Concentration: The Age-Old Question
Summary I’ve made the case for concentration before, and while I still advocate concentration, my original “time” argument was misplaced because of diminishing marginal returns on time. Concentration is largely a function of risk tolerance, which makes it far easier to find an appropriate level for an individual investor than it is for a professional. There is a lot of value in thinking about position sizing in terms of “starter” and “core” positions. Concentration is a subject I’ve written on before. In one of my first SA articles , I made the radical argument for a form of hyper-concentrated investing termed “Focus Investing” whereby one holds 3-10 positions… or even just one. Concentration is still a topic I give an inordinate amount of thought too and I wanted to share some of those thoughts here. This post also follows my first post on stock screening in a series communicating my investment process and philosophy. On Time My thoughts on position sizing have definitely evolved since my first article, and in hindsight, some of my arguments, while nice in theory, don’t hold in reality and my use of them demonstrated my inexperience as an investor. For example: Time The responsible investor follows each and every one of his holdings. It takes a constant amount of time per week to stay up on a company. I would advise at least an hour per week. Again this time is constant, whether that company makes up 2% of your portfolio or 100%… He could own 10 companies and still diligently follow them, but he’d have to devote ten hours instead of one. But wait, if he was willing to devote 10 hours total to stock market research when he held 10 companies, why not spend the same time researching, but while only holding one company? He could spend 5 hours per week keeping up on Apple and another 5 researching potential investments, comparing them against Apple, only considering them if they seemed much more attractive. My argument that investments require a constant amount of time and that 50% of an investor’s research time spent on a single investment is a good strategy ignores one very important principle: diminishing marginal returns or, more practically, the 80/20 rule. See one of my favorite Seeking Alpha articles , which discusses this subject, before continuing. The first hour of research yields more information and more valuable information than the 100th hour. The other problem with the time argument is sunk costs. We all know that a sunk cost should not influence decisions, but that they often do and, sadly, this is true even for decisions that we (the same people who are aware of the phenomenon) make. When you spend weeks researching a company and preparing an extensive, tidy investment thesis and article on the stock, it’s just harder not to take a position, independent of the actual prospects of the investment. While my time argument was somewhat off the mark (though it does hold in extreme cases; time is a serious problem for an actively managed portfolio of hundreds of stocks), I’ve still been a proponent of concentration, to a lesser extent, recently. Professional Constraints Concentration is largely a function of risk tolerance. This is not that meaningful if you are only managing your own money. All it means is that you must discover your risk tolerance and volatility tolerance and find a commensurate concentration level. Thing get tricky, however, when you are managing money for others. Introducing clients means more than one brain and in turn, risk tolerance is involved. What is the appropriate level now? If you are a manager like me with one strategy and one portfolio, then you should stick to the concentration level that you think is best for total returns, but I don’t think the story ends there. There needs to be some consideration that you are a professional and are managing other people’s money. There is a higher standard. This is one good reason to find like-minded clients. If you can withstand volatility, are long-term oriented, and are okay with concentration, look for clients with the same approach. Good luck- they’re rare! My other insight is that adopting a concentrated strategy as a new manager is tough because it requires credibility, to some extent, to be very concentrated. The base rate in investing is market returns and those are derived from a market portfolio, which is very diversified (500 stocks if we assume S&P 500 = market). Naturally, the more concentrated your portfolio gets, the more different it gets from the market and the further from the base rate its returns. You are going further out on a limb. It’s tough to do that with no professional track record. The logical next step is that if you’re a new manager you should be very diversified, but that’s a dangerous path I don’t want to take because it eliminates my positive optionality of earning extremely good returns and I’m in the investing industry for more than just money. Intellectual stimulation and an interesting, meaningful career is the most important thing I seek and the use of money as a means of keeping score and creating value is a big part of the financial aspect. In short, if I’m to succeed, I want to do so on my own terms and that rules out heavy diversification. If that means a slower ramp for my firm, so be it. “Starter” and “Core” I’m at a point now where my view on concentration is somewhat nuanced. Because I employ both deep research and empirical, systematic methods in my portfolio, not all positions will be sized equally – far from it. Right now, I have some positions that are 15-20% of my portfolio and some that are less than 1%. I think this dual concept of “starter” and “core” positions has been very helpful and is worth discussing. For me a starter position is 1% and a core position is much more than that, but the numbers don’t matter as much as the way of thinking about position sizing it represents. A starter position represents something that should, based on empirical evidence, outperform. That is a firm requirement. I talked about this extensively in a previous post . A starter position is also something I’ve done some research on, find interesting, and can model a good expected return with little to no downside on in an adverse case. But for some reason, it’s not fit to be a core position yet. The most common reasons are: I’m not sure I understand it, i.e. it may not be within my circle of competence The expected return I model is not high enough to exceed my absolute return hurdle, i.e. it’s not quite juicy enough I’ve not done enough research or thinking yet, i.e. I need more time Starter positions are crucial to my investment process because they allow me to slow down. Doing research needs to be a treasure hunt for me. I’m only interested in learning about companies when there is the possibility of it being in my portfolio and making me and my clients money. During the research process, it’s so tempting to act on research and invest. It’s hard to delay gratification. The problem is that good long-term investment decisions are made slowly. Gratification must be delayed. However, I’ve found that taking a small starter position up front helps to hold me over. Of course, I don’t do this for everything I research, but obviously far more than I end up taking core positions in as the chart above shows. It also provides an extra incentive to continue to dig deeper in the research. As Tom Gaynor says : When I buy some of something, I’m buying a library card. One of the reasons I buy some of something is to make myself think more deeply about it, read the reports and be more aware of it. It’s hard to overstate the positive impact starter positions have had for me. Not only have they performed well in aggregate, which is how I look at their performance, but they’ve rejuvenated me as an analyst. There was a rough patch where I only published four articles and made four investment decisions, not all of which were good ones, over a period of almost 8 months. (click to enlarge) I researched more than just four companies over this period, but not at as high a rate as I am now and not as effectively. The lack of gratification in the research process demotivated me. There’s no rule saying research needs to be fun for you to be a good investor, but for me I think it does need to be or I won’t find anything to invest in. It needs to be a treasure hunt and starter positions help a lot on that front. At the same time, when my research does, on rare occasions, generate what I think is a really good idea, I’m not going to only put 1% in it. There are times when the level of conviction and opportunity costs make anything but a big position a bad decision and that is when I am willing to take a core position. That is where concentration is needed. And in aggregate, you still end up with a pretty concentrated portfolio. More than half of my portfolio is in 6 stocks despite the large cash position. So I still advocate concentration, but clearly have a more nuanced view now and recognize that position sizing is a far more difficult issue than I initially had thought.