https://www.morningstar.com/funds/passive-investing-is-fueling-rise-mega-firms-that-could-affect-your-portfolio-unexpected-ways
Link posted to comply with Morningstar’s link-only policy; quotes reproduced under fair-use.
“The surge in passive investing doesn’t just mirror the market — it shapes it, creating new risks and opportunities.” — Morningstar, 26 Jun 2025
Article-in-a-nutshell
| Theme |
Evidence |
| Ownership concentration |
Top-10 stocks now control ≈ 36 % of SPY (SPDR holdings, 29 May 2025). |
| Flow-driven bid |
Passive inflows lift the largest firms the most, cutting their financing costs. |
| Amplification loop |
Higher weights ⇒ bigger passive demand ⇒ still higher weights (“autocatalytic”). |
| Volatility kicker |
Added idiosyncratic σ in mega-caps discourages arbitrage, letting mispricing persist. |
| Calendar quirk |
Largest names routinely beat the index in week 1 each month (pay-check flows). |
| Active-vs-Passive dynamics |
Passive dominance creates feedback loops that institutional investors may exploit. |
Your thoughts?
What knobs can we turn on P123 to hedge these risks and harvest alpha from these flow-driven anomalies?
1 Like
Well, the obvious answer is to buy short-dated ATM calls near the end of every month on the ten largest stocks in the S&P 500. (Put that on Monday's to-do list.)
You could download the transactions of the underlyings here: https://www.portfolio123.com/transact_real.jsp?portid=1864232 and see if that might work.
My back-of-the-envelope calculations show that you'd average 22% a month by doing this. Obviously there are quite a few months when you'd lose your entire investment. So you'd only want to do this with a small portion of your portfolio. (FWIW, the Kelly criterion says that the optimal investment is 31% of your portfolio. But that assumes the rest of your portfolio is in cash.)
2 Likes
A few more thoughts on the article. First, the authors claim to have shown that the return in the first week of the month is higher than in other weeks, which makes sense. But in the paper itself there's no empirical evidence to support this at all. Second, my own studies show that the bottom 10 or 50 stocks in the S&P 500 weighted by market cap outperform the top 10 or 50 stocks in the first week of each month and for other 10-day holding periods.
I think within the S&P500 most of it is noise. The most active, hyped stocks (PLTR, TSLA, NVDA) and many well-doing other stocks are not necessarily the TOP Megacaps. Recently, many of the biggest names like GOOG, AAPL did badly. So non-passive flows are still defining relative composition (even though maybe damped in comparison to the past)
What I find more interesting (yet I never heard anyone discussing this, even passive investing pessimists like Michael Green):
Let's assume passive really does NOT disturb pricing within the S&P500. It has a hard cut-off at the lower end regarding its constituents. A Russell1000 ETF is much less popular but has significant overlap with the S&P500. Russell 2000 ETFs are unpopular in passive and have no overlap. Some few US investors and basically EVERY exUS passive investor invests either in a QQQ or SPY ETF or the MSCi World (which is 70% US large)...
Conclusion: In sum, the S&P500 will receive continuous overweight inflows from global passive. Even if it's current "fair weight in a global LargeCap Portfolio is 70%, through the strong US LargeCap bias, it will receive maybe 80cts of each dollar invested...
This imo distorts the relative pricing between US Large and the rest (exUS and Small). There is evidence that just listing in the US already results in a premium for the same company.
And it's self reinforcing. QQQ and SPY behave smoother with higher returns and suck up all the global liquidity. This attractiveness brings new relative passive inflow (overseas investors ditching their MSCI World plan for QQQ) and also other flows like options and derivatives (liquidity needed). Overseas companies prefer to IPO in the US etc.
Thoughts?
PS: I would bet against passive flows but that also means in LargeCaps I would always prefer an ETF over own stock selection. Imo long-term you cannot win this game after fees.
For Small/Microcap selection this more blessing than curse, since the ever greater liquidity needs and flows battles at the big boy club could make mispricing in micro to mid more persistent
@yuvaltaylor: really appreciate you digging in and running your own tests; that kind of hands-on scrutiny is exactly what moves us forward. 
