How to think about the construction of a portfolio? I’m just talking about a portfolio of stocks, not a full asset allocation. There is plenty written about the latter; and comparatively not much on the former.
The question implies other questions, too: How many stocks? How to think about risk? How to think about diversification? And on it goes.
I get these questions often. And I have my own opinions about such things. But I decided to do another search through academic articles to see what’s out there. I’ve often said that while a wealth of material exists about how to value stocks and how to pick them, there is little about the actual continuous running of a portfolio.
In my search, I found my impression holds true. You can find mathematical treatments, but they seem impractical. Then I stumbled on a piece from 1969 by Frank E. Block titled “The Elements of Portfolio Construction,” published in the Financial Analysts Journal.
I was sucked right in by Block’s opening lines. He writes that building and running a portfolio are “among the weakest sectors of the literature of investing.” Block says that while there has been lots written by mathematically oriented scholars, “the problem is that a general theory of continuous portfolio management has not yet evolved.”
Confirmation bias? Probably! Anyway, I continued to read…
I know many revolutions around the sun separate us from 1969. But what I like about Block’s article is his treatment of the issues seems timeless. He doesn’t arrive at any magical answers. Nonetheless, I like the thought process and you may find it useful to go through some of the “elements of portfolio construction.” So in this post, I’m going to share my reflections on certain key insights from Block.
Block himself was a CFA and a VP of a bank. He was a practitioner – but with one foot in the academic world as he was also the assistant editor of the Financial Analysts Journal. So he comes at the subject with a good perspective from both camps.
In fact, he begins by noting some of the wide gaps between how investors think and how academics think.
How to Think About Risk
For example, take the idea of risk. For academics, “risk is measured by variability of rate of return, which they express in such familiar statistical terms as standard deviation, variance, semivariance, and so on.” This hasn’t changed since 1969!
For me, and I suspect most practitioners, risk is really about earning a poor return. And so when I’m trying to gauge risk I tend to focus on things that could impact long-term returns. I think about the financial leverage in the business. (One of Charlie Munger’s three L’s, which represent ways to go broke, is leverage). I think about its ability to generate cash. I think about the durability of the business itself - what kind of competition does it face? These are the kinds of things I focus on, not how much the price has bounced around.
This gap with academics here seems unbridgeable to me. And so a lot of the academic literature that does exist on portfolio construction is a no-go because it starts with this fundamental assumption that volatility is risk. If that’s how you begin, you're bound to offer advice that is a non-starter for most investors.
Block says as much:
“Occasionally, portfolio managers are startled to read an academic paper which seems to say that a particular mechanism may be used to accomplish something the portfolio manager is striving to avoid.”
Diversification is one area where this really comes out. Academics love diversification - but of a peculiar kind. Block gives us an example from William Sharpe, the inventor of the Sharpe ratio and one of the giants of academic finance. Sharpe wrote that "diversification enables the investor to escape all but the risk resulting from swings in economic activity – this type of risk remains even in efficient combinations."
But most portfolio managers would not consider that the main purpose of diversification. As Block says, an investor “does not wish to escape the unique characteristics of his portfolio holdings; he wants to participate in that uniqueness -- that is why he purchased those particular securities.”
Diversification for me is more about trying to mitigate unknowable events taking me out of the game. I hold ten stocks. There is no magic to the number ten; something in the 10-12 range seems ideal to me for a variety of reasons I've written about before. (I'm not dogmatic on this point; different numbers for different folks). My largest position is around 12%, but the stock went up to get there. I don’t like to push anything past 10% on cost and frequently I stop before that.
This strategy is not for everyone. You have to be okay with the (likely) higher volatility. You also have to be careful about your selections. With a ten-stock portfolio, there are many stocks I never even consider. Leveraged securities are a no-go. (“I learned this very early,” Peter Lynch once said, “It is very hard to go bankrupt if you don't have any debt.”) Companies that don't generate free cash flow get a quick boot. Ditto turnarounds. (“Turnarounds seldom turn," as Buffett wrote back in 1979). The list goes on.
How to Think About Diversification
So how many stocks should you own?
Block uses a basic probability exercise to get a feel for the number of stocks one should have in a portfolio.
First, he makes some assumptions, among them:
An investor picks three winners out of every ten.
“Winner” here means it performs better than the market average.
Of the remaining seven, six perform more or less in line with the market.
And one is a loser.
Further, one winner offsets one loser.
If success means beating, or at least matching, the market, how many stocks should you hold?
