Although we often hear (or read) the words “diversified” and “portfolio” together, we should question if it is really true that an investment portfolio can really be well diversified.
To speak about diversification, we first need to speak about “correlation”. When it comes to correlation, our creed can be summed up in a statement once made by Ray Dalio of Bridgewater Associates:
“People think that a thing called correlation exists. That’s wrong. What is really happening is that each market is behaving logically based on its own determinants, and as the nature of those determinants changes, what we call correlation changes”. (Jack D. Schwager, “Hedge Fund Market Wizards, How Winning Traders Win”, Wiley 2012)
The main idea behind portfolio diversification is that by determining the correlation between various markets is possible to build a portfolio where the profits realized by each portfolio component (i.e. each market invested in) are somehow offsetting the losses of one or more other components. For example: if an investment manager believes that rising Oil prices are directly correlated to rising stock market prices (i.e. the two markets go up or down together, in sync), she could be LONG Oil and SHORT the S&P500 Index and so when Oil prices start to rise and the S&P500 starts to fall, in theory the profits realized by the Oil market component should compensate the losses taken on the S&P500 market component. Another example could be that of a manager building a portfolio that invests in a number of different markets: stocks, bonds and currencies, with the goal of realizing profits in at least one or more markets at the same time, to compensate the losses in other markets invested in.
The risk with this type of approach to portfolio diversification is that there is a possibility that many of the markets you invest in goes in the opposite direction of your investment at once, for example Oil may fall while the S&P500 rises and at that point you will be completely on the wrong side of the game on all your bets.
Remember what we said before: correlation does not really exist, it is temporary and it tends to change in time, so it is hard to be able to make a bet on a persistent, stable correlation. The next image exemplifies the various Oil/S&P500 correlation scenarios encountered in the last few years, including one called “diverging” correlation where basically we have a an existing correlation (i.e. Oil up , S&P500 up) that progressively falters. You can see clearly that until the summer 2014 the trend of the Oil price was quite directly “correlated” to the price of the S&P500, then suddenly Oil prices started to fall while S&P500 prices kept rising.
This is one of the countless examples that could be made to show that the idea of exploiting a stable correlation between markets is faulted: two markets may be somehow correlated at times (aren’t they all somehow?), but it is very hard to determine when their current relationship will break apart and change sign.
Analysts and managers often stick the word “correlation” on two markets when their trends appear synchronized but that is an observation based on past data, there is no certainty that a previously existing correlation will continue into the future.
In conclusion, it’s easy to talk about “diversification” on paper, but unless you are certain that a stable correlation exists between two different markets, you cannot create any real diversification. Your diversification will be at best random and it will only work if a sufficient number of directional bets you’ve made is offsetting all the wrong bets.
Another way to think about the issue of (real) diversification is the following: there are only two possible market directions for each market: “up” or “down” (we exclude “sideways/flat” because that direction does not generate a profit or a loss). If we bet that one market will go up, the only way to offset for sure the potential losses on that market is by going SHORT the same market, so if that market falls, we will profit on the SHORT side, and have a loss on the LONG side (see image below). But how we are going to do this without having the losses on the LONG side offsetting completely the profits made on the LONG side? It’s clearly a problem without solution, and it reveals the cold truth about well-diversified portfolios: they do not exist.
There is actually an approach that can lead to real portfolio diversification, but it is not based simply on investing in several different markets. It is an approach that some very advanced hedge funds adopt when building their portfolio: we call it the “strategy-diversified” approach.
Market returns can be obtained applying a various number of approaches that can be grouped in five type of strategies (see image below): Trend and Reversion (based on Price input), Yield, Growth and Quality (based on Fundamental input).
Inevitably not all these strategies are going to work at the same time, for example a Mean Reversion strategy will work well during periods of high volatility when a Trend-following strategy will not work. However, a well-calibrated, mathematical, scientific combination of strategies exploiting a number of “market effects” (i.e. the manifestation of certain market behaviors) is what allows the building of a truly diversified investment portfolio that can surely reap at least some profits during any market scenario, but the diversification happens combining different strategies, not simply selecting different portfolio components!
Going back to our example of a manager that believes that rising Oil prices are directly correlated to rising stock market prices, she could go LONG Oil and SHORT the S&P500 Index and so when Oil prices start to rise and the S&P500 starts to fall, the profits realized by the Oil market component should compensate the losses taken on the S&P500 market component. However as we have explained previously, the logic of this approach is faulted at its core because:
a) there could be a scenario where the losses realized by one component completely offset the gains realized by the other component.
b) if all the existing “correlation” change, the manager may find herself on the wrong side of the market on all of its portfolio components.
The combination of different investment strategies is a highly complex task that occupies every day armies of scientists and mathematicians at the most advanced global hedge funds and certainly cannot be discused in deep in a simple blog post: the subject is vast and each firm has his own secret formula to balance various strategies and you are certainly not going to find that information publicly in a book or on the internet, but it is exactly that skill, the ability to create and combine strategies that work in as many as possible market scenarios, that makes it possible to create truly well-diversified investment portfolios.
[To read more articles like this follow Retracement Levels’ Blog.]
If you are interested in learning more about Retracement Levels’ market analysis and investment products please follow this link.
