UMBF - Reducing Luck: A Key Benefit Of An Unconcentrated Portfolio

2023-10-23 07:00:00 ET

Summary

• Investing is a learning process that requires time and a willingness to learn from mistakes and successes.
• Starting with a 1% portfolio weighting and not rebalancing allows winners to compound while limiting losses.
• Having a meaningful sample size of investments over time provides valuable feedback on investment techniques.

Introduction

This article will be the first of a series I write on the benefits of using an unconcentrated portfolio strategy. I've decided to write this series because I use an unconcentrated approach and I don't think many investors understand the benefits of doing so, since few famous investors take the approach (Ray Dalio is probably the exception to the rule). While there are some drawbacks, I will show how by making a few tweaks to the strategy one can minimize those drawbacks. Because of the scope of this topic, I have decided to break the discussion down into the individual topics I think are most important. This article's topic will be about data, sample sizes, and making adjustments as one learns about investing to reduce the role of luck, which are understated benefits of a well-designed, unconcentrated portfolio strategy.

Investing is an ongoing learning process

Imagine you walk into a room with some cash in your pocket and at the center of that room there is a poker table with 8 dapper, well-dressed players sitting around it. They are smiling and jolly and appear to be having a swell time. You catch their attention, and they wave you over, gesture toward an open seat, and say "Go ahead, have a seat, play a few hands with us." You tell them that you don't know how to play poker. They respond "Oh, it's easy. We'll teach you. Pull up a chair."

Most sensible people who know anything about poker would know that the odds are if you sit down at a poker table without knowing much about the game and having much experience, you will probably leave the table poorer than when you sat down. The incentives of this situation are aligned in such a way that the people already sitting at the table, waiting to take your money, are not likely to tell you the high odds of you losing money to them. Instead, their incentives are to make it seem like poker is an easy game that anyone can play, and your chances of winning are as good as anyone else's. It would be rare in this situation for them to gesture to you a shelf full of poker-playing strategy books, and say "Go read all those books, and then when you are finished, we'll play with you after you understand the game better."

Investing is very similar. Anyone who says or implies that investing is "easy", is probably trying to separate you from your money. And unless you get very lucky, they will succeed. It takes skill or an enormous amount of luck to do well as a stock investor over the long term. Developing skill in investing is just like most other skills, it takes time and a willingness to learn from mistakes and successes. People should assume they will not start a master.

With this framework in mind, if a person wanted to learn how to be a successful poker player it would make sense to read those books on how to play poker, watch some professional poker players and observe their techniques, then play some small-stakes games, carefully analyzing over time the strategies and tactics that they used, which of those failed and succeeded, and why. The alternative, of course, is they could just try to rely on luck, but I think most of us would agree that luck is not the best way to earn money playing poker over time.

It's usually at this point in the article a few readers skip right down to the comment section to tell me I'm wrong to compare poker with investing because they aren't the same thing. To them, I would say that the way most beginning investors start "investing" is much closer to poker than actual investing. But I would also say that it's much easier to earn money investing over time if one is willing to learn than it is to earn money playing poker even if one is willing to learn how to play poker well. But this is very different than saying investing is "easy". There have been times when stock investing was easier, like from 1983-1999, and in the period we just experienced since 2009. But there have also been long periods of time where investing was very difficult, like from 1968-1981 and 2000-2012. We should be careful about recency bias in this regard. We have experienced a very long run of time where investing has been relatively easy, for stocks, it has been easy from 2009 to the end of 2021, and for bonds, it has been easy from the early 1980s until the end of 2021. These long "easy" periods of investing are usually followed by much harder periods, and I don't think both the stock and bond markets have ever had "easy" runs that coincided for as long as these have.

Because nobody starts as a great investor, there is a learning process that must take place. Because of this, I am of the opinion that an investing strategy that allows an investor to (1) measure their results and have a decent sample size, and (2) minimize the damage of their mistakes while maximizing their successes, is probably going to be a pretty good investing strategy over time as long as the investor is willing to learn and improve along the way.

