Goldiloss
“Live to fight another day.”
A “Goldilocks” regime, defined as robust (but not too hot) economic growth and moderate (but not too cold) inflation, appears to now be the 2024 consensus among forecasters and investors. This follows 2023, when recession (“Deflation” or even worse “Stagflation”) was the consensus at the start of the year, but did not materialise as growth held up better than expected even as inflation came down.
I am not convinced “Goldilocks” is the most likely outcome this year, and suspect recency bias and incentives may be playing a role in distorting consensus views. Given typical lags in the economic impact of monetary tightening, it is unsurprising recession has not yet occurred, particularly after the Fed stepped in quickly to prevent a liquidity crisis emanating from the US Regional Banks (a more rapid, but unnecessarily costly, transmission mechanism for monetary policy that it seems proper the lender of last resort should prevent). This intervention allowed the Fed to continue hiking policy rates in 2023, with the lagged economic impact incident in the future, not past. Over the past three non-Covid cycles, it took 9 to 18 months from the last rate hike to recession, which would imply a recession starting 2Q24 to 1Q25.
Every cycle is different, but the sequencing of weakness in leading economic indicators when tight monetary policy results in a recession often follows a “rhyming” path. Weakness in Housing and Manufacturing tends to be most leading, and both have weakened significantly. Next up is corporate profitability, and EPS (particularly SMid Cap, responsible for most employment) appears to have peaked in 1Q 2023 and is steadily drifting lower. Services is often the next shoe to drop, more coincident with recession, with the ISM survey holding up just above 50 currently. Unemployment is last in the sequence, but is a lagging indicator, and even this is close to triggering the “Sahm rule”. Pointing out that “unemployment is low” is a strange counter-argument to the possibility of a coming downturn, given that unemployment is always lowest just prior to a recession occurring, and effectively a necessary precondition from a business cycle theory perspective.
There are of course seductive, and not unreasonable, arguments that “this time is different”. Covid and the consequent reflexive fiscal reaction exogenously perturbed the typical business cycle, resulting in short-term oscillations in economic data of such large amplitude that it is difficult to be confident where we are positioned in the underlying cycle. To the extent that inflation really was “transitory”, as the decline last year has (partially at least) vindicated, perhaps the Fed will be able to reverse its monetary tightening, and we will revert to an earlier phase of the cycle. “Immaculate disinflation” and lower interest rates would also weaken constraints on fiscal support continuing. The application of artificial intelligence could facilitate low inflation economic growth by raising productivity, while US onshoring and the necessary infrastructure to support it could create higher wage employment that supports income growth and a higher participation rate.
On the other hand, the distortions from Covid and fiscal policy cut both ways. Softening economic data and credit conditions heading into 2023, as post-Covid “revenge spending” was exhausted, may have encouraged premature consensus expectations of recession, despite the typical lag of monetary tightening not having had time to work its way through the system. This consensus likely then encouraged significant fiscal expansion last year, which reflexively supported economic activity and helped many leading indicators to stabilise at soft but not quite recessionary levels. As it became apparent that recession was not on the cards, and as leading indicators were flattered relative to the norm, the (impatient) consensus has now shifted away from anticipating a recession that occurs in line with typical lags. It would not be so surprising to observe fiscal support reflexively diminish (despite the election), and leading indicators catching up lower.
The intention of this memo is not to argue for a 2024 recession. No one knows whether this will occur. But “Goldilocks” does not seem obviously more likely, and it seems dangerous to overweight the probability of this outcome when making Asset Allocation decisions. My preferred approach is to construct a portfolio that can perform well across the most likely scenarios, not “bet” on any specific prediction. Having said this, I think a “Stagflation” scenario would be unstable and transition quickly to “Deflation” as “higher for longer” monetary policy comes back into play. Meanwhile, unless we get exponentially accelerating fiscal stimulus, it is difficult to see how the “Reflation” scenario would occur. For the reasons set out in Policy Agnosticism, it is difficult for fiscal authorities to sustain such policies for long, as feedback mechanisms push increasingly in the other direction. Remaining funds in Covid-related programmes that helped support fiscal in 2023 are increasingly exhausted. We think it reasonable to simplify to “Goldilocks” vs “Deflation” from a tactical perspective, and rely on our “core” Real Asset and De-Correlated strategic positions in case of the less likely (in the short-run) “Reflation” or “Stagflation”.
