|
In Strategy Matters: Momentum, rotation and value traps (October 15, 2015) we discussed the prospects for changes in sector leadership in the market. At that time, there had been a rapid and powerful rotation within several equity markets, resulting in a reversal of the trends that had been dominant throughout most of the year; in particular, EM and commodity-exposed stocks bounced strongly. This reversal of the previous performance led to much pain for many investors and raised questions about the sustainability of these moves. Our main messages were that: 1. While the momentum reversal in late September and early October had been sharp, it was not unprecedented. Since 2004, there have been eight previous momentum rotations in the market that were similar in magnitude to the one that started at the end of September. Many were short lived and we expected this one to be too. 2. We looked at value traps — long-run secular underperformers — and found that many enjoyed strong bounces in relative performance while still in a secular downtrend. Most such sectors historically experienced some bottoming of fundamental return drivers (such as ROE) before they started to recover in any sustainable way. 3. We argued that the rally in secular underperformers (commodities and capex plays in particular) at the start of October was more due to positioning than fundamentals. We expected the rotation to be short-lived and for the previous pattern of outperformance by more domestic, consumer-facing stocks to resume before long with renewed weakness in commodity- and industrial-capex-related sectors. Much has changed in the past month and, based on our measures of momentum, the leadership within the European equity market has once again reversed. The September/October market reversal has reversed strongly
EU Momentum Long/Short (GSRPEMEP Index), Jan-15 = 100
Source: Bloomberg, Goldman Sachs Global Investment Research. One important driver of the September/October rotation was the delayed expectations of a rise in US interest rates following the September 'dovish' Fed meeting. Coinciding with a bounce in commodity markets, this helped to trigger a rally in many EM FX cross rates as well as equities. But, with the more recent renewed confidence in an imminent rate 'lift off' in the US, coupled with yet further weakness in commodity markets, the previously weaker sectors have once again suffered poor performance. Previously weaker sectors have once again suffered poor performance
Market-relative price performance of European sectors, year to Sept. 28 and Oct. 9 to present
Source: Datastream, Goldman Sachs Global Investment Research. To examine rotations, we used the momentum basket GSRPEMEP (European Momentum Long/Short, price return/price return). This index is constructed using European stocks (approx. the largest 80% by market cap in Europe). The prior 11 months of price performance, one month ago, is standardised across regions/sectors and adjusted for volatility to provide a z-score (momentum factor) that is comparable across all stocks. The target portfolio goes long the top 20% of z-scores for the long leg, and the bottom 20% for the short leg. To this target portfolio, tradability constraints are applied (liquidity, turnover, sector/region weight) to generate the long and short baskets (GSRPEMEL/GSRPEMES=GSRPEMEP). It is rebalanced monthly. Looking at this index of momentum over a longer period reveals that the experiences at the start of October were by no means unique. We have found at least eight previous similar episodes (see below). The most recent downward shift in momentum was not extraordinary…
Left: EU Momentum Long/Short (GSRPEMEP Index), Log scale and Sep-04 = 100; Right: Examining the shifts in momentum
Source: Bloomberg, Goldman Sachs Global Investment Research. We found that the average had generated a shift of -15% in the momentum index. Of course, they were not all equivalent in length, thus when we considered the annualised shift in performance, the average was 84%. The reversal of the reversal! The reversals from these historical rotations has varied quite a bit. The speed depends to some degree on precisely which low we take as a measurement point. The table below is based on the low dates in the table above, but often a sharp reversal of momentum is followed by a sideways move for a while before a change in momentum asserts itself, meaning historical reversals of momentum may not have happened as quickly as the one we have just experienced. Nevertheless, the general pattern we have seen recently, with a reversal of the momentum unwind, is not unprecedented in terms of direction. That is to say, most momentum unwinds, such as the one that we experienced in October, do not last long. …as is the recent reversal of the reversal
Left: EU Momentum Long/Short (GSRPEMEP Index), Log scale and Sep-04 = 100; Right: Examining the shifts in momentum
Source: Bloomberg, Goldman Sachs Global Investment Research. Animportant feature of the market that is worthy of note is that, while previous leadership in the market has reinforced itself, the correlation of stock performance across the market has come down. Sector leadership has largely reverted to the pattern that prevailed before October, but stock performance has become less correlated. This is partly a reflection of the earnings season and more idiosyncratic news flow, but it is also probably a function of a less clear market direction. The strong momentum trend has re-asserted itself, but stock correlations are now falling
STOXX Europe 600 1m and 12m pairwise correlations
Source: Goldman Sachs Global Investment Research. Where do we go from here? In our view,many of the themes that we have looked at this year remain very much relevant. The domestic economies in Europe and the US are holding up well, and consumption is being supported by resilient labour markets. At the same time, the weakness in EM industrial end demand, coupled with low commodity prices and decreasing commodity capex, are serving as important and continued headwinds for many industrial stocks. In general, we think EM markets are likely to remain weak for two key reasons: 1. While US interest rate rises are now expected in December, the forwards imply a slower rise than we think is likely and the market-implied gap between our baseline projections and forwards grows over time from around 40bp at end-2016 to around 80-90bp at end-2017. Our US economics team’s modal forecast calls for 100bp of cumulative hikes during 2016 (Fed Funds would end the year at 1.4%) and a further 100bp of tightening during 2017 (at the end of which Fed Funds would stand at 2.4%). One of main reasons why we expect a higher trajectory for policy rates is that, with the US economy expected to continue to expand faster than its 1.75% potential rate of growth, pressures on wages and core inflation will build, but the market does not appear to be pricing this. Indeed, core US CPI inflation is already trailing just below 2%. Market prices imply that US policy rates will, at best, hover close to zero in real terms over the coming three years. If these expectations change, it may put more downward pressure on EM economies and markets. 2. EM imbalances have started to improve in many cases, owing to weak domestic demand and significant exchange rate adjustment. However, EM assets tend to bottom in the year before the end of the deleveraging cycle, suggesting that we could continue to see weakness at least well into 2016. Our economists note that credit gaps have already begun to shrink in Korea, Thailand and the CE-3 but gaps remain wide and bank earnings are still outsized in China, Brazil, Chile, Malaysia and Indonesia (for details, see: Emerging Markets Analyst: 15/18 - The EM Credit Cycle Part 2: Varying paths of deleveraging, November 13, 2015). Furthermore, we see more downward pressure on commodity markets, particularly in metals (for details, see: Metal Detector: Copper poised to move even lower November 11, 2015).
For these reasons,we continue to prefer more domestic-demand-exposed sectors and services, and remain negative on the capex cycle and areas exposed to this, particularly in EM.
|