Here is how I do this.

I start each new position with a 1% portfolio weighting, and I do not rebalance for rebalancing sake. This allows the ideas I correctly invested in to compound over the long term and limits the losses of any individual idea to about 1% of the portfolio. So, winners can potentially grow 10x or more, but losers have a limit. You can see that taking this approach could, in theory, produce a concentrated portfolio over a few decades, but that concentration would only happen with proven long-term winners. This solves the commonly stated problem of "diworsification", which is the critique often leveled at unconcentrated portfolios. The approach I describe above is especially beneficial to newer investors who are likely to make more mistakes with their stock selections.

But that's not the biggest benefit.

The biggest benefit is that the investor can achieve meaningful sample sizes of their investing strategy over time to see how they have performed on average. This data can give the investor valuable feedback regarding the strengths and weaknesses of their techniques over time. An investor with only a handful of investments, on the other hand, will not have this information. I'll use a sample of my stock ratings on Seeking Alpha as an example. I'll start in January 2022 at the peak of the market. I will include all single-stock articles with a "Sell" or "Strong Sell", and all the ones with a "Buy" or "Strong Buy", since then, and see how many I have and how they have compared relative to the S&P 500. If I wrote multiple articles with the same rating, I will use the oldest one within this time period, and if I changed my rating I use the returns between the two ratings. I have chosen this method because you can find all these articles on Seeking Alpha available on my profile page here .

I will share the total return since the publication of the article and also compare that to the returns of ( SPY ) and the equal-weighted S&P 500 ETF ( RSP ). During a time period this short, because my positions start equally weighted, it makes more sense to compare to the equal-weight index, but over a longer period of time, it might make more sense to compare to SPY which lets its winner run as my strategy is designed to do. I will share both benchmarks in the tables below.

Let's start with the "Sell" ratings that are currently negative. This is 17 total out of the 25 sell ratings over this period.

*AMD position half was sold, which you can read about in my article " After Doubling, It's Time To Trim AMD Stock ".

Next, I will share the Sell ratings that had positive returns.

Now let's look at the Buys that were negative:

And last, let's look at the Buys that were positive:

**I sold my SLGN position this past summer but haven't written a follow-up article about it. I remain long the rest of these positions.

The Buys had a 90% positive return outcome. 68% of the "Sells" had a negative return outcome. In terms of the magnitude of returns, the Buys, as a group averaged a +25.61% return with a sample size of 10. The Sells averaged a -3.66% return with a sample size of 25.

The Buys average return of +25.61% compared to +4.18% for the equal-weighted S&P 500 and +12.34% for the cap-weighted S&P 500. I think given the short time frame of less than 2 years we are dealing with, the equal-weighted S&P 500 is the best comparison and my "Buys" (All of which I actually bought myself) performed about 6x better, but even if one chose to use the normal S&P 500, the Buy's returns were 2x better. With a 90% positive return rate, when I analyze the performance of this strategy over this time period it looks very good, but it does have a relatively small sample size of 10 positions. Still, this data is much more useful than what is available to a person who has a very concentrated portfolio of just a few stocks, which really doesn't tell them anything meaningful about the role luck plays in their outcomes.

Also, since I have been using these two investing strategies for an average of about 5 years, I actually have more data on them than I shared above, which only includes public articles published within the past two years. So, just looking at the additional positions that were fully realized over the past 5 years I have about an 80% positive return rate, and an average 10% (or 1,000 basis point) outperformance compared to SPY if purchased and sold on the same dates. Which is actually pretty similar to what we see above. This is enough data to give me meaningful feedback on my strategies that investors who use a concentrated approach do not get. For example, I can see that my emerging market investments performed very poorly compared to my developed market investments, so I have invested in fewer EM stocks recently, and will likely get better returns going forward because of that.

