I tend to read a lot about systematic investing, but this is largely through the Fama-French literature and the other anomalies that come from it (momentum, for instance).
Today I thought I’d dig in to systematic macro. I enjoy few things more than a cup of coffee with a macro podcast on but I rarely hear people talk about it systematically, apart from CTAs or discussions these days about how having near term cash flows benefits value stocks during inflationary periods. It turns out that this isn’t all that different from the kinds of papers that Raffi, AQR and the like publish when exploring the sensitivity of systematic factors to say, inflation.
I went over “Macro Factor Investing with Style” in the 2022 Quantitative Special Issue of the Journal of Portfolio management by Swade et al. Here are a few neat things I learned from it. Note that their sample size consists of 2001-2021.
Firstly, they find that global equities (ACWI in this case) are negatively correlated to traditional systematic factors as measured by the difference between the factor indices (MSCI value, MSCI momentum, MSCI low vol and MSCI quality) and the ACWI itself. This isn’t quite macro but an interesting observation nonetheless. It’s been paying to have been in factor tilts as I write this in late 2022.
But on to the macro. This paper fccuses on a handful of macro factors; growth, defensive, inflation, FX, EM and Credit. More specifically:
Growth = MSCI ACWI total return
Defensive = 10 year UST rather than the defensive equity factor
Commodity = Bloomberg Commodity Index
EM = EM spread over MSCI World Index
Credit = Bloomberg US Agg (80%) + Bloomberg US Corporate HY (20%)
FX = MSCI EAFE currency index in USD (80%) + MSCI EM currency index in USD (20%).
The factors themselves seem to be a combination of macroeconmic understanding and attempts to improve model performance. The paper itself cites earlier work by Bass, Gladstone and Ang in 2017, suggesting that the R squared values of economic growth, real rates and inflation account for a large chunk of the dispersion in portfolio returns. The Credit, EM and and commodity factors were more so to increase model performance.
A few notable findings from their paper, though many more exist:
- The US tends to outperform the ACWI in positive growth environments, consistent with the experience of the last decade.
- Traditional systematic factors help mitigated equity market downturns.
- Interestingly, EM currencies and FX carry seem to do relatively well with increasing inflation (positive coefficients between 0.4 and 0.6 with inflation). As someone fixated on the roaring dollar over the past year, I wonder if this is just a sample size/regime change issue as opposed to a real relationship we can depend on.
- Around 70% of the excess returns of both HY Credit and EM Credit were explained by the Growth, Defensive and Inflation factors. Wow.
The rest of the paper goes in to trying to construct optimal portfolios based on a variety of macroeconomic environments (noninflationary growth, inflationary growth, deflationary and/or crisis) but perhaps I’ll dig in to this later as it’s getting late into the night.
Finally, I still don’t fully understand the difference between traditional systematic backtesting across macroeconomic environments, but maybe I’ll learn more as I continue reading this stuff.
DISCLAIMER: I am not a financial advisor. Nothing on this website is to be constituted as financial advice. All content here is solely for educational purposes. They solely represent the opinion of the author.
Bells and Tails 