Wednesday, September 2, 2015

How Much Diversification Do You Need?

Some investment experts advocate maximum diversification, which others deride it as “di-worse-ification.” In Ben Carlson’s recent article, he is somewhere in the middle saying you need to find the “right balance between eliminating unsystematic risk (risk that’s specific to single securities or industries) and di-worsification by adding too many overlapping funds.” Who is right? Your answer depends on your views on active investing.

At one extreme, suppose you knew for certain you’ve identified the one stock that will go up most in the next year. You’re not 90% sure or 99%. You’re 100% sure. Then you’d be crazy not to invest everything you have in that one stock. Of course, you’d also have to be crazy to be this certain about the stock.

As our crystal balls become cloudier, the need to diversify arises. Maybe you decide to put some of your money into other stocks, even though you have less confidence in these other stocks. You’ve decided that the protection against possibly being wrong about your top pick is worth the risk that your lesser picks will make less money.

Those who argue that you’re di-worse-ifying are saying that you’re diluting the potential returns from your top pick with lesser ideas. How far it makes sense to go with diversifying depends on how confident you are that your top stock picks will work out well.

At the low end of the confidence continuum, we have those who have no idea which stocks will perform well, have no idea if markets are going to go up or down, have no idea which money manager will perform well, and have no idea which financial advisor might beat the markets for us. These people are indexers.

Diversification is simple for indexers like me. We own all stocks for as low a cost as possible. There is no such thing as di-worse-ification because we have no opinions about one stock being better than others. There is no reason to fret over active mutual funds because index funds are cheaper and cover the same asset classes.

Another thing to consider when it comes to diversification is that all evidence points to very few people having stock-picking skill. Even among money managers you might choose to manage your money, very few have enough stock-picking skill to make up for their fees and expenses. And very few individual investors have the skill to identify these winning money managers.

So, almost everyone who thinks they can beat the market either with their own picks or by finding the right helper is wrong. Not everyone, but almost everyone. I had the stock-picking bug for about 12 years, but a careful study of my results beat that nonsense out of me. I’m a happy, diversified indexer now.


  1. I don't like the definition of diversification where we're talking about different funds that basically amount to being invested in equities. Full exposure to equities, no matter how it's shaken up through different funds, isn't proper diversification.

    Proper diversification is across asset classes, bonds, real estate, equities and cash.

    And the basis for diversification should be based on asset classes that are uncorrelated or negatively correlated - so that when one asset class crashes, the others either are unaffected, or go up instead of down.

    And the reason for all that is so that you can rebalance across asset classes to take advantage of those downturns.


    1. Good point. Diversification across asset classes would be in addition to Michael's discussion about diversification within an asset class.

      Michael is proposing to diversify using an index. He will never outperform the index and will only underperform the index due to his transaction costs and management expense costs.

      By deviating from the index, one can either outperform or underperform. Having fewer stocks than the index is the only sure way to deviate. Selection of the stocks then becomes important as Michael noted; and it's not his game.

      It all comes down to your own comfort level. Berkshire Hathaway's chairman Warren Buffett says:

      "If you can identify six wonderful businesses, that is all the diversification you need. And you will make a lot of money. And I can guarantee that going into the seventh one instead of putting more money into your first one is going to be terrible mistake. Very few people have gotten rich on their seventh best idea."

    2. @Glenn: When it comes to different asset classes, the argument is different than it is within an asset class. I have no idea which stocks are expected to outperform others, but I believe that over time stocks should outperform real estate, bonds, and cash. So, we diversify within an asset class to reduce volatility for the same expected return, but we diversify across asset classes to reduce volatility at the expense of lower expected returns. It's possible for reasonable (not overconfident) people to differ on their allocations to different asset classes. But it is mostly overconfidence that drives people to make picks within asset classes.

    3. @Anonymous: Given Buffett's stock-picking record, adding a 7th stock really is di-worse-ification. But for me, even adding the 5000th stock is just helping a tiny bit more.

