18 juillet 2017

The ‘Sintra Deleveraging’

July 18, 2017, Jean-Jacques Ohana, CFA and Dr. Christian Witt (both YCAP Asset Management)

Global sovereign yields suddenly spiked on June 26, 2017, when rumors spread that central bankers from around the world were discussing monetary tightening during a conference held in Sintra, a small Portuguese town. Since then, US and German sovereign yields jumped by +16 bps and +33 bps (see Figure 1), respectively. What was so special about this episode was the wider market response. Akin to the infamous ‘Taper Tantrum’ (2013) or the bond rout of 2015, stocks tanked in sync with bond markets in the first place. Diversification seized to exist in an instant. Surging yields forced many investors to first cut leveraged bond positions and then stock exposure, in particular defensive stocks (Utilities, Real Estate, Consumer Staples, Health Care) sensitive to the level of interest rates. This is also why we call this event the ‘Sintra Deleveraging’. With this case study, we attempt to characterize the ongoing deleveraging process by comparing it to these earlier periods of turmoil.

We perform several data adjustment to best contrast different asset classes. Otherwise, the analysis risks comparing apples to oranges. In a first step, we gather daily price data going back to the start of 2012 on German and US sovereign 10-year bonds (and yields), European and US High Yield bonds, the European and US equity indices and a global commodity index. Next, we compute daily returns from the price data. The daily returns are then normalized by their full-period standard deviations. We further determine the daily cross-section average of the individual standardized returns as a measure of global aggregate asset movements. Finally, a normalized performance indicator is derived by cumulating the normalized returns over time.

The case study considers three episodes: the ‘Taper Tantrum’ (2013), the bond rout of 2015 as well as the ‘Sintra Deleveraging’ (2017). The observation window starts one month prior to the initial shock and ends six months after. The precise date of the initial shock is derived from the drawdowns of popular risk-parity funds given their vulnerability to this type of shock. Using this technique, the dates of the initial shock are May 22, 2013 (‘Taper Tantrum’), April 27, 2015 (bond rout) and June 26, 2017 (‘Sintra Deleveraging’).

To get a general impression of how financial markets behaved over the course of the three episodes, we start by charting our normalized performance indicator based on the average cross-section return (see Figure 2). The indicator is rescaled to zero at the initial shock date. Two things catch our eyes. One, the average shock sustained on June 26, 2017 was stronger than its equivalents in 2013 and 2015, although it has fallen back in line with its two predecessors. Two, unlike 2013 and 2015, financial markets did not rally in the runup to the deleveraging shock. There are different ways to look at these findings. An optimistic assessment might conclude markets have started rolling over well in advance of the shock so that its repercussions will fade more quickly than in the past. A less favorable view, by contrast, holds that markets have been weak from the outset and are therefore particularly vulnerable. Ironically, the shock’s exceptionally strong initial but quickly fading magnitude lends credence to both interpretations.

Next, we perform an asset-by-asset examination to shed more light on the details. Let’s start with 10-year US and German sovereign bonds. Figures 3 and 4 illustrate that both the runup and dynamics of the 2017 deleveraging closely resemble the sell-off during the bond rout of 2015. This would imply bonds could come under infrequent but strong selling pressure. The contrast to the Taper Tantrum when bonds sold at a low pace over several months could barely be clearer. This raises the question of when the sell-off might run out of steam. Using both prior episodes as guidance, US as well as German sovereigns still have room for further declines. In 2013 (2015), the cumulative loss for the former reached 13 sigma (20 sigma) and 18 sigma (11 sigma) for the latter. The current sell-off amounts to an 8 sigma cumulative move for Bunds and a 5 sigma move for Treasuries. Thus, given the presumably unsteady rhythm of the bond rout and the still considerable potential for drawdowns, sovereign bond investors might be critical of any temporal tranquility.

These first impressions are further corroborated by the performance of global commodities (see Figure 5) which similarly track the trajectory of the bond rout of 2015. At the time, commodities initially rallied then reversed course and eventually plummeted. Before that, the Taper Tantrum saw commodities drop more gradually interrupted by some short-term rallies. Today, there is, however, a noteworthy difference to both examples. This time, commodities have already been in a slowdown before the rout even started. Although it might be tempting to conclude commodities are therefore up to a horrific correction, the varying initial market conditions render such a long shot pure speculation. Rather the key takeaway is that investors should vigilantly monitor commodity markets over the months to come.

Furthermore, neither equity markets nor high yield markets offer much insight into where they could be headed. While European stocks (see Figure 6) have rapidly resumed their bull market rallies in the past, US equities (see Figure 7) have done so only once. The other time US stocks were eventually flat while suffering considerable drawdowns in the meantime. In addition, neither stock market showed relevant momentum prior to the initial shock on June 26, 2017. The absence of a clear pattern also applies to European and US High Yield bonds (see Figures 8 and 9). During the Taper Tantrum, they roughly imitated equity markets, while in 2015 High Yield bonds seemed intimately intertwined with lackluster commodity prices. In any case, Equity and High Yield markets on both sides of the Atlantic failed to produce a clear pattern related to sovereign bonds.

To summarize our findings, the ‘Sintra Deleveraging’ has been characterized by considerable and concentrated losses across sovereign bonds. Using a normalized performance indicator, our case study illustrates that the current episode started off very similar to the Taper Tantrum (2013) or the bond rout of 2015. But at the epicenter of the deleveraging, i.e. the bond market, the current sell-off looks more like the unsteady bond rout of 2015. Given that both earlier events saw bond markets drop considerably further before hitting rock bottom, we worry about further bond weakness. It is also noteworthy that neither Equity nor High Yield and commodity markets offer much insight into where other financial markets could be headed given their disparate behavior during previous periods of bond calamity. For the time being, however, stock markets keep up well with Emerging Equities outperforming their developed market peers.

Figure 1: 10-Year US and German Sovereign Bond Yields

Figure 2: Normalized Performance Indicator: Average Cross-Section Returns. Initial Shock Date = 0.

Figure 3: Normalized Performance Indicator: German 10-Year Sovereign Bonds (BUNDS). Initial Shock Date = 0.

Figure 4: Normalized Performance Indicator: US 10-Year Sovereign Bonds (US Treasury). Initial Shock Date = 0.

Figure 5: Normalized Performance Indicator: Commodities. Initial Shock Date = 0.

Figure 6: Normalized Performance Indicator: European Equities (Euro Stoxx 50). Initial Shock Date = 0.

Figure 7: Normalized Performance Indicator: US Equities (S&P 500). Initial Shock Date = 0.

Figure 8: Normalized Performance Indicator: European High-Yield Bonds (HY EU). Initial Shock Date = 0.

Figure 9: Normalized Performance Indicator: US High-Yield Bonds (HY US). Initial Shock Date = 0.

Figure 10: Performance of Equity Markets Since June 26, 2017