- Comments (19)
- Text Size: Down Up
moneysense.ca, 11/03/10
Active Share: Not an infallible metric
You have to pity the plight of the active investor as study after study demonstrates the merits of indexing for the average investor. But a recent column in the Globe & Mail by Dan Richards titled “Only the truly active fund managers lead the pack” provided many active investors an opportunity to taunt those in the passive camp. “So, what do you say to that, Mr. Index Investor?” captures the gist of many of the emails I received.
A confession first: I haven’t fully read the paper titled How active is your fund manager? A new measure that predicts performance by Martijn Cremers and Antti Petajisto of Yale University that Mr. Richards refers in his Globe article. Still, it quickly becomes pretty clear that measuring how much a fund manager deviates from the index is not an infallible metric.
First, it is necessary to define one of the metrics the Yale Professors devised to measure how much a fund deviates from the index. They call it “Active Share”, which is simply a measure of how the weightings of stocks in a mutual fund portfolio deviates from the benchmark index. Long-only mutual funds will have an active share between 0 and 100. Index mutual funds whose holdings mirror the weightings in the index have an active share of 0. A mutual fund whose holdings have no overlap with the index have an active share of 100. The authors find that mutual funds with the highest active share outperform their benchmarks by 1.49-1.59% per year after fees and transaction costs.
Unfortunately for proponents of active investing, Active Share does not provide an infallible way to pick mutual funds. To understand why, let’s do a thought experiment. Let’s assume that we have an equal-weighted index composed of just 10 stocks. Let’s further assume that 10 actively-managed mutual funds are available and each fund holds one of the 10 stocks in the index. Each fund has an active share of 90 (just 10% overlap with the index). But it is clear that they can’t all beat the index. In fact, as a group, their performance will trail the index by their average fees.
PS: I have some exciting news to share with you about the new direction the blog will be taking shortly. Later in the day, you’ll notice some changes on the blog (those of you who are reading this through RSS or email will have to click through to find out). Sorry to be mysterious and all but we are working feverishly putting on the finishing touches and we are not quite ready just yet.
moneysense.ca, 11/03/10
Dude, MoneySense barfed all over your site. I sincerely hope I’m reading this before the finishing touches are applied.
Wow. I hope you’ll be making some good coin from this switch.
Congratulations?
I’m with Tom, I liked the ghetto look.
@Tom: Can you let me know exactly what you don’t like?
@Phil: The terms of the deal are confidential. I added some more details in this post but the bottomline is that not much has changed around here (apart from the look and feel):
http://www.canadiancapitalist.com/a-new-direction/
@Matt: Ouch. Ghetto look? I’m so offended
What am I missing? You wrote:
“But it is clear that they can’t all beat the index. In fact, as a group, their performance will trail the index by their average fees.”
If all of the index stocks were flat and the other 90% of non index stocks held by each of the actiuve funds all went thru the roof, why wouldn’t all of the active funds beat the idex (which would be flat)?
Kim
It’s all good, CC!
Hi CC,
I agree Moneysense just barfed over your site
what was a lovely uncluttered site with info and feedback now loses all that and the eye is sore looking at all the moneysense content
I suggest you set up another site like the last one where honest discussion takes place From being Socrates you have gone to Hugh Heffner. And like Heffner I am sure Moneysense must pay well to buy your followers. I dont think either will be a winner in this arrangement
I do hope Money sense will not be editing your content so it is suitable from their perspective since this is the impression the site now gives – a mouthpiece of Moneysense
Rajeev
[...] Capitalist discusses Active Share as it pertains to managed investment funds. I also discussed it a while ago for more colour if you really want your jaw to [...]
[...] Canadian Capitalist gave his opinion on recent column from Globe & Mail titled “Only the truly active fund managers lead the pack”. Is actively managed funds really better than passively managed ones? [...]
You never answered Kim’s post regarding the faulty reasoning in your “thought experiment”. I fear that your blind religuous faith in what Bogle has sold to you is clouding your analytical abilities. There is no need to pity us active investors; we are content in our superior index-beating results.
@Kim: Someone has own the stocks in an index. In my example, each of the ten funds owns 1 stock from the index. The other 9 stocks are owned by the other 9 funds. If the other 90% of the stocks go through the roof, it will be reflected in the index performance. The index returns will be through the roof as well.