It is interesting to consider whether a “Goldilocks Portfolio” or a “Deflation Portfolio” makes more sense from a tactical perspective. I will not discuss specific positions, to avoid being misconstrued as giving investment advice, but it should be obvious that even a very simple portfolio consisting of Equities and Bonds can be weighted to perform better in one or the other of the outcomes. More complex portfolios can be constructed with the same objective. Let’s consider schematically how each might perform across scenario outcomes:
What are the justifications for the expected performances across the scenario outcomes presented in this table?
In the case of a “Goldilocks Outcome”, the “Goldilocks Portfolio” should make money, but since this already seems to be priced into markets to some extent and risky asset valuations are stretched already, these gains are only likely to be modest.
In the case of a “Goldilocks Outcome”, the “Deflation Portfolio” should make money, as this scenario envisages yields declining and most likely trigger an “everything rally”. Gains might be smaller due to this portfolio not including as many risky assets.
In the case of a “Deflation Outcome”, the “Goldilocks Portfolio” might outright implode, because risky assets in particular are not pricing this at all, and likely have very substantial downside due to how stretched valuations and fundamentals have become.
In the case of a “Deflation Outcome”, the “Deflation Portfolio” should perform strongly, particularly as yields have risen over the past few years, and there is significant potential to benefit from cheap Convexity via both fixed income and elsewhere.
If we assume 50 / 50 odds of “Goldilocks” vs “Deflation” it seems fairly clear which portfolio is the dominant strategy. Even if it is true that “Goldilocks” is likelier than “Deflation” in 2024, risk-reward suggests the “Deflation Portfolio” may still be the dominant (ie: optimal) strategy. Furthermore, this is not a single game, but repeated. After every time interval (let’s assume quarter) during which “Goldilocks” conditions held, investors will have to play the same game again, but with valuations and positioning likely stretched even further (ie: a payoff structure that makes positioning for Goldilocks even more dangerous), and the likelihood of “Deflation” commencing in the next time interval having increased (assuming the effects of Fed rate hikes come through with a lag). Consequently, this game looks a lot like “picking up pennies in front of a steamroller”, and unless career incentives make it impossible not to do this (the “Deflation Portfolio” will tend to lag beta / benchmarks in the “Goldilocks Outcome”), it seems sensible to repeat the game with “Deflation Portfolio” positioning. Of course, if the facts change and the payoff matrix looks different, it will be necessary to be flexible and not stubbornly stick with the same stance that underperforms over a sustained period. But at this point in time, the “Deflation Portfolio” seems the most sensible approach from an absolute return perspective. For those with “career risk” relating to benchmarks, the payoff matrix looks somewhat different. This helps explain elevated risky asset valuations, and suggests caution is needed when attempting to interpret “what markets are pricing in”.
The title of this memo is not a claim that the “Goldilocks Outcome” will not occur, or that a “Goldilocks Portfolio” will suffer losses. Instead, “Goldiloss” refers to the top right quadrant of the “Absolute Return” matrix, and how this informs my Tactical Asset Allocation reasoning. Side-stepping the risk of the losses indicated in this quadrant, rather than “predicting” the future or “betting” on any particular scenario, is the basis for this current tactical stance. That is emphatically not to claim there is no risk of losses from the “Deflation Portfolio”. It would tend to perform poorly in the “Reflation” scenario, and might be mixed at best in a “Stagflation” scenario. There is an implicit assumption in this analysis that inflation will not re-accelerate, which is a view many might reasonably take issue with. However, as discussed above, both these regimes appear unsustainable for long, and could offer good potential entry points to increase positioning for “Deflation” if they occur in the short-run (contingent on being open-minded that the facts might have changed, and our stance might need to do likewise).
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.