Diversification or Di-worse-ification?

Hi Bylo-

The problem with this approach is that you are likely to get index performance (in gross terms) but your net return will be lower because your weighted average MER will be higher. In another thread thbox cited a study that showed even combining just two funds dramatically increased the probability of index performance (maybe thbox can repost that info, because I can't remember where it is).

If you go for this strategy, I think you want to make sure you don't buy funds with competing styles. A value manager and a growth manager will likely cancel each other out.

A valid approach might be to buy one index fund and one other actively managed fund that has a manager and/or style you are strategically comfortable with. This way, if the active manager wins, your returns will be somewhat in excess of the index. On the other hand, if you lose with your choosen active fund, your results will not be completely compromised relative to the index.

One other thing Bylo, it occurs to me you could implement the above without active management if you like. For example, you could buy the SP500 with SPDRS or an index and then overweight your large cap US investments with a value or growth index fund. I'm actually thinking of doing this myself and would probably buy Vanagaurd's Value Index via Waterhouse.

Bylo, I acknowledge that you know much more about these things than me, so I cannot suggest a solution to the problem you pose. Let me instead counter with another. Last year I thought I did my homework. Altamira Equity and the Trimark Canadian Funds were highly rated by most of the people who make money writing about them. Moreover, they represented two different and complementary investing styles according to one author. The one (Altamira) sought out the sector likely to outperform, and overweighted in this. The other (Trimark) sought out individual companies which were likely to do well. The one concentrated on smaller companies and traded often. The other took large positions is bluer chips and held them. How did my wise strategy end up. You said it yourself: Witness the recent underperformance of such former high-flyers as Altamira Equity or...Trimark Canadian. So where did I go wrong in my attempts to diversify?

PK,

FWIW, in typically Canadian fashion I've started to implement a part-index/part-active compromise solution. My US portfolio at Waterhouse now stands at about ¼ Vanguard Index 500 Value (VIVAX - tracks S&P 500/BARRA Value index), ¼ BRK.b (tracks the Oracle "index") and ½ Vanguard Total Market Index (VTSMX - tracks Wilshire 5000). (Of course I've also owned PH&N US Equity for many years. It has performed about the same as the S&P 500, but if nothing else the tax implications are too painful to consider a switch.)

rge,

Thanks for the kind words. I don't think you necessarily went wrong. Remember mutual funds are long term investments. Would you believe that if you'd bought Altamira Equity 10 years ago then -- despite the recent poor performance -- your original investment would have grown by over 20% CAR? Not too shabby. Frankly (sorry!) I don't know what to make of Mersch's departure and the new "team." Notwithstanding Mersch's problems with the OSC, IMHO the fund outgrew Mersch's sector rotation, small-cap oriented, etc. style and/or his ability to adapt. As for Trimark, I'm in the camp that can't believe Krembil has suddenly lost his marbles. [you owe me, thbox ;-)]

Greetings, fellow travellers:

PK:

The study, done in the US in 1995 by Prudential Diversified Investment Strategies found that - on average - combinations of even two growth funds had a .97 correlation with the index.

rge:

The whole point of diversifying by investment strategy ( a dumb idea if you ask me) is that they cancel each other out so that a bad year by one style is cancelled out by the good? year of the other. Over the long term, each style will have ups and downs, so - who cares which style you use. (Unless, like PK, you've read Fama and French, and believe that value is fundamentally better than growth.)

BTW rge; FWIW, the ONLY domestic fund I own is Trimark Cdn (the DSC version no less) and I'm not going no place - leastways until Krembil hangs up his spurs.

Bylo:

debt recorded.

:-)

Clearly enough there are at least two different issues at the forefront of this discussion -- maximizing return and minimizing risk.

The principal argument for holding "complementary funds," it seems to me, is reducing volatility, rather than maximizing returns. That reduced volatility may well have the consequence of an investor's underperforming the index -- but so what, if reducing short-term risk is the objective?

Secondly, the argument for indexing vs. active management seems to me much stronger considered relative to the more diversified U.S. market than in the Canadian context. I would guess that likewise a Prudential-type study of the correlation of fund combinations to the index might produce less conclusive results in Canada.

At any rate, I think there's a strong argument for agnosticism in discussions of this sort. For my own part, risk management is a main consideration. So far I've resisted the temptation to go the index route in favor of "deworsification."

(thbox, this is off-topic, and worse yet, none of my business, but I get the occasional vibe of left-wing political associations somewhere in your past -- or perhaps just an academic interest in that respect. Or am I completely out to lunch here?)

Good educational question. Not sure I can add much to what Bylo, PK, rge, Ray, and thbox have said, but I suppose that is no reason not to try.

It seems we all [at least myself for sure] may have some degree of self-contradiction or irrationality in our make-up. We know our statistics – yet we still try to beat it. We want the safety of diversity but hope for the better return at the same time. We may accept the random walk model but still think there are “better funds”, etc we can identify.

I would agree that owning “two good” similar funds [in the same “market”] will only provide a smoother and quicker reversion to the mean. And it is the same mean that either of the two funds would tend to by itself anyway – in due course.

My take on the Altamira and Trimark example used would be that here is another statistical example of how top-quartile funds at one time will be bottom-quartile funds at another.time. Or just because both funds flipped three heads [good returns] in a row in the past, there is no guarantee that both cannot flip tails [poor returns] together now.

So, I would argue that “lower cost” is the only logical criteria for picking between two similar alternatives. And it may as well be the boring index – because it protects you from the down-side fluctation of a “bad year” – relative to the index. [of course the index itself can have a bad year]

On this issue – it seems to me that in Canada, since we do not have a real low cost index fund [like Vanguard], – TIPs makes far more sense than the equivalent but higher-priced Canadian Equity index fund. In fact I don’t see the rationale for Canadian equity index funds with a MER near 1% at all. Why would anybody buy them?

