In Value investors: is this their biggest opportunity since the Tech Bubble?, we looked at how spreads have changed for stocks in the MSCI World, US and Europe indices and in different global macro sectors. We found that at the end of 2020, spreads across the board had risen to levels last seen at the peak of the tech bubble.
With such a large dispersion, the risk of spread compression is now high. What can we expect in periods of spread compression for common factor styles and their multi-factor combinations?
Which factors do well in periods of value spread compression?
In table 1-A, we show the performance of the value, quality, momentum and low risk factor styles for stocks in the MSCI World index based on similar robust macro-sector neutral and beta-neutral long-short factor strategies.
We have included an equal risk contribution multi-factor composite based on the aggregation of the four styles. For illustration purposes, we have added a low-size factor, also based on a macro-sector neutral and beta-neutral strategy.
In table 1-B, we repeated the exercise while relying on more traditional approaches to construct the factor portfolios.
- Here we used the HML (high minus low) value factor strategy that relies on the book-equity-to-price ratio.
- We included a proxy of the quality factor, RMW (robust minus weak), relying just on return on capital employed.
- We included the momentum factor, UMD (up minus down), based on 12 month returns minus one month returns.
- The raw low volatility factor is based on three-year historical volatility. Raw factors are neither sector-neutral nor beta-neutral. Their performance is more representative of what many equity investors tend to have in mind when discussing equity factors.
Tables 1-A and 1-B: Information ratios of equity factor styles conditional to value spread regimes for stocks in the MSCI World universe; based on monthly net returns in USD; 31 July 1995 – 30 Nov 2020; no transaction costs
| Table 1-A Factors | period | expansion | compression |
| Value | 0.8 | 0.1 | 2.0 |
| Quality | 1.4 | 1.4 | 1.3 |
| Low Vol | 0.6 | 0.6 | 0.5 |
| Momentum | 0.6 | 0.8 | 0.3 |
| Multi-factor Composite | 1.3 | 1.1 | 1.6 |
| Low Size | 0.0 | -0.5 | 0.8 |
| MSCI World | 0.3 | 0.2 | 0.6 |
| Sample size (months) | 305 | 134 | 171 |
| Table 1-B | Information ratio | Information ratio | Information ratio |
| Full | Value Spread | Value Spread | |
| Factors | period | Expansion | Compression |
| HML Value Raw | 0.1 | -0.4 | 1.3 |
| RMW Quality Raw | 0.6 | 0.7 | 0.4 |
| Low Vol Raw | 0.2 | 0.3 | 0.0 |
| UMD Momentum Raw | 0.3 | 0.5 | 0.0 |
| Multi-factor Composite Raw | 0.7 | 0.3 | 1.2 |
| Low Size Raw | 0.1 | -0.5 | 0.9 |
| MSCI World | 0.3 | 0.2 | 0.6 |
| Sample size (months) | 305 | 134 | 171 |
Factors in Table 1-A are macro-sector and beta-neutral and based on a target volatility approach with each style based on a diversified set of factors. Factors in Table 1-B are neither sector nor beta-neutral, and each style depends on one single raw factor. Source: Bloomberg, FactSet, Worldscope, IBES, Exshare-ICE, BNP Paribas Asset Management. For illustration purposes only.
For the more robust sector and beta neutral factor styles, the performance of value has been significantly better in periods of value compression and less good in periods of value spread expansion.
The same has been the case for smaller capitalisation stocks, perhaps even more so: The size factor only seems to be positive in periods of value spread compression. Quality, momentum and low volatility, appear to be less sensitive to changes in value spreads.
For the simpler raw factor styles, the differences in factor performance have been more pronounced as a function of the regime of changes in value spread: Value and size factors have outperformed only in periods of value spread compression and clearly underperformed in periods of value spread expansion.
In turn, quality, momentum and low volatility factor styles have done well when value spreads expanded and less well when value spreads compressed.
Multi-factor portfolios tend to do well in periods of spread compression
Multi-factor strategies tend to do particularly well in periods of value spread compression. They can also do well in periods of value spread expansion, in particular when based on the more robust sector and beta-neutral factor styles, but still do less well than in spread compression periods.
It is interesting to notice that investing in the best single raw factor with a perfect foresight timing strategy based on the value spread would have barely outperformed the more diversified multi-factor composites.
Takeaways
The recent poor performance of value stocks, smaller-cap stocks and multi-factor strategies can be understood in light of the expansion of value spreads.
With value spreads now at 20-year plus highs and with a compression of spreads on the cards, we believe that the coming years are likely to be particularly favourable for value stocks, small caps and multi-factor strategies.
This article originally appeared on Investors' Corner.
