One of the easiest things to overlook when constructing a system is which markets you intend to trade and what percentage of capital you should allocate to each of those markets. The guys at Darwin Investment Strategies are experts in this department. They have recently gone back to basics and are working on a series of posts that break the idea of diversification down into easy to digest portions.
Their most recent article addresses the subject from the very beginning and discusses why diversification needs to go beyond the typical 60%/40% equities to bonds that is the industry standard. They argue that the 60/40 approach is only productive in one of the four possible economic regimes.
Generally we can say that the global economy is defined by four regimes, each of which combines a growth vector with an inflation vector. A period of accelerating growth in combination with rising inflation might be termed an ‘inflationary boom’, while a combination of accelerating growth simultaneous with falling inflation might be represent a ‘disinflationary boom’. On the other side of the axis, a period of slowing growth combined with rising inflation is often termed ‘stagflation’, and finally we might call a period of decelerating growth concurrent with falling inflation a ‘deflationary bust’.
After defining each of the four possible economic regimes, they move forward by placing each possible asset class on the two-dimensional chart where it performs best. Then, they begin to address how they should weight a portfolio across each of the four quadrants so that it is able to perform regardless of which economic regime is present.
To the extent that we are able to assign a relative probability to each regime over our rebalance horizon, we can bias our allocations in favour of some regimes over others. Unfortunately, it is very difficult to know with any confidence which regime we are actually in at any point in time, and much harder still to predict when the regime will change, and the direction of transition. If we were to assume that each of the four regimes is equally likely at any point in time, we might be wise to allocate 25% to each quadrant.
They also address the concern of whether to allocate 25% of the entire capital or not, but decide to keep things simple for this post.
To keep things simple, as a first approximation we will assume that we want to equally distribute capital across the four quadrants in order to ensure the portfolio is equally resilient to all four major states of the world via structural diversification.
They decide to split the capital equally across each of the four quadrants. Then they split each quadrants capital equally across each of the different assets in that quadrant. They proceed to discuss the pros and cons of this type of portfolio.
The advantage of the structural diversification framework outlined in this article is that it relies on an understanding of the various drivers of asset returns that is consistent with most financial and economic theory. These are excellent qualities so long as assets behave as they theoretically should, and the universe is coherent and thoughtfully diversified. However, managers are often faced with noisy, incoherent universes that have dramatically unbalanced exposures that would seriously impair the effectiveness of this simple approach. Further, risk characteristics and correlations can change quite dramatically through time, and a static allocation like this policy portfolio is unable to respond to these changes. As a result, this portfolio is vulnerable to extreme shocks.
They conclude the piece by reminding us that this portfolio would not be an ideal solution, but it does provide us with an interesting model to guide our thinking about this topic. Putting this level of thought into the markets that our systems trade could go a long way towards producing consistent returns in all types of economic regimes.