Quants Corner

We have seen a very busy week in terms of data and news flows last week and there is one word which is now been increasingly used by different institutions from different parts of the globe: “Uncertainty”.

Here is a quick look:

The OECD cut its global growth forecasts by 0.2 percentage for 2019 to 3.3 percent and by 0.1 to 3.4 percent for 2020. The report stated that high policy uncertainty, ongoing trade tensions, and a further erosion of business and consumer confidence were all contributing to the slowdown.

The ECB growth forecast for 2019 was lowered down from 1.7 percent in December to 1.1 percent and growth forecast for 2020 was lowered down from 1.7 percent in December to 1.5 percent. Draghi commenting on these downward adjustments to Eurozone growth said: “The fact that the climate has become more uncertain doesn’t mean that one has to stay put. In a dark room you move with tiny steps.”

On Friday, US jobs growth shocked market participants. 20.000 new jobs created against the market expectation of 180.000. FED Chair Jerome Powell said on an interview on Sunday in CBS : “What’s happened in the last 90 or so days is that we’ve seen increasing evidence of the global economy slowing down” and “The principal risks to our economy now seem to be coming from slower growth in China and Europe and also risk events such as Brexit.”

In a world of increased uncertainty and qualitative information flow, quantitative analysts tend to have harder time in explaining trends and price movements. This week, we analyse how this increase in economic uncertainty effects fund flows through a multi-asset perspective with the help of the indices generated by Baker, Bloom and Davis, 2016 (BBD).

Fund flows between different asset classes tend to deviate due to differing interest of market participants. The amount of this deviation tends to increase in times of asset allocation shifts. When different asset classes have similar sized inflows or outflows, this would mean the overall asset allocation of global investors is not changing. An increase in this deviation on the other hand would mean there is an increased interest to some asset-classes compared to other asset classes.

Using monthly EPFR data, we calculate cross sectional deviation in fund flows (as % of the total AuM) using flows to 7 different equity regions and 10 different fixed income asset classes. Equity regions are Asia ex-Japan, Europe ex-UK, Japan, Latin America Pacific, UK and USA. Fixed Income assets classes are Global, Emerging market, Western Europe, High yield, US Municipal and US treasury long-short-intermediate bonds , floating rate notes and cash. To quantify uncertainty, we use economic policy uncertainty (EPU) indices. These indices are generated by counting the frequency of words (uncertainty or uncertain) in the newspapers and have been cited heavily in the literature.

Global economic policy uncertainty (EPU) is at its highest point since 1997 and global investors started to integrate this information to their pricing kernel. By the end of January 2019, the 2-year rolling correlation between the two is at its highest point since 2015 and is around 50%. (Figure 2). In comparison to longer-term - which is around minus %15 – this is a historically significant correlation and signals a structural change in investor behaviour towards uncertainty. Moreover, comparing these correlations with regional indices, China related uncertainty index have seen the most increase in correlation with the cross-sectional deviation across our 17 asset class universe. Using weekly data of EPFR, we also share more recent developments (Figure 4). The cross-sectional deviation continue to stay at elevated levels by the week ending on 06 March 2019.

For more insight subscribe here: https://bit.ly/2PlsLcl

Figure 1: Uncertainty started to effect asset allocation shifts

Figure 2: Correlation btw Uncertainty Index vs deviation in flows is at its highest

Figure 3: Uncertainty on outlook for China is affecting cross-sectional fund flows

Figure 4: Weekly Multi-Asset Cross-Sectional Flow(%) Deviation stays elevated

Reference
Baker, Scott R., Nicholas Bloom, and Steven J. Davis. "Measuring economic policy uncertainty." The Quarterly Journal of Economics 131.4 (2016): 1593-1636.