The S&P 500 Is No Longer Diversified - And That Changes Everything About Vanguard ETF Advice

The false narrative here isn't hard to spot. Financial writers line up three S&P 500 ETFs, compare their expense ratios to three decimal places, and declare winners. The headline always promises a choice. The reality is that VOO, IVV, and SPY track the same index, hold the same stocks, and deliver nearly identical returns. The actual decision isn't which fund to pick. It's whether the S&P 500 itself still functions as the diversified anchor investors think they're buying.

Because it doesn't.

The S&P 500 used to be 500 companies spread across sectors. Today it's ten stocks with 490 tag-alongs. Vanguard's flagship S&P 500 ETF, VOO, which traded at $686.65 as of July 31, 2026, and carries $973 billion in assets, holds NVIDIA at 7.50%, Apple at 6.58%, Microsoft at 4.29%, and Amazon at 3.61%. The top ten holdings represent 36.33% of the fund. The broader index concentration - roughly 40% by market cap - has exceeded the peak of the dot-com bubble, when the top ten held around 27%. During a 28-session rally between late March and early May 2026, Nomura analysts found that just ten stocks drove 69% of the index's gains. The other 490 companies were passengers.

That is not diversification. That is a concentrated tech bet wrapped in an index fund costume.

The valuation data doesn't make this less uncomfortable. The S&P 500's Shiller CAPE ratio - the cyclically adjusted price-to-earnings metric that smooths earnings over ten years to strip out temporary spikes - sits at 41.37 as of July 2026, up from 37.47 a year ago. The trailing twelve-month P/E was 25.3 as of June 2026. Goldman Sachs notes the forward P/E of roughly 21 times earnings ranks in the 88th percentile relative to the past 40 years. These aren't numbers that scream "panic sell," but they are numbers that demand position sizing discipline. You are buying earnings power that the market has already priced for exceptional performance.

Then there's the dividend problem. VOO's trailing twelve-month dividend yield is 1.057%. The broader S&P 500 yield stood at 1.09% as of June 23, 2026, which is 33% below its long-term average of 1.63%. The year-over-year decline in yield was 14.64%. For an investor who needs income, or who wants some margin of safety from cash returned to shareholders, a 1% yield at a CAPE of 41 is a thin rope to hang a portfolio on. Companies in the index are returning capital through buybacks rather than dividends - and while buybacks aren't inherently bad, they don't carry the same commitment as a dividend. Dividends are a promise; buybacks are a transaction, and companies can stop buying back stock on a quarterly basis without any announcement.

The earnings growth story is the bull case, and it's not hollow. FactSet projects S&P 500 earnings growth of roughly 15% for calendar year 2026 - the third straight year of double-digit growth, with all 11 sectors expected to contribute. Q2 2026 earnings growth was projected at 23.3%, with Goldman Sachs upgrading its full-year EPS forecast to $340, representing 24% annual growth, and raising its S&P 500 year-end target to 8,000 from 7,600. AI infrastructure beneficiaries alone are expected to account for roughly half of that growth. Hyperscalers are on track to spend $754 billion on capex this year, up 83% from 2025, with $905 billion expected in 2027.

But the earnings growth paradox is that the very strength driving the index is also the source of its concentration risk. When half of the S&P 500's earnings expansion comes from companies that are already the largest holdings, the market isn't broadening - it's leaning harder on the same names. Goldman Sachs flags that a sharp increase in momentum and narrow market breadth are emerging as cautionary signals. The sustainability of this momentum depends on whether enterprise AI investment translates into recurring profits. Goldman embeds just a 0.4 percentage point boost to S&P 500 EPS growth from AI productivity in 2026, rising to 1.5 points in 2027. That is a small fraction of the capex being thrown at the problem.

So back to the ETF question itself. If you're buying S&P 500 exposure, VOO at 0.03% expense ratio is the rational choice. Its 5-year annualized return of 13.56% is functionally identical to iShares Core S&P 500 ETF (IVV) at 13.60% and SPDR S&P 500 ETF Trust (SPY) at 13.64%. SPY charges 0.095%, roughly three times VOO's fee, which translates to $6.45 more per $10,000 invested annually. Over a 20-year horizon, that compounds to a real difference. VOO and IVV are interchangeable at this point. You only need one.

The real question that the "three best ETFs" article never asks is what your S&P 500 allocation should look like at these valuations. In my opinion, the answer depends on whether you're an income investor, a growth accumulator, or someone sitting on dry powder waiting for a better entry. VOO's 10.12% year-to-date return through July 31, 2026, is impressive - but the trailing 19.53% one-year return means anyone buying now faces a steep recovery bar if the index corrects. A CAPE of 41 historically implies lower returns over the next decade. That doesn't mean a crash is coming. It means the math works against you if you're allocating a disproportionate share of your portfolio to U.S. large-cap growth at these multiples.

For investors who can tolerate short-term volatility and have a long time horizon, a core S&P 500 ETF position still makes sense as part of a diversified framework. VOO is the cleanest way to get it. But that core should be balanced against higher-yield income positions, international value exposure, and perhaps a defensive allocation to gold or cash that provides ballast when the ten stocks everyone owns start moving in the same direction. The S&P 500 hasn't stopped working as an index. It just stopped being what investors think they're buying.

Bottom line: VOO is the best Vanguard S&P 500 ETF because it's the cheapest and tracks the index precisely. I rate it a Buy for investors who understand that they are really buying a concentrated tech-forward portfolio, not a diversified basket of 500 companies. The decision isn't about picking between VOO, IVV, and SPY. It's about sizing your U.S. large-cap allocation relative to everything else when the CAPE is 41, the dividend yield is 1%, and ten stocks control 36% of the fund.