Via the excellent Damien Boey at Credit Suisse:
The “quant quake” is happening again.
Overnight, and indeed, over the past week, we have seen a very sharp rotation into value factors and out of momentum factors. This follows on from the sharp rotation we saw from late August to early September. To be sure, there was a brief intercession in the rotation from mid-September to early October. But now, on a sector neutral, equally-weighted basis, the value index has broken through its September peak, while the momentum index is close to re-testing its September lows.
For the most part, value factor performance has largely mirrored the evolution of the yield curve. As the yield curve has steepened (flattened), value factors have outperformed (underperformed). Momentum on the other hand has been heavily caught up with bond proxies, with performance varying inversely with the level of bond yields. Interestingly, overnight, we saw bonds rally and the yield curve flatten – but yet we also saw an extreme shift towards value and away from momentum. For readers keen on statistical arbitrage, we suspect this reflects a bit of “catch up”, given that value has undershot the level implied by the yield curve recently.
Indeed, “catch up” is the operative term when it comes to Australian value factor performance. Australian value portfolios have generally lagged behind their global counterparts recently. Indeed, if we did not have the WAAX stocks in our value short portfolio, Australian value performance would look very ordinary indeed. The international divergence is all the more stark when we consider that the Australian yield curve has steepened by more than the US curve over the past month, thanks to an RBA rate cut and dovish guidance. In part, the underperformance of Australian value reflects the undershooting we have seen in the domestic demand cycle and stock specific developments, resulting in a distinct lack of confidence in the earnings of cheap domestic exposures. But equally, our leading indicators suggest that things will ultimately get better, because financial conditions have eased. Undershooting today creates a lot more room for overshooting tomorrow.
In recent articles, we have outlined the case for value to continue bouncing. For further discussion, please see our recent articles “Tail risk or apocalypse” dated 5 September 2009, and “Uncorrelate me” dated 30 August 2019. Now that value has experienced a decent bounce, and is now testing technical barriers, we need to think about the broader question of whether we are merely seeing a temporary bounce, or the start of something much bigger. After all, the 2016-17 value bounce was very short-lived, even though it felt quite permanent in the heat of the moment. And recent history tells us that value only outperforms once every three to five years.
There are a few pertinent things to consider in addressing this question, which we have discussed previously:
1. How much can bond yields rise from here, and yield curves steepen? Higher yields, and steeper curves could continue to support value. And our risk premium models tell us that bond yields need to rise, while central banks are telling us that they still need to cut. Ironically, the liquidity problems the US repo markets are facing are keeping the Fed on the back foot, with a bias towards easing. Indeed, the Fed’s solution of permanent open market operations, is concentrated at the short-end of the curve. Bond valuations and policy biases favour curve steepening and value.
2. How much room is there for passive investing to underperform? A key reason why we have favoured value recently is that very flat or inverted yield curves are consistent with rising correlations between components of passive portfolios, and underperformance due to a loss of diversification. In this macro environment, it is not possible to be all things to all people. We are now being forced out the risk spectrum by easy central bank policy, such that we have to express a view about the world in our portfolio construction. Should bond yields rise, we would expect value equities to outperform, and quality equities to underperform, with the net balance of forces (cap-weighted) likely to result in passive portfolio underperformance. To hedge against this risk, the natural option is to take on more value exposure. Put differently, a real rush to the exits on passive investing should result in a material uplift to value factor performance. Interestingly, passive, and levered passive (risk parity) portfolios have not underperformed the past few months. We have not seen the unwind of passive yet, nor its effects on value investing.
3. What is the balance of fiscal and monetary policies in the next phase of easing? Fiscal policy in most economies represents money printing, and is directly inflationary. On the other hand, rate cuts are likely to become less effective in stimulating growth, either because of diminishing returns to the credit creation process (ie exhaustion of the pool of credit worthy borrowers), diminishing returns to asset price inflation (as wealth effects become less powerful) or because savers lose out more than borrowers benefit. If one believes that fiscal policy is the only way to go, and that productivity deterioration is destroying spare capacity, the next easing cycle could be more inflationary than previous cycles, and the very nature of defensive positioning will change. Bonds and related exposures will look less attractive than inflation hedges like commodities. And surprise surprise – value factors do not favour bond proxies.
4. Finally, how much more room is there for Australian value to catch up to the value in the rest of the world? Undershooting today creates room for overshooting tomorrow. Perhaps Australian value is experiencing more of a delayed reaction than a decoupling from the slope of the yield curve.
Having said all of this, it would be remiss of us to not mention the other side of the coin. What if higher bond yields cause such an unravelling in positioning in asset allocation land, that de-leveraging and sharply slower growth occur? What if fiscal policy makers choose to dig in, forcing monetary policy makers to do all the work, an in a rather inefficient way? What if the return of quantitative easing makes us revise down our estimates of neutral rates globally, supporting and justifying expensive bond valuations for longer (ie what if there is no bond bubble, even though the conventional indicators are saying there is one)? And looking beyond the next year, what if fiscal policy makers stimulate today, but forget to add stimulus tomorrow? This risk is material and relevant today when we consider how gridlocked many governments currently are, and that very low interest rates make a dollar tomorrow worth just as much as a dollar today.
Right now, all we can definitively say is that:
1. Multiple dispersion is still very wide. Even if one cares about both value and quality, a blended factor portfolio still favours many value names in the long basket, and many quality names in the short basket.
2. Notwithstanding longer term concerns about the impotence of monetary policy, monetary policy is proving remarkably effective today. The likely pick up in the global credit impulse buys more time for a fiscal response which is so urgently needed for reflation and sustained value factor outperformance.
3. Central bankers, and to a lesser extent, fiscal policy makers, deserve some credit for having lowered volatility the way they have.
4. There has rarely been such a large divergence between our strategic (medium-term) factor portfolio signals favouring defensive momentum, and our tactical factor portfolio signals favouring value and quality at the expense of momentum. Right now, the tactical signals are overwhelming the strategic signals.
Therefore, there is still roughly a 3-month window for the value bounce to turn into something with a longer life.
I still view it all as more head fake than quant quake.