What we test
The hypothesis is simple: investing each year in the largest companies by market capitalization might capture some of the market leaders’ performance. We examine portfolios of one to ten companies—TOP 1–10—against holding SPY. The ranking comes from historical Portfolio123 snapshots: it is an S&P 500 proxy, not a certified official constituent history.
How we construct the portfolios
Each portfolio starts with USD 1 million and selects companies from the previous year’s snapshot. We test equal and market-capitalization weights. Rebalancing takes place at the opening of the first trading session of the year: continuing holdings are retained, only the differences are traded and departing constituents are replaced.
Dividends are included through total returns, and modeled IBKR/ARCA commissions are deducted from portfolio equity. SPY shares the start date and dividend treatment but remains invested without annual turnover. This is a historical simulation, not the record of a real account.
Full-period results
TOP 2, TOP 3, TOP 5 and TOP 6 exceed SPY’s CAGR under both weighting schemes in the main scenario. TOP 2 has the highest observed value, but choosing it after seeing the results would be retrospective selection, not a rule validated for the future.
| Portfolio | Equal weights | Market capitalization |
|---|---|---|
| TOP 1 | 6.47% | 6.47% |
| TOP 2 | 9.70% | 9.46% |
| TOP 3 | 8.27% | 8.13% |
| TOP 4 | 7.89% | 7.79% |
| TOP 5 | 8.31% | 8.27% |
| TOP 6 | 7.97% | 8.25% |
| TOP 7 | 6.62% | 7.45% |
| TOP 8 | 6.18% | 7.05% |
| TOP 9 | 6.66% | 7.24% |
| TOP 10 | 6.20% | 6.90% |
| SPY | 7.95% | 7.95% |
Buying large companies does not eliminate risk. Every portfolio in this window experiences a deeper maximum drawdown than SPY. TOP 1’s maximum drawdown reaches -81.95%, compared with -55.19% for SPY.
The period changes the picture
In 2000–2009, every portfolio underperforms SPY; in 2010–2025, every portfolio outperforms it. TOP 1 moves from a CAGR of -12.21%, compared with SPY’s -1.10% in the first decade, to 20.11%, compared with 14.02% in the later period. These are descriptive windows chosen after reviewing the curves, not an out-of-sample test.
The paper also compares accounts starting in 2001. Both windows are retained: the year 2000 is not removed. The comparison shows how much the starting point matters and what changes when the incomplete execution intervals of three historical securities are avoided.
Study notes
The constituent proxy and certain events are not certified. A neutral execution approximation is used for three historical securities, with ±10% sensitivities; for equally weighted TOP 4 and TOP 6, that stress can change the sign of performance relative to SPY. Spreads, slippage, market impact and taxes are excluded. The V2 extensions on reweighting and volatility-based exposure are analyzed separately in the paper, not in the tracker’s twenty annual variants.
Conclusion
In the 2000–2025 simulation, some portfolios of the largest S&P 500 companies exceed SPY’s annualized return, although all experience deeper maximum drawdowns. The comparison across decades also shows that leadership by market capitalization has not delivered a consistent advantage.
The study does not offer a single answer for or against concentration: portfolio size, weighting and the starting point change the outcome. Observed returns must be read alongside the risk taken and the period analyzed, not as identifying an optimal TOP portfolio for the future.
The PDF retains the complete curves, sources, assumptions, risk tables and appendices.