I started buying my first ETF in late 2019. I was 27, working in Seoul, and had been reading Bogle and Malkiel for a few months. The plan was simple: open a Korean brokerage’s overseas account, buy a few shares of VTI every month, and let it compound for thirty years. The plan made it fourteen weeks before reality kicked it in the teeth.
This post is my honest record of that first year — the one where COVID broke out, I panicked, I sold, and I learned that DCA is much harder to actually do than to read about. If you’re about to start your own first year, I wrote the post I wish someone had handed me in February 2020.
The First Three Months: Easy Mode
January 2020. VTI was around $165. DCA — dollar-cost averaging, the practice of buying a fixed dollar amount of an ETF on a schedule regardless of price — felt obvious in the way obvious things often do before they get tested.
I bought a small slice every two weeks. In January and February the market drifted up. My account balance went up. I refreshed the brokerage app probably once a day, sometimes twice, the way you’d check a video game where you’re winning.
If my first year had ended in February, this would be a useless post. Everyone is a long-term investor in a bull market. The interesting test of DCA is what happens when the price goes the wrong way for a long time. Mine arrived in week thirteen.
March 2020: The Sell Button
VTI dropped from roughly $165 to $117 in five weeks. That’s about -29% from peak to trough — and the speed was what broke my plan. The 2008 crash, which I’d read about in books, happened over more than a year. The COVID crash compressed a similar move into a month. There was no time to mentally prepare. Every morning the market gapped down, and the brokerage app started showing my carefully accumulated position as a bigger and bigger negative number.
The book I had been reading was clear: keep DCA-ing through drops. Buy the dips. The lower price is a feature.
I held the line for about three weeks. Then on a specific Tuesday in mid-March, after the worst news cycle of my adult life, I sold most of my VTI position. Not all of it. About 70%, at an average price somewhere around $128.
Here is the part of the story that is hardest to write honestly: I did not sell because I had a thesis. I sold because I could not stand watching the number drop one more time. I told myself I’d buy back when “things stabilized.” I did not have a definition of stabilized. I think, in retrospect, I meant “when I felt comfortable again.”
What Stabilized Means, In Hindsight
The bottom of the COVID crash for VTI was March 23, 2020 — about $117. By the end of April, VTI was already back near $145. By July it had reclaimed the February peak. By the end of 2020, VTI closed near $192, up roughly 16% from where the year started.
I did not buy back near $128. I bought back, slowly and embarrassedly, in late summer at prices in the $160s and $170s. The shares I sold at $128 cost me roughly $40–50 each to replace. On the partial position I had panic-sold, that’s a permanent loss of around 25–35% of those shares’ future compounding base. Not a temporary loss. A real, locked-in loss against a counterfactual where I’d done nothing.
Here’s what that looked like in numbers, simplified. Say I had 100 shares at the moment I sold, average cost $145.
| Action | Shares | Avg cost basis | Approximate net result by year-end 2020 |
|---|---|---|---|
| Hold all 100 shares through the crash | 100 | $145 | ~$192 × 100 = $19,200 |
| Sell 70 at $128, rebuy 70 at $170 | 100 | ~$157 (blended) | ~$19,200 ending value, but with ~$2,940 lost in the round-trip vs. holding |
The end-of-year value looks similar in dollar terms because the recovery was so fast. The damage shows up in the cost basis — the per-share price you’ll be measured against forever — and in the lost compounding on the dollars that exited the market at the bottom and didn’t fully come back. (For the uninitiated: cost basis is the price you paid for a share. It’s used to calculate taxes when you eventually sell, and it’s also the mental anchor against which you measure gains and losses.)
In a slower bear market, the damage would have been much worse. I got lucky on the timing of my mistake.
The Lesson the Math Teaches
Burton Malkiel writes that missing the best 10 days in any given decade typically wipes out the majority of that decade’s return. I had read that sentence. I did not believe it the way I needed to believe it.
The COVID crash version: between March 23 and April 30, 2020, VTI returned roughly 27% in five weeks. If you sold on March 17, like I mostly did, and waited for “stabilization” before buying back, you missed essentially all of that move. The single best week was the one immediately after the worst week. There was no signal you could have used to catch the bottom — the bottom only became visible in retrospect, after the recovery had already happened.
This is the practical version of “time in the market beats timing the market”:
- The worst days of a bear market are usually clustered next to the best days of the recovery.
- Selling at the bottom feels like risk reduction. It is actually the most expensive form of risk-taking — locking in losses before participating in the rebound.
- The investors who get the recovery are mostly the ones who didn’t sell. Not because they were brave. Because they didn’t act.
What I Wish Someone Had Told Me Before March
Three things, in plainer language than the books I had been reading.
One: your DCA plan is the plan. It is the entire plan. The plan is not “DCA when the market goes up; revisit the plan when it goes down.” If you have to revisit the plan during a drawdown, the drawdown will revisit you in the worst possible way. Either you DCA through everything, or you don’t have a DCA plan, you have a moving target.