@Dale: Thanks for the reminder to reply Kim’s post. There is nothing faulty in my thought experiment. In any case, I don’t know why you are jumping up and down. Do you use Active Share to narrow your fund picks? If not, how is this relevant in any way, shape or form to what you do?
@CC I misinterpreted your example. I was assuming that the one index stock made up only 10% of the holdings of each fund and that the other 90% was made up of non index stocks (i.e. that’s how you come up with an Active Share of 90). If each of your funds only owns one index stock and nothing else, then I’m not sure how your thought experiment is all that relevant to the Active Share concept.
@Kim: The example in this post is vastly simplified. Studies have shown that mutual fund portfolios in aggregate resemble their index. What I’m saying is that even if all the mutual funds have a high active share (for example, if there are just 10 funds invested in the S&P 500 and each of the fund holds 10 of the 500 stocks in the index, assuming equal weight, the active share for the funds will be 90), but the majority of them will under perform the index because of their fees. In fact, the paper itself points out that mutual funds used to have high active shares but as a group they trailed the index in the past. It’s not clear to me why the data for 1990-2003 supports a different conclusion and if high Active Share funds do generate alpha, where it is coming from (because if one group is winning, it has to at the expense of someone else. The question is: who are the losers?).
@CC Your “thought experiment”, now clarified to deal with Kim’s valid point, isn’t as inane as I first assumed. The “losers” are the investors in closet-index funds who are stupidly paying a 2+% MER for a fund which isn’t “actively” manged at all. I was “jumping up and down” because now that you are part of Canada’s “finest personal finance mag” (or whatever you called it) your humble readers will be holding you to an even higher standard.
P.S. I’m starting my own blog, to be called “The Other Canadian Capitalist”.
@Dale: Have you read the Active Share paper? If you have, how do you infer that the “losers” are the closet-indexers? The paper does support the notion that a fund that pretends to be active but hugs the index trails the benchmarks. Of course it will. If I have an index fund and charge investors 2%, it shouldn’t be surprising if my performance trails the index by 2%. Investors in the fund are losing to fees, not to other active investors. If truly active investors are able to earn alpha, they should be getting it from somewhere. My question is where? And what makes you think this excellent state of affairs will last (after all, the 14 year period of the study isn’t really long-term by any stretch of the imagination)?
CC In your thought experiment and the later musings, it seems that you have equated “average performance” with “index performance”. They may often be similiar, but they are not identical. Your whole notion of “alpha” as a useful yardstick seems to be result from your equation as well. It seems to me that over-perfoming an index can result from myriad factors including: holdings which differ from a chosen index, weightings, MERs, cash, turnover frequency, and the constituents (and weightings) of the index arbitrarily chose for comparitive purposes. Perhaps the biggest factor is “Omega”, which will be explained in greater detail in the newest blog on the block, “The Other Canadian Capitalist”. Have a good weekend. PS……… I don’t like the light blue sub-headings in your new format………..
PS 2…….MoneySense hasn’t been around sufficiently long for it to be hailed as Canda’s best PF magazine—(although it is very good); the title goes to Cdn Money Saver.
@Dale: I’m perfectly aware of the sources of alpha, thank you very much. But you seem to have forgotten that the discussion in this post is on Active Share, specifically the findings by Cremers and Petajisto that stock pickers are able to generate alpha net of expenses. The funds that make factor bets have not generated positive alpha, so it is moot to discuss them here:
“Economically, these results suggest that the most active diversifi ed stock pickers and concentrated stock pickers have enough skill to generate alphas that remain positive even after fees and transaction costs. In contrast, funds focusing on factor bets seem to have zero to negative skill, which leads to particularly bad performance after fees. Hence, it appears that there are some mispricings in individual stocks that active managers can exploit, but broader factor portfolios are either too efficiently priced to allow any alphas or too difficult for the managers to predict. Closet indexers, unsurprisingly, exhibit zero skill but underperform because of their expenses.”
@ You know perfectly well who the “losers” are in a pure mathematical sense; so do I. I was simply making the point, that in a broader sense, closet indexers are the “losers” because they probably believe they are in an “active” fund when they are not, and hence they are stupidly paying a 2.7% MER, when they should be paying .17%. My other rant stems from my observation that index-zealots are, to my way of thinking, obsessed with the “wrong” tools for measuring their performance. “Alpha” is a slide-rule, and it has it’s place, but index-zealots should put down their slide-rules and use their telescopes more often.
I am now prepared to let you have the last word.