These "Brexit" Market Moves Were Bigger Than Anything Seen In 2008... And What Comes Next

When the general population is asked to define a moment of paradigmatic instability in the history of financial markets, inevitably 2008 - in which there was a unprecedented divergence between risk perception

and the ultimate reality which saw Lehman fail and lead to the near collapse of the financial system - is the most cited answer. However, on Friday various markets saw volatile moves that put 2008 in the dust: in fact, the historic collapse in GBPUSD was not only far greater than any such move seen in history, but was an unprecedented 12-sigma move.

As Bank of America notes, post Britain’s vote to Leave the EU, the Euro STOXX 50 experienced its largest ever 1-day loss, GBPUSD reached 30yr lows and EURJPY, EURUSD & 10yr Bunds experienced 1d moves that were more significant than on any day in 2008. Gaps in risk perception that were evident even as recently as last week (VSTOXX-VIX spread was as wide as 20pts), may narrow further as spillover risk to global assets remains high. Indeed the Critical Stress Indicator of BofAML’s GFSITM index triggered on 13-Jun, suggesting cross asset stresses had risen by enough to lead to widespread contagion, absent policy intervention.

More details:

While derivatives markets like Sterling options clearly anticipated event risk around the referendum as early as April, there were record gaps in risk perception across regions and asset classes even days before the vote. Bank of America believes that gaps will further narrow as markets outside Europe realize the risks Brexit creates for global markets. For instance, at the time of writing the VSTOXX (V2X)-VIX spread, which traded as wide as 20 pts in the last week, decreased to ~13pts (EU close) as US equity volatility catches up with that in Europe.

But back to the resultant moves, some of which had no precedented in history, indeed some markets were more shocking than 2008, something which leads BofA to concludes that substantial "spillover risks remain." As the chart below shows, after the Brexit vote, the Euro STOXX 50 (SX5E) experienced its largest 1-day move ever today.

 

To put this and other monstrous asset class moves in context, Chart 2 normalises each asset class’ daily move by the standard deviation of the previous 100d (daily) returns, to reveal the X-Sigma nature of any given move. The findings:

  • Sterling had a 12 sigma move (vs history) on 24-Jun: GBPUSD, risk perception on which has been among the most stretched metrics across the GFSI leading up to EU referendum, reacted the most across the assets in Chart 2. This highlights just how important it is to pay attention to metrics close to the source of stress.
  • GBPUSD, EURJPY, EURUSD & 10yr Bunds were more shocked than even ‘08: Price action suggests FX & bond markets were more shocked by Brexit than they were at any time in 2008.
  • Scope of further contagion remains high: The GFSI’s Critical Stress Signal triggered on 13-Jun, as FX & equity stresses rose sharply. The markets’ reaction to Brexit has also witnessed a narrowing of some gaps in risk perception like the V2XVIX spread; indeed the VIX was >2x more surprised than the V2X today. Given the highly significant moves in cross asset stress and the relative underperformance of V2X (see later) versus EU equities, this metric will be a key barometer for monitoring contagion.

So unless central banks step in, how bad can it get? Some BofA forecasts:

The V2X lagged in the largest 1-d SX5E drop; may rise to ~46 if sell-off persists

 

The V2X underperformed the (~8.5%) fall in the SX5E today – rising by ~3.5 pts, while a historical relationship would suggest an increase of ~10 pts (Chart 2). Our equity strategists called for a 16% fall in the STOXX 600 in the short term vs yesterday’s levels, which translates to a further 10% drop in the SX5E from 24-Jun. Based on a historical relationship of SX5E 1wk returns & V2X 1wk changes, a 10% drop in the SX5E implies a rise of ~11 vol pts in the V2X (on average), which would take it near its 2011 high of 46 (Chart 3). Chart 4 extends the analysis from Chart 3 & highlights potential reaction of V2X in the coming month under various SX5E return scenarios. For instance, if the SX5E was to fall further by 16% in the next month, the VSTOXX is likely to be ~ 55.

 

 

 

How to read the projections in Chart 4: “Twk” stands for “time in weeks” from now to the end of a given projection. For instance, the green projection line in 2wks corresponds to an SX5E return of -8.5% [=6% * sqrt(2)] and predicts a V2X level of ~45.

 

The next question: what happens after such a spike, absent further exogenous shocks?

BofA uses an exponential decay process that is ‘fitted’ to each of the V2X, VIX and VNKY according to historical vol spike (see source of Chart 6 for details). Importantly, the decay process is independent of the starting level of vol, so it may be used to track the normalisation of a range of vol shocks: from the largest spikes in 2008 to the midsized ones in 2010-11 as well as the most recent (smaller) spikes (Chart 6, Chart 7, Chart 8). Note that in recent years, vol spikes have faded faster than in previous periods given the strong hand of central bank intervention (particularly in the US).

Finally, where could the VIX go?

The VIX outperformed the SPX decline; however, it remains low vs the V2X US equity vol, which lagged the broader rise in stress in the lead up to the EU referendum, has outperformed today’s decline in the SPX (as of the EU close). This has led to a narrowing of the gap between US vs EU equity vol. The trend is likely to continue if the sell-off continues.

In short: unless some exogenous entity, read central banks, openly intervene to restore calm and normalcy in the market, traders may be set for a long, painful period of volatility that has now been repricted sharply higher (something the banks will be delighted to use as an excuse to explain another quarter of subpar revenues) and will remain there absent some actions, not words, from the Yellen-Draghi-Carney trio.