Diversification and concentration

When less is more

Clichés, mantras, dogmas. They are an inevitable part of our existence. Investing is no exception. Diversification is one of those widely cited yet utterly misunderstood dogmas.

Diversification is a tool for constructing a portfolio, but it is not the only one. For example, diversified concentration is a viable approach. It is the one I rely on for my book management. Before I elaborate on how it works, let's discuss diversification.

Diversification has been abused as the only way to play the game. MPT (Modern Portfolio Theory) promotes it as the ultimate portfolio composition tool. However, MPT is based on flawed assumptions. Let’s name a few:

• All market participants are entirely rational.

• Markets move, following a random walk pattern.

• Markets are efficient or EMH(Efficient Markets Hypothesis).

• Returns distribution follows the Bell curve.

The core tenets of the MPT are that all people make well-argued rational decisions and the markets move according to random walk patterns. Hence, the markets are efficient, resulting in a normal distribution of returns. However, we must take those premises with a grain of salt. They illustrate certain market anomalies, not the market modus operandi.

Market participants are irrational, and rationality is applied only to justify our irrationality. Markets are path-dependent because the agents have a memory for the past. Hence, the subsequent price fluctuation is not independent from the previous. Markets swing between inefficient efficiency and efficient inefficiency. Gaussian distribution is an exception, not a rule because the markets are not fully efficient or inefficient.

So, the MPT and EMH give an imperfect picture of the markets and investing. Despite that, we take their postulates for granted, and diversification is one of them.

Some thoughts on diversification

Blindly following MPT is not the only issue. Market participants too often do not scrutinize their beliefs. Assuming that MPT is bulletproof, we believe that broad diversification is the best way to manage a portfolio.

Naively mirroring MPT/diversification leads to widespread misconceptions:  

• Investing following the principle of "the more, the better" without considering regions, industries, and sectors, the leading criterion is quantity. A portfolio including more than 20 stocks is too diversified.

• Investing in multiple companies but only in one region and industry. Buying ten gold mining companies reduces individual company risk but retains systematic risk inherent in the gold mining industry. Hence, we have a position in one gold miner.

• Investing in one or two companies because we're "sure" nothing can go wrong. Usually, that fallacy comes with overconfidence and zero risk management. Sooner rather than later, we pay a steep price for our negligence.

• All positions in the portfolio are equal. We put all Alpha ideas under one denominator. Even the ideas with the best risk reward and highest odds get the same size as more average ideas.

In the following paragraphs, I discuss those fallacies and their exceptions.

Over diversification

Overdiversification is a classic. It emphasizes quantity, not quality. The questions below illustrate the issues of overdiversification.

• Can we find dozens or even hundreds of ideas with high odds and attractive risk reward?

• How thoroughly can we research so many ideas (companies) to conclude that they are great?

• How effectively can we manage them once they are part of our portfolio?

To all three questions, my answer is: No, we can't.

Therefore, over-diversification is harmful in most cases. Exceptions are factors-based strategies. Nevertheless, they can be performed via exposure to indexes and funds. That way, the burden of following and managing tenths of positions is reduced to a few large positions.

Overconcentration on region and industry

Investing in multiple companies from one region/industry is not that popular. This does not mean that it is not dangerous. I've tried it myself. Years ago, I was a gold bug. So, I was all in gold, physical, and mining companies and thought I was diversified. Fortunately, I got away with it. I sold most stocks at a good profit and restructured my portfolio.

Concentration in just one industry is dangerous because we expose ourselves to its industry/region-specific risk. To go back to my example. If I had stayed only in gold mining companies, in case of a severe downturn in the gold price, my portfolio would have followed a decline in the gold spot.

If we like gold and want broad exposure, using indexes and ETFs is much wiser. They provide extensive diversification to the idiosyncratic risk yet relieve the weight of following tenths or even hundreds of companies. On top of that, most liquid thematic ETFs offer liquid call options. Accordingly, we can express our ideas via ETF and LEAPS calls. The former gives Beta to the chosen theme, while the latter delivers the proverbial Alpha.

Overconcentration in one or a few companies

Betting on just one or a few companies is usually too risky. We expose ourselves to the company’s idiosyncratic risk. There is one exception.

Investors who actively manage their books occasionally stumble across high-risk reward ideas with significantly skewed odds in their favor. At that point, going (almost) all in is the right thing to do. Of course, under a few considerations.

Our first task (as always) is to define risk and position size. Many market participants confuse both metrics. I will illustrate the difference with a simple mental experiment. We have a portfolio of $100 and bought one share of Company XYZ for $100/share. What is the risk and the position size relative to our portfolio?

The position size is 100%, yet the risk is not mandatory to be 100%. If we set a stop loss at $99, we have a 1% risk on positions XYZ. The same logic applies; if we set a stop loss at $90, we have a 10% risk on position XYZ. Conversely, if we do not consider our emergency exit strategy, i.e., position stop loss, technically, we have a 100% risk.

So, if the odds and risk-reward ratio are tilted in our favor, we can bet 100% of our book on one position while keeping the risk at low single-digit levels. Soros and Druckenmiller's approach is a case in point. They may go all in when they spot a lucrative bet, yet they have prudently defined risk.

Another crucial aspect of this approach is time. Define the time frame of the positions. Remember, the odds of being right are a function of the time we stay in positions. That’s why derivatives have time value, too. The longer the derivative's life span, the higher the chances it becomes ITM. On the other hand, the longer we stay in position, the higher the opportunity cost. Hence, we must balance between time preferences and the opportunity cost.

Such opportunities do not come daily, so we usually balance our books with moderately but not equally sized bets. This leads me to the next fallacy.

Equal sizing

The equal weighting of all positions in the portfolio often goes along with over-diversification. Having many different stocks with equal weighting indicates that we attribute equal credibility to all ideas. This is a consequence of mediocre or missing analysis. There aren't that many great companies, and among the few available, the degree of credibility also varies widely.

A partial exception is a portfolio based on the All-Weather portfolio (or its spin-offs like Butterfly and Permanent Portfolio). The proportions are fixed yet not identical. Once per year, we must rebalance our book to achieve the prescribed asset distribution. Even in that case, however, the weight of the positions is not equal.

My way: diversified concentration

Simplicity beats complexity. Always. I follow that motto in my portfolio management. 

Here are my core rules for managing my book:

• Play a few major themes, preferably below five. For 2024, I bet on PGMs, shipping, uranium, and LatAm.

• Apply two strategies: event-driven investing and mining investing for non-geologists. I wrote about both approaches here and here.

• Use equity and options (if available) for event-driven plays. I combine ETFs, options, and occasionally equity for broad thematic plays.

• Keep a total number of assets below 20. The most significant positions, i.e., the core assets, are ETFs/ETCs and event-driven equity plays. The remaining are scattered among multiple small bets via LEAPS calls.

• Rely on the fractional Kelly formula to estimate the position size.

• When the stars are aligned, go for the jugular.

Every point deserves a separate article. However, my goal today is to examine diversification as a tool and my interpretation of portfolio management.

The investing process has three steps: research, valuation, and execution. The last one represents risk management. Risk management is divided into two segments: micro-management, which concerns individual position management, and macro-management, which concerns portfolio composition.

Both aspects are equally important and not interchangeable. Micro defines position sizing, risk, and execution metrics (stop loss, take profit, options). Macro defines the portfolio metrics: number of themes, distribution, total risk, correlation, and maximum potential drawdown.

Today, I shared my approach to macro management, i.e., portfolio compositions. In the coming weeks, I will cover the aspects of micromanagement.