De-Correlation
“When the well is dry, we know the worth of water.”
As described in my last post, the third leg of our current Asset Allocation strategy is to search for those niches within the investment universe we think are reasonably likely to remain De-Correlated across future scenarios. Our aim is to thread the needle between the reinvestment risk, inherent in hiding at the front end of the yield curve, and the investment risk, inherent in attempts to assign probabilities to forward-looking scenarios and weight our exposures to factors according to predicted scenario-dependent return configurations. Such De-Correlated investments offer the (hopefully not illusory) promise of an “anchor” allocation in the portfolio less sensitive to timing. But how do we go about looking for such investments and deciding whether they will remain De-Correlated across future scenarios? And what other criteria must be met for us to be comfortable investing? Answering these questions will be the focus of this post.
When searching for De-Correlated investments, our initial focus is how much relative liquidity has already found its way into the niche. “Relative” in this context refers to the ratio of $ liquidity allocated / $ liquidity required. Both numerator and denominator are imprecise. “Liquidity allocated” is difficult to measure, particularly given leverage. “Liquidity required” is a nebulous concept that refers to the capital that can be usefully deployed into an economic niche to deliver something of value for which there is potential demand. Though this idea can be easily criticised, some kind of normalisation is necessary to determine whether the liquidity in a given niche is “ample” or “scarce”. A given $ amount might flood a niche Private Credit segment but represent a drop in the ocean if deployed into physical Infrastructure.
Aside from relative liquidity, which usually can only be approximately estimated, there are other clues that might indicate to us that an investment is likely to remain De-Correlated. The first is the extent of financialization, which is both a proxy for liquidity and directly relevant to whether the capital-flows tail starts wagging the investment-returns dog. A niche that has been heavily financialised, and has ample liquidity allocated into it, is unlikely to remain De-Correlated when the tide goes out. “Financialisation” in this context is a relative concept that compares the influence of flows from speculative decisions to flows from underlying real economy decisions, and therefore is also difficult to measure exactly. However, an indirect assessment of the extent to which the financial industry has turned investment in a given niche into a “product” can be used as one reasonable proxy. As a general principle, a niche that is not well known, has not been turned into a product by intermediaries (especially flow products), and is difficult to access (or even find) is more likely De-Correlated.
Of course, low relative liquidity and financialisation are only the most important “necessary but not sufficient conditions”. We also need to decompose the investment at the risk factor level, to become comfortable that factor sensitivities to GDP and other Cyclical variables are not too high. Perhaps most important is the credit cycle, as this factor tends to drive many others across financial markets, and indeed is strongly related to liquidity via positive feedback loops. An investment niche that is not highly dependent on GDP growth, not vulnerable to prices inflated by debt and/or liquidity, and resistant to default cycles is a good candidate for De-Correlation. Additionally, we always apply a “value-added” sense check, which is to say we need to be able to articulate how capital allocated to the opportunity is used for something useful that people should fundamentally be willing to exchange value for. If we are unable to do so, demand might be illusory, with the niche likely correlated with broad speculative sentiment. This helps protect us from the seductive logic of artificial scarcity, used to push speculative products such as Crypto tokens. Finally, we generally need to identify some kind of market inefficiency to be willing to believe the investment niche is a De-Correlated opportunity, rather than just a bad idea that others are avoiding for good reason. Credible examples include unintended consequences of regulation, insufficient scale to financialise, or the incentives of market participants.
Since public markets tend to be liquid and are highly correlated during stressed periods, our search for De-Correlation generally involves private markets. Additionally, many potentially De-Correlated investment types are unusual, requiring specific domain expertise to select and manage suitable positions. This might imply we are the disadvantaged party in an information asymmetry situation. There may be an elevated risk that we are incorrect in our analysis — that the niche has poor risk-reward and / or is not really De-Correlated. We might be exploited by the agents we work with to access the niche. To mitigate these risks, we must carry out rigorous due diligence, relating to the platform through which we gain exposure, as well as by looking through to the underlying positions to validate through bottom-up analysis and scenario testing that the aggregate position is likely to behave as we expect from a top-down perspective. We also need to check that the idiosyncratic risk-reward of each individual position is desirable to avoid being taken in by top-down narratives or marketing. This is a highly resource-intensive process, but there is no free lunch, and this work is non-negotiable for us to be willing to commit capital.
An investment niche that meets the criteria set out above will usually possess modest valuations (though this is yet another example of a necessarily relative concept), limited overall leverage, deliver value into some baseline unmet demand, and ideally be accessed via a trustworthy stable platform. These features seem consistent with limited downside risk. Meanwhile, a less crowded niche will often offer relatively attractive returns due to risk premia being more insulated from search-for-yield dynamics. This implies that more De-Correlated niches might interestingly also often offer better risk-reward, a triple plus from a portfolio construction perspective. We might reasonably expect such opportunities to be rare, and so we should resist having too much confidence that we are correct in our analysis, especially given the information asymmetries. We must remain constructively Sceptical (ie: neither Naive nor Cynical) and remember “too good to be true usually is”. Most importantly, we must keep our aggregate exposure to any single investment niche within prudent risk limits, and then diligently seek orthogonal niches, rather than allocating excessively to any one potentially De-Correlated idea we think promising.
Looking beyond the portfolio construction merits of De-Correlated investments, allocating to these niches also tends to be a good fit with Titanium Birch’s mission. Where constrained “relative liquidity” conditions exist, this implies that additional capital allocated to that activity can contribute towards satisfying unmet demand. The additional time and effort that it takes us to find and diligence such opportunities may be rewarded not just with higher investment returns, lower downside risk, and more attractive correlation properties, but also by the incremental capital we deploy enhancing human wellbeing by more than its alternative potential uses. Such outcomes appear consistent and inter-related theoretically. Following the herd is for many market participants most safe from a career perspective, but in aggregate this can result in capital flowing to where it is already too abundant, while it remains scarce where it might be more useful. Titanium Birch is well placed to take advantage of, and help reduce, this inefficiency by aligning incentives with long-term performance and our mission.
Disclaimer: The content in this blog post should not be taken as investment advice and does not constitute any offer or solicitation offering or recommending any investment product.