PK – on the issue of SPDRS, I was told that there is some technical [and obscure] reason that strictly speaking they cannot be part of the foreign content in a SDRRSP. They don’t qualify – but I don’t know why, something about being a special kind of trust perhaps(??). Do you know anything about this?

Having said all this I could still give some reasons for diversification:

(1) Just for-the-hell-of-it because life is more interesting to have “bet” on a “good horse” [and watch it fail to win like it did yesterday].

(2) If the two funds cross-correlate, they will of course provide a smoother ride, i.e. smoother fluctuations (risk), about the mean.

(3) If one of the funds is in a different market – say with a higher mean return (and so higher volitility-risk), like the US market, or European market or the value market (a la Fama and French).

In my own case I have "diversified" in the sense that our investments are split between two extreme camps. In the one camp are the equivalent of both the TSE and S&P index investments. In the other camp are two high-tech stocks.

Ray W:

After the revolution............

:-)

I’m till reading Frank Arsmtrong’s “Investment Strategies for the 21st Century”. In Chapter 22 he writes:

Subscribe to the – “Journal of Finance”

Does anybody out there subscribe? Is it a “good investment” of time and money?

I hope my note is not compromising the thread topic. Seems to me this journal could be relevant.

Hi George - Re SPDRS as foreign content in a SDRRSP, I don't know. I own mine outside an RRSP via the Canadian Shareowners Association, who until recently didn't even offer RRSPs. It has never occured to me that they couldn't be in a SDRRSP--but then I've never needed to find out. Maybe someone else knows the answer to your question for sure -- if not, I'll send an e-mail to CSA and see what they say.

I've always had problems with the argument that buying a bunch of funds will give index like performance at a higher MER so you should only buy one fund. This view is caused, I believe, by emphasizing the best outcomes when examining the first fund and the worst outcomes on the second fund when you should be weighting all outcomes by their probability of occurring. If you think that buying two (or more) funds will give you index like performance, then you believe that on average, funds do no better than the index so your one fund bet is just a gamble that your fund will turn out better than what you expect will occur on average. You are taking extra risk for no additional expected gain. A better response, if this is what you believe, is to buy an index fund. Suppose you think that, with careful selection, managed funds can beat the index by say, 1%, on average. Then you should buy as many good funds as practical. Your goal should be to get index like performance plus an additional 1 percent per year. You will reduce the idiosyncratic risk from the individual funds, and capture the additional returns that you believe occurs on average. You want your portfolio return to closely correlate with the index but at a one or two percent higher level. Less risk, and still getting the extra return. I guess I'm saying that you should carefully examine your expectations for your funds. Do you really believe that your first fund is so good that you definitely want to buy it but your next best fund doesn't come close to making the cut? Or are you focusing only on the good outcomes for the first fund and ignoring the bad. If you decide that your fund can not beat the index on the average, then buying index funds will reduce risk and provide the same expected returns.

Resent because the previous posting didn't have paragraph breaks:

I've always had problems with the argument that buying a bunch of funds will give index like performance at a higher MER so you should only buy one fund. This view is caused, I believe, by emphasizing the best outcomes when examining the first fund and the worst outcomes on the second fund when you should be weighting all outcomes by their probability of occurring.

If you think that buying two (or more) funds will give you index like performance, then you believe that on average, funds do no better than the index so your one fund bet is just a gamble that your fund will turn out better than what you expect will occur on average. You are taking extra risk for no additional expected gain. A better response if this is what you believe, is to buy an index fund.

Suppose you think that, with careful selection, managed funds can beat the index by say, 1%, on average. Then you should buy as many good funds as practical. Your goal should be to get index like performance plus an additional 1 percent per year. You will reduce the idiosyncratic risk from the individual funds, and capture the additional returns that you believe occurs on average. You want your portfolio return to closely correlate with the index but at a one or two percent higher level. Less risk, and still getting the extra return.

I guess I'm saying that you should carefully examine your expectations for your funds. Do you really believe that your first fund is so good that you definitely want to buy it but your next best fund doesn't come close to making the cut? Or are you focusing only on the good outcomes for the first fund and ignoring the bad. If you decide that your fund can not beat the index on the average, then buying index funds will reduce risk and provide the same expected returns.

thbox, I herewith acknowledge my "dumbneess." I guess I should have read your book rather than Duff Young's. BTW, have you informed him that his recommendations are also "dumb," or is there a professional code of closing the ranks among financial pundits? ;<(

rge:

I certainly didn't mean to imply that I thought you were dumb, and if offence was taken, I apologize. On re-reading your posting, I couldn't find anything to suggest you subscribed to the idea of "style diversification" and the remark was intended as a repsonse to PK's posting. Again, sorry if I offended you.

On the issue of "professional courtesy", I haven't had the opportunity to discuss the concept with Duff, but I think its clear from my post here (and elsewhere) that I think the appeal of the concept of style diversification has more to do with its marketing potential than its investment soundness.

:-)

Another minor reason for diversifying among a couple of "equivalent funds" is the ability to make some adjustment in asset allocation through switches within one or more fund families in a SDRRSP rather than pay excessive transaction fees to a discount brokerage. So, for example, you might hold PH&N Bond and Canadian Equity as well as Bissett Bond and Canadian Equity rather than PH&N Bond and Bissett Canadian Equity on their own and have to cough up $40 to Action Direct if you wanted to implement a shift from fixed income to equity or vice versa. Of course, with E-Trade I suppose you could avoid this and simply pick the mutual funds you wanted without concern about switches. But that would mean another dreadful round of T2033s...