It is a truth universally acknowledged that a passive investor in possession of a quote from William Sharpe must be in want of a megaphone.
One of Sharpe’s most popular lines comes from his 1991 essay ‘The Arithmetic of Active Management,’ in which he claimed that, ‘Properly measured, the average actively managed dollar must underperform the average passively managed dollar, net of costs.’
‘Over any specified time period, the market return will be a weighted average of the returns on the securities within the market, using beginning market values as weights,’ the Nobel laureate continued. ‘Each passive manager will obtain precisely the market return, before costs. From this, it follows (as the night from the day) that the return on the average actively managed dollar must equal the market return. Why? Because the market return must equal a weighted average of the returns on the passive and active segments of the market.’
The essence of this is that active managers, in aggregate and whether operating within a mutual fund structure or not, own the whole market. The passive investor earns the return from the whole of the market, which is to say the average of the active managers. Passive funds tend to have lower costs than active funds, so indexers outperform the average manager on a net basis.
But does that model reflect reality? New research by Sanjeev Bhojraj of Cornell University and Young Jun Cho of Singapore Management University examines whether active and passive investors really do own the entire market in a meaningful way.
Passive is the new active
Narrowing down their dataset for simplicity, Bhojraj and Cho took the Center for Research in Security Prices’ universe of more than 3,500 domestic US stocks as their benchmark and compared it with the aggregate holdings of passive domestic equity funds. They employed the active share technique, which calculates what proportion of a portfolio’s holdings differs from the benchmark’s holdings.
Active share is more commonly used to measure how closely active managers are hugging their indices, but in this case it also revealed the active biases in nominally passive vehicles. An S&P 500 ETF, for example, will have a high active share relative to the broader US stock market because it only owns large caps, meaning that investors in such funds are not really making a passive allocation to US equities as a whole.
Unsurprisingly, then, Bhojraj and Cho discovered that passive funds had a remarkably high active share in the early days of index investing. The figure stood at 45% in 1990, rising to a peak of 47% two years later, before dwindling to its current value of 25%.
‘The 1990s were characterized by passive funds that were heavily weighted toward large-cap stocks,’ the authors observed. ‘This means that, in aggregate, the performance of passive funds was heavily influenced by the performance of large-cap stocks. The drop in the active share of passive funds suggests an increasing similarity in portfolio holdings between the stocks held by passive funds in aggregate and the broad universe of stocks. This is consistent with more money flowing into passive small- and mid-cap funds, making it more difficult for active funds to attract investors in this space.’
Interestingly, the active share of active managers in aggregate is only a little higher than the figure for passive funds, at 31% today compared with 44% in 1990.
The primary reason why the active share for passive investors isn’t closer to zero is that ETFs systematically underweight small caps. As well as favoring small caps, active managers have a substantially stronger bias to momentum names and recent IPOs, plus a more modest preference for growth stocks.
This in turn has implications for the performance debate between active and passive investors. Looking at their full sample period of 1992 to 2017, Bhojraj and Cho found that passive vehicles on average outperformed active funds by about five basis points (bps) per month on an equal-weighted basis and by 10 bps per month in asset-weighted terms.
But this was a tale of two periods. Between 1992 and 1998, passive beat active by a massive 24 bps per month. However, from 1999 to 2017, the difference in performance was not statistically significant. Furthermore, when they controlled for factors, the academics determined that there was no difference between active and passive performance, even after fees.
‘Our evidence suggests that passive and active funds in aggregate hold positions that are different from each other as well as when compared with the universe of stocks,’ Bhojraj and Cho concluded. ‘We also find that the holdings differ between the groups along systematic dimensions. This raises the possibility that returns across the two groups could vary at any point in time. For example, active funds’ greater exposure to small/mid-cap and momentum stocks – as well as their smaller exposure to value stocks – suggests that active funds are likely to outperform during the periods when small, value and momentum stocks perform well.’