As my Bloomberg News colleagues Rachel Evans and Dani Burger reported last week, Vanguard wants to expand its footprint in Europe and Asia-Pacific. Roughly 94 percent of the company’s assets under management are now in the U.S.
Vanguard, however, won’t be peddling the market-cap index funds for which it’s famous in America. Instead, according to incoming Chief Investment Officer Greg Davis, it will “offer low-cost, high-quality, active products.” Davis forgot to mention old-fashioned. The mutual funds will feature real-life stock pickers as opposed to the robots that run the new generation of active funds known as smart beta.
The old countries apparently prefer the old methods. In Europe, traditional active management accounts for 86 percent of assets, compared with 65 percent in the U.S. And it isn’t cheap. European active funds charge an average of 1.02 percent in annual fees.
Vanguard intends to give overseas investors a break with its low-cost stock picking. In the process, it will also help answer two questions that bedevil active managers.
The first is whether there’s merit in active management. Vanguard has shown over the last four decades that most active managers lose to a low-cost market-cap index fund. But it’s not entirely clear how much of that failing is due to poor stock picking and how much can be attributed to high fees.
The second question is whether active management is better left to computers, which can pick stocks more cheaply and systematically than humans.
The answers will be years in the making, of course, but who wants to wait that long? I decided to search for clues in Morningstar’s mutual fund data. I looked at the alpha -- or outperformance relative to a benchmark, as calculated by Morningstar -- generated by the institutional shares of 4,604 mutual funds over the last three years through July, the longest period for which numbers are available.
I found that only 48 percent of those funds generated alpha during the period net of fees. But gross of fees, that number jumped to 88 percent.
Institutional shares typically pay lower fees than other mutual fund share classes. So I also looked at the C-shares of 2,951 mutual funds, which typically pay the highest fees. Only 21 percent of those funds generated alpha net of fees. But gross of fees, 96 percent managed to beat their benchmarks.
Those numbers support the theory behind Vanguard’s push into active management. Namely, the only thing wrong with active is high fees.
But who should do the stock picking, humans or bots?
For that, I compared the institutional shares of 135 actively managed U.S. large-cap growth mutual funds and 131 value funds with their smart beta counterparts. Specifically, I compared the gross returns of the growth funds over the last 10 years through July with the return of the MSCI USA Quality Index, and I compared the value funds with the MSCI USA Enhanced Value Index.
The bots were better. Only 27 percent of the growth funds beat the Quality Index during the period, and only 24 percent of the value funds beat the Enhanced Value Index.
Obviously, these are limited samples, so it wouldn’t be smart to declare ironclad conclusions. It’s also worth noting that Vanguard isn’t new to active management -- roughly a third of its assets are active, including some smart beta mutual funds and ETFs.
Still, Vanguard’s new commitment to active management overseas will be telling, and I think these numbers offer a peek into what Vanguard’s experience will show. My prediction: Stock picking is a worthy endeavor if done cheaply, and the picking is better left to the bots.
This column does not necessarily reflect the opinion of Bloomberg LP and its owners.
Nir Kaissar is a Bloomberg Gadfly columnist covering the markets. He is the founder of Unison Advisors, an asset management firm. He has worked as a lawyer at Sullivan & Cromwell and a consultant at Ernst & Young.