You are on page 1of 9

EU Referendum market insights; what

was actually going on behind the


scenes?

June 2016
EU Referendum market insights; what was actually going on behind the scenes?

Page | 2
EU Referendum market insights; what was actually going on behind the scenes?

EU Referendum market insights; what


was actually going on behind the
scenes?
Research note covering activities in the financial markets leading up to the UK
Referendum on EU membership continuation.

Introduction
This research takes a deeper look at the events before and after the UK Referendum
on EU Membership held on Thursday 23 June 2016. It also provides a bit more context
on inter-market correlations and other discerning takeaways. The deep insights
provided by the AlgoDynamix Risk Analytics Platform (ALDX-RAP) were exceptionally
helpful in highlighting numerous market anomalies across EU, Japanese and North
American equity indices. Scenario testing on the potential consequences of a British
Exit (‘Brexit’) has been considered and some ‘implied’ Brexit possibility was perhaps
showing in the FX options market [1].

Little or no outcome certainty was available from the ‘traditional’’ sources including
numerous pollsters [2] and in fact even more uncertainties from the bookmakers based
on potentially very skewed betting sizes [3]. Indeed anecdotal evidence suggests that
even some of the more experienced market participants were wrongly positioned [4].

Deep Data Insights

The AlgoDynamix Risk Analytics Platform (ALDX-RAP) is based on ‘deep data’


algorithms that are continuously analysing the order book flow and are detecting non-
random behaviour of market participants. Ongoing analysis is used over time to cluster
the behaviour of these structured market participants. Excessive structure, especially
over longer time frames, results in the anomaly detections shown on the RAP web
interface, see Figure 1.

Page | 3
EU Referendum market insights; what was actually going on behind the scenes?

From Figure 1: recent anomalies alerts - within the first 24 hours - are shown in red.
Older anomaly events are shown in yellow. Note that email alerts are automatically
sent every time a new anomaly is detected. Emails are time stamped and digitally
signed; this is especially relevant for both ease of product use and ex-ante
performance validation. Note that the underlying analytics engines are non-predictive;
they are reporting in real time what is going on in the underlying data sets and as
such should be viewed as deep data insights and not some form of probabilistic
indicator. The API can be used to gain additional insights to help improve the decision
making process; this includes cluster sizes and more granular price points.

Page | 4
EU Referendum market insights; what was actually going on behind the scenes?

Results

The results for various instruments are shown in figures 2 to 5. The input data sets
were streamed in real time from the major Futures exchanges including CME and
Eurex.

Figure 2: EURO STOXX 50® Index Futures SEPTEMBER 2016, with anomalies shown as A0, A1, A2 and A3.

Anomalies on the EURO STOXX 50® were already detected on Tuesday 21 June 3pm
UK time, about 2 and half days before the actual crash, see figure 2. The frequency
and intensity of the alerts increased more and more as we approached the results of
the referendum. Note that anomaly alerts are usually provided hours or days in
advance of the potential event. Different instruments and asset classes can have very
different ‘typical’ behaviours. Just like most other financial time series parameters,
these underlying properties will vary over time (‘non stationary data sets’) so any
inference from historical parameters should be used with extreme caution, if at all.

Very similar results also apply to the DAX Futures contract, see Figure 3. This isn’t
very surprising considering the very tight correlation between these 2 equity
instruments [6].

Page | 5
EU Referendum market insights; what was actually going on behind the scenes?

Figure 3: DAX SEPTEMBER 2016, with anomalies shown as A1, A2, A3 and A4.

The results for the SP500 and NIKKEI 225 YEN INDEX are shown in figures 4 and 5
respectively. Note that the anomaly alerts appear somewhat later and less frequently
than the European indices.

Figure 4: NIKKEI 225 YEN INDEX SEPTEMBER 2016, with ‘Brexit’ related anomalies shown as A2 and A3.

Page | 6
EU Referendum market insights; what was actually going on behind the scenes?

In the case of the SP500 this is only detected a few hours before the actual crash.
This does suggest a ‘panic contagion’ between these global equity indices with varying
delays possibly due to different levels of correlations between these instruments.

Figure 5: S&P 500 INDEX FUTURE (Continuous)

Also of great interest is that more anomalies happened after the crash on the non-
European indices, see Figure 4 and Figure 5. This implies that lower equity prices on
these non-European equity instruments could be contributing to ongoing lower prices
across more global equity indices.

Conclusion

These above examples of ex-ante performance of the AlgoDynamix Risk Analytics


Platform are interesting from both a portfolio and exposure protection point of view,
as well as offering a better understanding of global inter equity contagion.

Page | 7
EU Referendum market insights; what was actually going on behind the scenes?

As also demonstrated on other occasions [7], the deep data insights provided by
AlgoDynamix can be used for numerous applications including improvements in:
trading, hedging, derivative overlay strategies and option protection strategies.

References
[1] Implied Distributions from GBPUSD Risk-Reversals and Implication for Brexit Scenarios
Iain Clark , Efficient Frontier Consulting Ltd
Saeed Amen , Thalesians Ltd
June 13, 2016
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2794888

[2] http://www.bloomberg.com/news/articles/2016-06-05/brexit-pollsters-feeling-pressure-
with-uncertainty-as-only-given

[3] http://www.zerohedge.com/news/2016-06-22/something-strange-emerges-when-looking-
behind-brexit-bookie-odds

[4] http://www.bloomberg.com/news/articles/2016-06-27/soros-was-long-the-pound-before-
brexit-vote-says-spokesman

[5] https://vimeo.com/163691471

[6] COPULA BASED VaR APPROACH FOR EUROPEAN STOCKS PORTFOLIO Mária
Bohdalová, Michal Greguš, Comenius University in Bratislava,
maria.bohdalova@fm.uniba.sk, michal.gregus@fm.uniba.sk

[7] Overview of Deep Data Analytics Applied to Portfolio Tail Risk Hedging
Jeremy Sosabowski, Wei Yin Teo.
Big Data for Trading and Risk Management Stream, QuanTech Conference, 22nd April 2016

Page | 8
EU Referendum market insights; what was actually going on behind the scenes?

Disclaimer
AlgoDynamix Ltd. (The Company) is a company registered in England and Wales with a company
number of 08853134 with a registered office at 5th Floor Salisbury House 744-750 Finsbury Circus
London EC2M 5QQ. The Company is not a Registered Investment Advisor, Broker/Dealer, Financial
Analyst, Financial Bank, Securities Broker or Financial Planner. The Information in this document is
provided for information purposes only. The Information is not intended to be and does not constitute
financial advice or any other advice, is general in nature and not specific to you. Before using the
Company’s information to make an investment decision, you should seek the advice of a qualified and
registered securities professional and undertake your own due diligence. None of the information in
this document is intended as investment advice, as an offer or solicitation of an offer to buy or sell, or
as a recommendation, endorsement, or sponsorship of any security, Company, or fund. The Company
is not responsible for any investment decision made by you. You are responsible for your own
investment research and investment decisions.

Page | 9

You might also like