On the first point: the evidence is in the draft, it’s just hiding in Appendix E (p. 76). And, to be fair to the authors, the piece is still a working paper, so the headline tables haven’t all migrated to the main text yet.
Below is their Table E.8. The coefficients are excess returns for each large-cap portfolio during the first seven trading days of the month (that’s the authors’ definition of “week 1,” presumably to capture pay-period flows that can land anytime up to the 7th). Newey-West t-stats are in parentheses.
| Variables |
RTop10 |
RTop50 |
RTop100 |
RTop150 |
RTop200 |
| Top 10 × Month Start |
0.0282 (2.28) |
— |
— |
— |
— |
| Top 50 × Month Start |
— |
0.0220 (2.82) |
— |
— |
— |
| Top 100 × Month Start |
— |
— |
0.0140 (2.10) |
— |
— |
| Top 150 × Month Start |
— |
— |
— |
0.00647 (1.07) |
— |
| Top 200 × Month Start |
— |
— |
— |
— |
0.000915 (0.16) |
| Observations |
2,974,202 |
2,974,202 |
2,974,202 |
2,974,202 |
2,974,202 |
| Firm × Month FE |
Y |
Y |
Y |
Y |
Y |
| Month × Month Start FE |
Y |
Y |
Y |
Y |
Y |
| Adjusted R² |
0.00656 |
0.00656 |
0.00656 |
0.00656 |
0.00655 |
Table E.8 – Excess Returns on Large Stocks at Beginning of Month
Your second finding, bottom-cap stocks beating the top-cap names in the same window, is really intriguing. The current draft doesn’t test that at all. It would be great if the authors
- ran the same “Month Start” regression on the bottom 10/50 portfolios, or
- at least showed simple mean-return comparisons (bottom − top) for those 10-day windows.
That would be a neat robustness check, and maybe even an extension, before the paper is finalised.
Hope this helps, and kudos again for digging in and running your own tests!
Another intriguing detail is that the data they used dates from 1996 to 2020. Passive flows were relatively negligible during the earlier part of this sample. Passive flows have been massive since 2020, yet data from that period was deliberately excluded.
1 Like
@yuvaltaylor great spot! Surprising that the dataset cuts off just before the 2021-24 wave of passive inflows. Adding those quarters might reshape, or even upend, the results. No clue why the authors drew the line there, but calling it out keeps the discussion honest. 
Bottom line: academic papers are great for inspiration, but we shouldn’t take them at face value. We need to rerun the numbers ourselves before baking any findings into our models.
@Doney1000 good points. Here are four quick, data‑based checks to see if passive flows are bending the market:
- Inside‑S&P gap
Plot equal‑weight vs. cap‑weight S&P 500 returns since 2020. If the gap keeps growing, a few mega‑caps (NVDA, TSLA, etc.) are doing most of the lifting.
- Flow bias
Compare net ETF inflows into big‑cap funds (SPY + IVV) with flows into small‑cap funds (IWM + VTWO). Line that up with the Russell 1000 − 2000 return gap. If they move together, fresh money is chasing the biggest names.
- U.S. listing premium
Look at valuation multiples of U.S.‑listed ADRs versus their home‑market shares. A bigger U.S. premium after 2020 would show extra demand for a U.S. ticker.
- Small‑cap trade
Back‑test a long micro/small‑cap value, short cap‑weight S&P 500 strategy with realistic costs, splitting the run into pre‑ETF boom and post‑ETF boom. If the spread stays positive outside of crash rebounds, the distortion is real.
Bottom line: single‑stock “noise” inside the 500 might be limited, but passive flows can still tilt prices by size and country. The data is public, could be worth crunching to see how strong the tilt really is.
1 Like
Regarding point 1 and 4 I like to look at Yardeni's data as first naive check.
Shifting sector composition and their P/E's also have huge effect of course but even there cap-weighted construction of sector indices shows multiples going up and to the right for almost all industries...
Hard to tell if SMid is too cheap today or Large was too cheap in 2010-2018 (or both?) but something seems odd. Especially past 2020...