In January 2018 I split a $20,000 starter portfolio in half. One half — call it Portfolio A — went into a single broad-market index ETF (VTI), with dividends set to automatically reinvest. The other half — Portfolio B — became my "active" sleeve, where I would pick individual dividend-paying stocks, hold each for at least a year, and reinvest dividends only into the names I most wanted to add to. I kept both halves separate in two sub-accounts at the same broker and logged each year-end balance to a spreadsheet.
This was not a real experiment. The sample size is one of each, the picks were not randomized, and my two halves drifted apart in ways that were partly market-driven. But eight years on, the gap is wider than I had expected, and the reasons matter.
The experiment, briefly
Portfolio A: $10,000 into VTI. Automatic monthly contribution of $250 added through July 2024, then paused. Dividends reinvested mechanically. Total contributions over the period: $19,750.
Portfolio B: $10,000 split across five individual dividend names. Same $250 monthly contribution, allocated to whichever name I most wanted to add to that month. Dividends paid in cash, reinvested manually, sometimes into the position that paid them, sometimes into a different name. Same $19,750 of contributions.
Same broker, same start date, same tax treatment, same external macro environment. Different process.
What actually happened
| Metric | Portfolio A (index, DRIP) | Portfolio B (active picks) |
|---|---|---|
| Total contributed | $19,750 | $19,750 |
| Value at Dec 31, 2025 | $36,840 | $30,720 |
| Total return | +87% | +56% |
| Decisions made (8 years) | ~12 (rebalancing only) | ~190 |
The numbers above are pre-tax. Post-tax the gap narrows slightly — the active sleeve had higher turnover and more capital-gains realizations along the way — but Portfolio A still wins by enough that the explanation cannot be tax timing.
To my surprise, even within Portfolio B, the names I added to mechanically — the ones I bought as part of an automatic dollar-cost-average schedule — outperformed the names I "researched and chose." The smaller subset of B that was mechanical did roughly +73%, while the truly active subset did about +44%. The active half of the active sleeve lost to the index by a wide margin.
Why the boring side won
Three reasons, in order of magnitude.
Concentration penalty. Portfolio B held five to seven names at any given time. Even with reasonable picks, one of those names dropped 60% in 2022 and never recovered. The same drop, in a portfolio of 4,000 names, would have been a rounding error.
Behavioral drift. Across 190 manual decisions, I made a handful of bad ones — selling a position into a panic that bounced back, holding a position past a fundamental change. I am not unusually undisciplined; the count of decisions itself was the problem. The index half made roughly twelve.
Dividend reinvestment, automated, was mechanical and fast. Every dollar of dividend got reinvested at the next-available price. In Portfolio B, dividends accumulated as cash and I batched them — sometimes letting cash sit for weeks while I tried to decide which name to add to. Those weeks were drag.
The hardest argument against an active strategy is not that it underperforms. It is that the underperformance is mostly your own fault, and you cannot stop being yourself.
The cost I underestimated
The big number is the return. The small number is the hours spent. Across eight years I conservatively spent 200 hours managing Portfolio B — reading 10-Ks, monitoring earnings, deciding when to buy or trim. Across the same period Portfolio A took maybe fifteen hours, mostly the annual rebalance.
If I value my time at $40/hour — well below my freelance writing rate, which is the relevant comparable — that is $8,000 of opportunity cost. Portfolio B underperformed by $6,120 in raw terms and took $8,000 of my time to do it.
What I do now
The split is dead. I closed Portfolio B in early 2024 — sold the worst three names, transferred the remaining four into the same brokerage account as VTI, and let everything drift toward index-shaped. Roughly 80% of my equity exposure is now broad-market index funds, mostly VTI and VXUS. The remaining 20% sits in a handful of individual dividend names I have held for so long that selling them would create a capital-gains event I would rather not pay.
I still read 10-Ks. I still write about individual companies. I no longer try to outperform an index with my taxable money. The experiment cost me less than tuition would have, and I learned the same thing.
One small footnote about reinvestment. The single switch that probably mattered most across Portfolio A's life was the default to "reinvest dividends." Every broker offers this; you tick a box once. It is the closest thing to free outperformance available to a retail investor over a multi-decade window. If your dividends are sitting as cash in a brokerage account right now, tick the box this week.