The ten largest companies in the S&P 500 now account for roughly 40 percent of the index's total value, close to double the share they held a decade ago and above the level reached at the peak of the dot com bubble. By any conventional textbook rule, a portfolio that has drifted this far from its intended weights is overdue for a haircut. Yet a growing body of quantitative research is telling investors the opposite. Trim those winners on schedule, it argues, and you are not managing risk. You are fighting the single most persistent force in markets: momentum.
That is the split running through portfolio management right now, and it is sharper than the usual hedging language of financial media lets on. On one side sit the risk control traditionalists: Vanguard's research desk, wealth managers such as the Bahnsen Group, and decades of behavioral finance research built on the work of Nobel laureate William Sharpe. On the other sit the momentum realists, a camp backed by trend following quants, index reconstitution researchers at Research Affiliates, and a wave of recent backtests showing that mechanical rebalancing has been a quiet drag on returns for years. Both sides are working from real data. They just disagree about what that data is telling them to do next.
The Discipline Camp: Rebalancing as Risk Insurance, Not a Return Strategy
The traditionalist case starts with a distinction most retail investors miss entirely: rebalancing was never sold as a way to boost returns. Vanguard's analysis of the 1926 to 2019 period found that a periodically rebalanced 60/40 portfolio produced a higher Sharpe ratio, meaning better return per unit of risk taken, than an unbalanced version, even in stretches where its absolute return was lower. Sharpe's own research reaches the same conclusion: the primary benefit is risk control, not alpha generation. A more recent Vanguard paper on target date funds goes further, showing that a threshold based approach, rebalancing once an allocation drifts past a set band rather than on a rigid calendar, can actually reduce transaction costs while holding risk closer to target than either a monthly or quarterly schedule.
Advisors making this case point to history rather than models. The RBC Wealth Management concentration study traces how the platform economy pushed the top ten weight from roughly 19 percent in 1990 to a first peak near 27 percent during the 2000 tech boom, an episode that ended in a sharp and painful unwind. Guinness Global Investors draws the same parallel to financials before the 2008 crisis and to pandemic era lockdown winners in 2020, arguing that narrow leadership has repeatedly amplified downside once sentiment turned. JPMorgan's asset management team adds a valuation layer to the argument: the top ten stocks currently trade near 26 times earnings, about 25 percent above their long run average, a premium that barely moved even during this year's spring pullback. Put simply, the case is not that today's mega caps are bad businesses. It is that an unrebalanced portfolio is quietly making an enormous, undiversified bet on a handful of them continuing to defy gravity at a premium price.
The Momentum Camp: Selling Winners Fights the Tape
The counter argument does not deny any of that data. It challenges the assumption sitting underneath it, that markets reliably mean revert on a time horizon short enough for a calendar rebalance to catch. A widely cited industry analysis of trending markets found that rebalancing actively hurts returns when a trend is underway, because letting an outperforming allocation grow larger captures a compounding effect that a scheduled trim simply cuts off. A recent fourteen year backtest across four rebalancing frequencies reached a similar conclusion: shorter rebalancing windows repeatedly fought short term momentum across 168 separate rebalancing events, and that friction compounded into a measurable performance gap over time. Momentum, in other words, is not a fringe theory. It shows up in the return data of AQR's momentum focused funds, which have posted some of their strongest relative years precisely during the stretch when concentration has been climbing.
There is a second, less obvious version of this argument, and it comes from inside the indexing world itself. Research Affiliates' Vitali Kalesnik has documented what he calls conventional indexing's own rebalancing problem: in the year before an S&P 500 index committee adds a stock, that stock tends to outperform the one it is replacing by roughly 83 percent, only to give back about 23 percent of that edge in the year after the swap actually happens. That is a structural buy high, sell low pattern baked into the mechanics of index reconstitution itself, and it is a reminder that rebalancing of any kind, whether it is an index committee's or an individual investor's, has real, measurable costs that are easy to romanticize away as pure discipline.
Same Data, Different Clocks
Here is why two camps can look at an identical concentration chart and reach opposite conclusions: they are answering different questions on different time horizons. The discipline camp is optimizing for risk adjusted outcomes and defending against a tail event, the low probability but high consequence scenario where a handful of mega caps correct sharply and take an unrebalanced portfolio down with them. The momentum camp is optimizing for the base case, the much more probable scenario where a trend simply continues for another year or two, and mechanical selling shows up as a cost with no offsetting benefit. Both effects are real. Momentum genuinely persists over months to a few years. Mean reversion genuinely tends to win over full market cycles. The rebalancing decision is really a bet about which clock is closer to running out, and nobody rings a bell when it does.
Where the Evidence Actually Leans Right Now
Given where things stand, the stronger case currently belongs to the discipline camp, though not for a rigid, calendar driven version of it. The scale of today's concentration is not a modest deviation from historical norms; it is a record, sitting above the dot com peak, with valuations to match. Goldman Sachs' Global Investment Research group, in analysis cited by Tema ETFs, has tied current concentration levels to depressed forward ten year return estimates for the index as a whole, based on the historical relationship between concentration and subsequent performance. That is not a timing call on next quarter. It is a statement about the odds an investor is accepting by leaving a portfolio's risk profile unmanaged at these levels. The smarter middle ground, and the one Vanguard's own research supports, is threshold based rebalancing rather than a fixed date on the calendar: let a position run within a defined band so short term momentum is not fought unnecessarily, but act once genuine drift breaches that band, using tax sheltered accounts and new contributions first, a sequencing Morningstar's Christine Benz has long recommended to avoid triggering needless tax bills on appreciated winners.
What to Watch
The tiebreaker will not come from another op ed. It will come from market breadth: whether gains start showing up outside the current mega cap leaders, whether the effective number of stocks driving the index's returns starts rising again, and whether that 25 percent valuation premium on the top ten begins to compress even as their earnings keep growing. Bond yields near 4 percent currently make the destination side of a rebalance genuinely productive rather than a forced retreat into dead weight, which lowers the cost of erring toward discipline. Watch those three signals over the next few quarters, and you will have a better answer than either camp's models can currently give you on their own.