Block presents the following table:
Intuitively, that feels right. What might be surprising is how only 15 stocks give you a 90% chance of success. Of course, you could tweak his assumptions and you could get lower numbers, but the gist of it holds: You don’t need as many stocks as you may think to get the job done. (The above table makes a general point; don’t get too wrapped up in the precision of it just because there are numbers).
But there is more. I didn’t say over what time interval we measure such returns. Let’s say it’s one year, or even five years. Either way, “the likelihood of [the 15-stock portfolio] being unsuccessful in two consecutive time intervals is only 1%, or a 99% success ratio.”
Thus, the key takeaway, which Block emphasizes with italics: “The amount of diversification needed is a function of the time horizon.”
Put another way: The longer the time horizon you have, the less need you have for diversification.
What Have You Done For Me This Morning?
I suspect this is one reason why mutual funds frequently own so many stocks. Since mutual funds are frequently priced, they are frequently evaluated and few managers want to risk being out-of-step for long.
Block addresses this, too:
"A particular problem of modern day investing is the pressure for short term performance. Many accounts are priced monthly or quarterly and reviewed by the customer or a committee to see how they have performed in such brief periods of time. Mutual funds are priced daily, and mutual fund managers almost face the question: 'What have you done for me this morning?'"
An old problem. And hard to fix unless you have the right set up and good partners. Individual investors have no such pressures – to their great advantage!
I aim to own my stocks forever. Not literally forever, but I abide by Buffett’s dictum: “If you aren't thinking about owning a stock for ten years, don't even think about owning it for ten minutes.” It’s one of the best filters I use.
Academic finance tends to think differently. Block mentions Harry Markowitz and his idea of an “efficient portfolio” that maximizes the risk-return combination. Markowitz may be a dated example, but Block makes a good point: any idea of “efficiency” implies a lot of trading:
“It is not hard to imagine the designing and construction of an ‘efficient portfolio’ at a given point in time. Yet, the moment that either prices or expectations change, one no longer has a perfect portfolio. Purchases and sales would have to be made continuously… The manager would find that its time horizon had collapsed to virtually zero and that its turnover was nearly infinite.”
Obviously, you could rebalance every quarter or year, but such strategies hold no appeal to me. I have nightmares about the taxes. But aside from that, I disagree with Block that an “efficient portfolio” is “not hard to imagine.” The assumption here is that you can accurately measure risk and also reliably estimate potential return. I don’t think so. I find the idea of some optimized portfolio to be pure fantasy. Real life is messy.
Also, mathematically driven models can recommend you do seemingly nutty things. Block notes there is no constraint inherent in Markowitz’s efficient portfolio, so, for example, it might dictate you put 90% in one stock and 10% in Treasury Bills. This reminds me of the popular Kelly Criterion, which is about bet sizing. It, too, can recommend dangerously large bets depending on the inputs. Even people who love the Kelly Criterion often recommend taking positions that are half, or one quarter, the size of what the model says.
I don’t worry about whether my portfolio is “optimized” at any given moment. And I dislike the idea you sometimes hear about how “if you hold a stock at today’s price, it’s like buying it.” I like to let my portfolio grow like an unruly garden.
Cutting Flowers, Watering Weeds
But the mantra of diversification has investors cut back on winners and add to losers – a process Peter Lynch described as “cutting the flowers and watering the weeds.”
Block agrees:
“The traditional attitude about the appropriateness of size often tends to force the portfolio manager to sell part of those stocks which have done the best and to add to those that have performed poorly. More often than not, both moves turn out to be mistakes.”
The ability to let winners run and become a large part of one’s portfolio is key to long-term outperformance. I expect (hope!) at least one of the stocks I own performs so well that it becomes a very large position over time. It would be tragic to cut that wonderful process short by trimming back a winner regularly.
One thing that emerges from Block’s essay is how various “portfolio elements are deeply intertwined.” As he says, “it seems impossible to discuss any single element without considering the inter-relationship with several of the other elements.” Diversification, time horizon and risk, for example, all reflect on each other.
Unfortunately, there are no easy recipes to follow when making (and managing) your own portfolio. But the above gets at a rough set of guidelines to consider. Portfolio management may be one of those things in life, like a recipe, where you have to adjust it to suit your taste.
In any event, it’s fun to think about and I could’ve written a lot more – there is so much to consider. For now, I'll sign off and wish you the best in running your own portfolio. And if you have come across any good material on this subject, please send it my way!
***
Published on January 17, 2023
Please see our disclaimers