  2. The original quote that you mentioned talks about "overlapping funds", which is a great example of di-worsification. Like the classic mistake "I bought 5 Canadian equity funds so I'm diversified now", or a dividend investor who decides to invest in all 5 big Canadian banks to spread out their risk. Of course that's not the true definition of diversification but many people believe it does fit.

    1. @Richard: I come to the same conclusion, but for a different angle. I have no idea which Canadian fund manager might outperform. So, buying multiple such funds really does provide some useful diversification. The part that makes no sense is paying the fees for even one such fund. I choose to replace them all with a cheap index fund.

    2. Michael, put another way, buying multiple actively managed funds does provide diversification, but only of active manager risk. But why take any active manager risk, just buy an index fund and save the fees.

  3. My thinking has matured on this....and I agree... as I get older I realize my crystal ball is very cloudy. Hence, the only way I can be successful is to diversify more and reduce risk in doing so. Investing in more individual stocks might not be best, hence, I'm indexing more with products like VTI and VXUS.

    I am hopeful this is the right thing to do (because I cannot predict the future).

    1. @Mark: It took me a while to admit this to myself. I hope things work out for you.

  4. As above, true diversification lies across asset classes, not simply stretched through different stock markets. (How do you define "the market" and if you think you are diversified, do you measure against a benchmark?)

    If you have all your money in public equities and don't own truly ALL of them, then you are doubly concentrated and stock-picking.

    It's silly to pretend we are passive diversifiers when 99.99% of us are active concentrated asset pickers.

    1. @SST: I see active/passive as a continuum rather than binary. There are those who say that to be passive and diversified, you need to own every asset class on earth in proportion to their capitalizations. This is one extreme. I think it's reasonable to have different return and volatility expectations for stocks, bonds, real estate, cash, commodities, etc. Once you believe these asset classes have different expectations, it becomes reasonable to not own some of them, or at least make you own selection of allocation.

      But you're right that most investors take it much further than this. They engage in market timing and individual security selection. I guess someone has to participate in price discovery.

    2. re: "I believe that over time stocks should outperform real estate, bonds, and cash."

      But there is no way for you to be certain, and it's a dangerous mindset for an investor to exclude any asset (i.e. concentrate on a single asset).

      The 30 year stretch from 1980-2010 bonds gave out a higher return (w/ lower volatility & risk) than public equities. Thirty years is a good chunk of your investment lifetime to ignore something like that. Over the last 45 years there were two full decades (almost half the time!) where gold trounced stock returns. With Canadian real estate rising 5%/year over the last 30 years, tag on a 5% net rental profit, and a landlord/RE investor handily beats stocks.
      (Cash is more of a mechanism rather than a true investment, so everything beats it!)

      I'm not saying all these trends will be repeated, but if an investor thinks that stocks will always be the best investment, "over time", they will be (and have been) wrong -- a lot.

      re: "But why take any active manager risk, just buy an index fund and save the fees."

      Don't kid yourself, an index fund places you under two levels of active management risk -- index and fund (but pay for the fund only).

    3. @SST: It's not about whether stocks will eventually prove to be better over 30 years. It's about a priori expectations. Choosing stocks over a mix of assets amounts to choosing higher expected returns and higher volatility. By making this choice, I'm content to take the small chance that stocks will underperform other assets for a shot at the likelihood that they will outperform.

      It's always possible to cherry pick start and end dates to make one asset class or another look good or bad. Your contention that when "an investor thinks that stocks will always be the best investment, 'over time', they will be (and have been) wrong -- a lot" depends greatly on the definition of "a lot". For 30-year periods, stocks losing to anything else is rare, and the margin is small. For the period from the end of 1980 to the end of 2010, Canadian bonds beat Canadian stocks by 0.8%/year.

      It's misleading to refer to risks associated with index funds as "active management risk." Index fund managers could prove to be incompetent at following an index or they could be corrupt, but this is very different from being poor stock-pickers.

    4. I'll side with Cullen Roche's summary of diversification: "Stocks are an important, but relatively minor slice of an aggregate portfolio."