Two: the book wisdom is correct, but it is harder than the books make it sound. Reading “stay invested through the drawdown” feels like an obvious technical instruction. Doing it while watching your account go red every morning for thirty straight trading days is an emotional task on a different scale. If your plan depends on you behaving rationally during a 30% drop, you need to think harder about what your plan actually requires of you.
Three: build the plan for the version of yourself that can break. This is the part that took me until year three to internalize. The 27-year-old me who built the DCA plan was a calm-day version of myself who had read books. The crash-day version of me was a different person — one who had not slept well, one who was afraid of losing my savings during a global emergency, one who could not see the whole horizon. The plan needs to work for the worst version of you, not the best.
For me, that translated into three concrete adjustments by year two:
- Automate the contributions. If the buy happens without me clicking a button, the worst-case version of me has fewer chances to override it. (Some Korean brokerages don’t support automatic ETF purchases for overseas accounts; in that case, the best substitute I found is a calendar reminder on the same day each month, treated as non-negotiable.)
- Stop checking the app daily. I now check it weekly at most, often less. The frequency of looking is the frequency of having opinions, and the opinions, in 2020, were the problem.
- Pre-decide what I’d do in a 30% drop. I wrote a one-page note on what I’d do at -10%, -20%, -30%, and -40%. The -30% answer is “buy double, automatically, if cash allows.” Pre-deciding turns a panic moment into an instruction-following moment.
A note from my own portfolio: the thing that survived from year one was not my returns. It was the rule that I would never again sell during a drawdown, no matter what the news cycle said. That rule has been tested twice since — once in 2022 (-19.5% calendar year for VTI, about -26% peak-to-trough), once in a faster early-2025 drop. Both times I held. Both times the recovery happened without my participation in any clever way. I just kept buying.
The Six-Month Update I Did Not Want to Write
By early September 2020, my year-to-date return was modestly positive — around 6–8%, depending on how I’d handled the rebuy timing. Holding straight through would have been about 3–5% better than that. The market was at all-time highs.
I had spent six months feeling smart for selling and stupid for buying back. The lesson the year was teaching me wasn’t “you sold at the wrong time.” It was “you cannot trust your instincts during a drawdown to make any decision better than the do-nothing default.”
That is the lesson DCA is supposed to teach. The book versions had not landed. The lived version did.
What Year One Looked Like, Stripped of Emotions
If you’re a beginner trying to imagine what a first year of DCA into VTI realistically looks like, here’s the unvarnished version of mine:
| Phase | What happened | What I did right | What I did wrong |
|---|---|---|---|
| Jan–Feb 2020 | Smooth uptrend | Started DCA on schedule | Built no resilience for a downtrend |
| Mar–Apr 2020 | -29% peak-to-trough | Held for first 3 weeks | Panic-sold ~70% near the bottom |
| May–Jul 2020 | Sharp recovery | Realized I was wrong | Hesitated to buy back, lost most of the move |
| Aug–Dec 2020 | Drift to all-time highs | Bought back, resumed DCA | Cost basis permanently elevated |
End-of-year return was up modestly. End-of-year lesson was the most expensive education I’ve paid for, by a wide margin.
What I’d Tell Someone Starting Their First Year Now
If you’re considering opening a brokerage account and starting your own first year of DCA into VTI or VOO or VT (any broad index ETF), here are the only three rules I’d hand you:
-
Pick one fund and one schedule, then automate or pre-commit. VTI/VOO/VT are all reasonable cores. Don’t optimize the choice — pick and start. The amount you invest matters more than which broad index you chose. If you’re unsure between US-only and global, my view (with 30% international, late 2025 voice) is in the VT vs. VTI + VXUS guide.
-
Pre-decide what you’ll do in a 30% drop. Write it down. Tape it inside your wallet. The decision you make in the moment, on the worst news day, will not be the decision you would have made calmly. Write the calm-day decision down before you need it.
-
Plan to hold for at least three years before judging anything. First-year results in stocks tell you nothing. The 2008 first-year investors who held look like geniuses now. The 1929 first-year investors who held also did fine, eventually, after a very bad decade. The number that matters is your decade, not your year.
If your first year matches mine — emotionally turbulent, modestly profitable, deeply educational — that’s a much better outcome than it might feel like in the moment. The investors who quit after a hard first year are the ones who never get the long-run compounding the books promised them. The ones who stay become the second-year version of themselves, then the fifth-year, then the tenth.
I’m in year six now. The first year was the one that broke me, the second was the one I rebuilt on, and the years since have mostly been quiet. That quiet is, in retrospect, the whole point.
For the broader case for buy-and-hold and the math of why the recovery years matter more than the crash years, see my market corrections guide and what to expect from S&P 500 returns. They are the framework. This post is what the framework felt like during the year I was learning to trust it.