where the money sits. That's a decision about the world, and it would look identical whether the market were at a high or down 30%.

A number going down is not information about the world. It's information about what other people were willing to pay this week.

Where my own argument falls apart

Four places, and I'd rather name them than have someone name them for me.

The first is that AQR's average buy-the-dip variant did produce a small positive excess return, about 0.5% a year. So dip-buying isn't stupid. It's unreliable, which is different and in some ways worse, because unreliable things work often enough to teach you the wrong lesson.

The second is Warren Buffett, who ended the first quarter of 2026 sitting on a record $397 billion in cash. The most patient investor alive keeps dry powder and waits for prices. I have no clever answer beyond the obvious one: he's doing a job I'm not doing, with an information advantage I don't have, and his own advice to people like me has never been to copy that part.

The third is that my "do nothing" has only ever been tested while a salary kept landing. A drop while I'm accumulating is genuinely cheap for me. I have not yet found out what I'm like when the same chart appears and no payday is coming.

The fourth is the uncomfortable one. My boring world fund is not the neutral thing I like to call it. The largest seven US companies now make up roughly a third of the S&P 500 — the highest concentration in modern market history, against a long-run average of about 24% for the top ten — and the US is close to 60% of the "world" in most global index funds. I still think that concentration is an outcome rather than a forecast I have to be right about, and I'd rather not choose which tech to own than choose wrong. But "I deleted the decision" is doing some quiet work there too. The index made a decision. I just didn't make it myself.

The boring part is the only hard part

Deleting the decision took an afternoon. Living with it has taken years, and the years are the bit nobody sells you.

The setup is genuinely trivial: money leaves on payday, buys the fund, and I find out what happened later. The hard part is doing that through a September where everyone around you found the word opportunity again, and the article you just read is in your feed, and your ego is standing at your shoulder pointing out that you did read about the chip cycle and you do work in tech.

That's what the Cardano line in my portfolio is for. Not sentimentality. A monument to the last time I thought a falling price was talking to me specifically.

If I get one thing across to our daughter about money eventually, I'd like it to be this: the most profitable thing I ever did was make a decision once, properly, and then take away my own ability to revisit it every time the news got loud. Which is why, when the market drops, the most useful thing I own is a setup that doesn't ask me what I think.