The "Sell's average return of -3.66% compared to a -2.51% return for the equal-weight S&P 500 and a +4.76% average return for the normal S&P 500. In this case, we have a significantly bigger sample size of 25 and these stocks come from a variety of industries. This gives us useful information about my valuation process. I can see that it's likely if I sell stocks when they appear to be very overvalued, I'm not likely to suffer much of an opportunity cost over time. Sure, there will be an occasional outlier that will go on to make significantly higher highs and never come down, but those are outliers and over time they will be drowned out by the Sells that go on to lose money or produce below-market-average returns. None of this information is available in a significant scale to the concentrated investor until they have been investing for decades.

Because becoming a better investor involves a continuous process of learning over time, I can review this data and research the areas that might be outliers, then try to figure out what is going on, and determine whether there is an adjustment I can make to improve. For example, when I look at AMD ( AMD ), which produced a 100% return, that is actually within the range of return I was expecting when I bought the stock, it just came quicker than I would normally have expected. With the "Sell" stocks, Caterpillar ( CAT ) and Tesla ( TSLA ) I would expect to be deeply cyclical during a recession, and we haven't had a recession, yet, so the fact that they have risen this year doesn't mean much with regard to my sell thesis. They will still probably both produce negative returns at some point when measured from the time my sell articles came out. But, with Palo Alto ( PANW ), I think I may have genuinely made an error with my valuation and might have fundamentally underestimated their future earnings growth potential. I probably should have categorized that business differently and used a different type of analysis more suited to fast-growth businesses. This is something I would pay closer attention to in the future so that I am less likely to sell a stock like Pala Alto too soon and suffer an opportunity cost.

Reducing Luck

We can never know for sure how much of our success or failure is the result of luck. It takes a decent sample size and a significant amount of time to become more confident that luck, rather than one's process and strategy, produces the investing results they end up with. Time is something we can't speed up, but the sample size is something we can control, and by taking an equal-weighted approach regarding one's initial position sizes, an investor can gather useful information regarding how their strategies are working, and also likely identify areas where things are working better or worse so they can be adjusted and improved.

I can say with a lot of confidence that most investors are not skilled enough to produce good returns with a concentrated portfolio over the long term. And no investors start off with adequate skills to produce good returns with a concentrated portfolio without relying on a great deal of luck. We will all hear from those investors who were lucky enough to put most of their net worth in Monster Beverage ( MNST ), NVIDIA ( NVDA ), and Apple ( AAPL ) in 2003, and are now quite wealthy. But we don't hear from the 99,000 losers who tried that and failed. Additionally, I have yet to encounter one of these concentrated investors who was using a process in 2003 that identified 2 out of the 3 listed above. It's almost always only one where they happened to get some special insight at a good time, made a big bet, and stuck with it. Good for them. I feel good for them the same way I feel good for the factory worker who wins the Powerball lottery. But just as the lottery isn't a good investing strategy, even if it produces a few big winners, very concentrated investing unless one owns a controlling interest in the business, isn't a good investing strategy for the vast majority of people either.

Conclusion

New investors are made to feel like they are much better investors than they actually are so they can be separated from their money by more experienced market participants. Investors should approach investing with a large dose of humility. That doesn't mean investors should throw their hands up, give up, become a Boglehead, or put all their money in a Target-Date Fund. Ignorance is not a good investing strategy any more than hubris is. Just look at all the investors who listened to the advice of investing in a 60/40 allocation who have been crushed in the past two years. It's okay for an investor to start out indexing, but over time it is worth it to learn how to invest in individual stocks. The easiest way to do this is to start out with 50% cash, 50% S&P 500, and add 1% weighted individual positions equally from those two pools of money when the investor thinks they have found a good investment.

For new investors, I would actually limit the universe of stocks they choose from to the S&P 500 constituents. Then track how the individual positions are doing over time compared to those two default positions (cash and SPY/VOO). Unless we are in a recession or deep bear market, don't add more than one position a month, and after a few years, you'll have enough data to tell you something useful about your investing strategy. Use that information to improve and keep investing. You might do so badly, you realize individual investing isn't for you. But for a lot of people, if they keep learning, they can be excellent investors by the time they are nearing or are in retirement, which is when it is most important not to make big mistakes.

For further details see: