Cyclicality and Self Similarity

In Howard Marks’ book Mastering the Market Cycle, he describes a multitude of cyclical phenomena that operate on, and compose, markets. The term cycle is perhaps misleading, as these are not cycles in the sense of looping back on themselves (as in bicycle), but rather oscillating forces around secular procession. In the interest of brevity, I am going to continue using Marks’ term.

The point of the book is to describe various cyclical phenoma and then detail how understanding these cycles can position one best to take advantage of, and not be burned by, them from an investing viewpoint. Marks contends that most of us are not cycle-aware and interpret cause-effect relationships as merely discrete events occuring through time. We are often blind to the forces driving events and how those events feed back into those same forces. Without this understanding, all we see is a mess of confusing noise, which is much of what 24 hours news comes across as.

Cycles act like repetitive processes, never the same, but bearing similarities through time. Their shape (length and extent) changes, and the events they manifest are certainly always different, but Marks argues that to some extent the same cycles are always at play in the background. As an example, let’s take what Marks considers to be one of the most important cycles operating on/within markets: the credit cycle.

The credit cycle can be thought of as the opening and closing of the credit window, which is a metaphor for the availability of credit. In “good times” credit is widely available, and so the window is considered open, but, as we saw in 2008-2009, the window can close very dramatically if economic conditions change, leaving people without access to credit who would otherwise have it at their fingertips. Marks details the progression of the cycle:

1. Stringent borrowing rules and risk aversion amongst lenders aid stable economic conditions.
2. The stable economic conditions lead to lower prospective returns in what are considered safe loans and so lead to a relaxing of the risk aversion amongst creditors as they seek higher returns.
3. Relaxed risk aversion means more loans are issued to riskier borrowers and a race to the bottom takes place as creditors pursue ever-riskier borrowers.
4. The continued pursuit of ever riskier investments leads to an eventual buckling as loans turn bad.
5. Investors, panicked by the environment, withdraw their credit except for to the safest borrowers. And so the process repeats.

For many, the most obvious example of this process would be the early-2000 period and subsequent financial crisis of 2008, but the process is at play throughout history. Indeed, Marks does not explicitly say this, but I think these cycles also operate at various scales. The cycle that led to the financial crisis of 2008 was a large scale, pretty much global, cycle¹, but we have experienced smaller credit cycles. In the UK specifically, we may have entered our own phase of a cycle with the September 2022 mini-budget. Though the catalyst in this case was government policy, and therefore the specifics differ, we seem to have entered our own iteration, consisting of more risk aversion than in the years prior to COVID-19.

Taking the idea to the extreme, we know there are geological/biospherical cycles at play, and that these will have an effect on smaller cycles by defining the environment they operate within. Geological periods are often eras of great flourishing for certain groups of flora and fauna, punctuated by extinction events that completely shake up the playing field afterwards. Indeed, without the K-Pg event mammals would likely not have become the dominant animals, and perhaps there’d be no computer for me to type this on. Many experts believe we are currently in an extinction event as we enter the holocene, driven by increasing human impact on the biopshere.

What I realised reading Marks’ book, is that cycles at differing scales will influence each other. Marks notes that each cycle he details exerts its forces on the others, but he makes no hierarchical distinction. I think there is one. For our purposes, it is clear that the geological cycle is a “lower-level” cycle than the credit one. The genesis of the holocene will change the playing field that economies operate on, all economic cycles will be influenced by it, though how quickly and to what extent it is unclear. But the geological cycle has been influenced itself by higher-level economic cycles. Our economic activity is feeding back and driving geological change².

The pattern recognition part of my brain cannot help but draw these connections. At various scales we can observe cyclical phenomena. A world in processional flux. So as we zoom in from the geological scale, we see the climate scale, eventually into Marks’ topic of the economic scale, which itself is composed of multiple scales, for the global economy, and at regional and national levels too. We can probably keep this going to ever higher constructs: it is cycles all the way down (or up).

Mathematicians call this property self-similarity. At different scales, we observe the same or similar structures. Perhaps the most famous example of this is the mandelbrot set, which presents indistinguishably similar structures on its boundary as one zooms further into its infinite regress.

I think cycles have a certain self-similarity. Of course, their specific domains, events, etc. are wildly different. No geologist is consulted on monetary policy, but the rhythmic movements are shared. Geological cycles take the shape of calm punctuated by extreme stress at the boundary of each period. So too we see this in the economic domain. The shape is found in politics, where lengthy periods of consensus are suddenly broken apart as new ideas take root more rapidly than anyone could have imagined.

What I think this means is that by thinking laterally, by studying other areas outside our specialities, we can make connections across subject field boundaries. Because these different areas share similar structures, we can take lessons from each of them. I am in no doubt that other structures are common across a variety of domains. Cycles are but one that can be observed so widely. The challenge is being receptive enough to see them, and to strip away the specifics to observe the similarity in the structures.