Portfolio management, conceptually, is about risk and return and about correlation and diversification. It is taught in academia as Modern Portfolio Theory; Markowitz’s efficient frontier, optimal portfolios and the like. While we firmly agree with the concepts of risk and diversification, it seems to us that this theory is too often put into practice using backwards-looking price correlations. These can fail when short- and medium-term trends break down, a prime example of which was at the end of the 2000’s. Remember the senior financier saying that their risk models had failed because 25-standard-deviation events, which statistically should only happen once in 100,000 years, had happened several days in a row I’m not sure what that told the financier, but it told us that the mathematical models needed work. To construct a diversified long-term stock portfolio that can avoid being caught out at times such as these, we need to look somewhere else. As such, we at Kennox look for diversification of the long-term (meaning five to ten years) earnings and cash flow streams of the companies that make up our portfolio. In the long term, these are the factors that will provide true diversification.
With that as our starting point, let us explain how we think about portfolio management. We view the question from two different angles, the combination of which gives us the confidence that we have a robust portfolio.
The first angle is that our portfolio is made up of individual stocks each of which we believe has an asymmetric risk profile. Never make predictions, especially about the future, as Mark Twain said. Instead of making specific predictions about what we guess might happen, we look for stocks that have limited downside with upside potential, no matter what the future brings. We do this by buying stocks that have identifiable underlying profits and cash flows, are conservatively financed, and are trading at valuations where we believe there is a built-in margin of safety. If each stock exhibits this risk profile, the entire portfolio should be resilient to significant tumbles.
At the same time, we aim to achieve what we call “duration diversification”, which we use to mean that not all the stocks are likely to perform at the same time. In our experience, the best portfolios have a few stocks drive performance in any one time period, followed by different stocks performing in the next period. Portfolio performance in this vein is more desirable than having a spectacular jump in all the share prices followed by a long plateau of flat returns. Thus we search out opportunities where we believe the upsides are uncorrelated.
This leads us to the second angle, a slightly different view of the same problem. As we never know what the future will bring, we also examine our portfolio looking for individual events or factors (economic, political, technological, natural disaster, or other) that could hinder the performance of a large number of our stocks simultaneously. If we find such an event, we try to find a new stock that (a) will benefit from that event to offset the impact on the portfolio as a whole, but also (b) is a strong investment in its own right. An example of this is oil; the global economy will be reliant on oil for many years to come and a dramatic spike in the price would negatively impact most of our stocks. As such, owning a company that will benefit from this is an attractive proposition and creates a more robust portfolio.
The result of our process is a portfolio with an assorted spread of stocks: recoveries; niche market players; ballast (steady long term holdings); unloved (trough earnings at trough multiples); cyclical; high growth potential, etc. As well, our portfolio ends up being diversified across traditional metrics such as geography, sector, and company size.
One last point: the Strategic Value Fund benefits from being unconstrained and we make the most of this. Some of the characteristics we seek are more easily found amongst smaller capitalised stocks, others are more readily discovered in certain geographies, and other characteristics are peculiar to certain industries. Without the ability to roam freely in all corners of the market, our approach to diversification would be significantly hampered.