The BIS Triennial Central Bank Survey - National Bureau of ...

1 Restricted The BIS Triennial Central bank SurveyAndreas Schrimpf NBER SI 2016 IFM Data Sources Project July, 2016 Disclaimer:Any views presented here are those of the author and do not represent necessarily those of the BISR estricted 2 Introduction FX and interest rate derivatives are some of the world s largest and most active financial markets .. OTC market with bank dealersas liquidity providers Trading is fragmented and relatively opaque Crucial markets, yet difficult to get data! In April 2016, BIS coordinated the Triennial Central bank Survey , known as the Triennial Survey conducted every threeyears since April 1986 11thedition provides the most comprehensive and internationally consistent information on size and structure of global OTC trading in FX and interest rate derivatives marketsRestricted 3 Introduction Focus is usually on the FX part of the Survey For instance, the 2013 Triennial showed that global trading volumein FX markets reached $ trillion (per day).

2 Global FX turnover: by instrument ( net-net ) USD bln Headline FX turnover figures widely reported and publicised .. But, underlying data remain largely unexplored by researchers .. Expected release of the 2016 figures: early-Sept this yearRestricted description What is the Survey about? How is the Survey conducted? of key features of the and weaknesses to the dataRestricted 5 Data descriptionFor more details, see King M. and C. Mallo (2010): A user s guide to the Triennial Central bank Survey of foreign exchange market activity , BIS Quarterly Review, December 2010 Restricted 6 Triennial Central bank Survey of FX market activityThe Triennial collects data exchange and interest rate derivatives turnovermeasured in notional amounts (April) amounts outstandingand gross market values of foreign exchange instruments and other OTC derivatives (end of June)

3 Turnoveris defined as the aggregate gross notional amount of all transactions struck in April (regardless of whether delivery or settlement was made during that month) Daily average turnover is computed by dividing aggregate monthly turnover by number of trading days in April for each country All figures reported in USD equivalents. Non-USD amounts converted into USD using the exchange rate prevailing on the date of the trade Focus in this presentation is on FX turnoverRestricted 7 How is the Survey conducted?In April this year, 52 Central banks surveyed 1,283 reporting dealersNY Fed, for instance, surveyed 24banks FRBNY reports aggregate results to BIS, who coordinates the Survey BIS does not see data for individual banksBIS adjusts the data for inter-dealer double counting For inter-dealer transactions, dealers specify whether the trade is local or cross-border 1)Local trades between dealers are halved ( Net-Gross basis)2)Cross-border trades between dealers are halved ( Net-Net basis)Gross-Gross (raw data) Net-Gross(location) Net-Net(global)

4 Restricted 8 What is reported in the Triennial ? FX partDealers report trades, broken down by instruments: swapsoOnly forward leg of the swap is forwardsoAlso includes non-deliverable forwards (NDFs) productsDealers report trades, broken down by currencies and currency pairs: -38currencies and 47 individual crossesRestricted 9 What is reported in the Triennial ? FX partDealers report trades, broken down by counterparties: DealersLarge banks and securities houses that participate in foreign exchange markets as market makers and have active business with end customers captures inter-dealer trading Other financial institutions-o/w Non-reporting banks-o/w Institutional investors-o/w Hedge funds and prop.

5 Trading firms-o/w Official sector financial institutions Non-financial corporates, governments and retail customersRestricted 10 What is reported in the Triennial ? IR partDealers report trades, broken down by rate and other productsDealers report trades, broken down by financial customersDealers report trades, broken down by 38currenciesRestricted 11 Additional features In 2007, Survey questions on trade execution methods were introduced (further refined 2013/2016) Electronic vs. Voice Direct (Bilateral) vs. Indirect (Brokered) Different types of platforms ( SBP, MBP, dark pools) In 2013 additional o/w items were introduced Retail-driventransactions ( retail aggregators) Prime-brokerage( hedge funds or PTFs trading in the PB s name with other wholesale market participants) Non-deliverable forwards (NDFs) In 2016, a complementary Survey question was introduced to gauge internalizationratiosRestricted 12 Overview of some key data featuresFor more details, see Rime D.

6 And A. Schrimpf (2013): The Anatomy of the global FX market through the lens of the 2013 Triennial Survey , BIS Quarterly Review, December 13 Dissecting the drivers of FX turnover growthGlobal FX market turnover, by counterparty and by instrument1Ta b l e 1 Turnover in 20132(USD bn)Absolute change from 20102(USD bn)Growth since 2010 (%)Contribution to FX market growth3(%)Global FX market5,3451,37335100By counterpartyReporting dealers2,0705263438 Other financial institutions2,8099144867 Non-financial customers465 66 12 5By instrumentSpot2,0465583841 Outright forwards6802054315FX swaps2,2284692734FX options337130639 Currency swaps54112611 Adjusted for local and cross border inter-dealer double-counting, ie net-net basis.

7 Daily averages in billions of US to the total increase of $1,373 billion from 2010 to : 2013 Triennial Central bank 14 Financial institutions (other than dealers) are the main source of FX turnover growth .. Tu r n o ve r w i t h other financial institutions (OFIs) rose by $ to $ trillion, a 48% increase Continues the trend of the past three surveysGlobal FX turnover by counterparty ( net-net )USD blnReportingdealers39%Other-financialins titutions53%Non-financialcustomers\9%FX turnover 2013: counterparty sharesRestricted 15 Top 10 currency distribution in April 2013FX market turnover for selected currencies ( net-net ) two currencies are involved in each transaction, the sum of the percentage shares of individual currencies totals 200% instead of 100%.

8 2 Turnover for yearsprior to 2013 may be underestimated owing to incomplete reporting of offshore trading in previous surveys. Methodological changes in the 2013 surveyensured morecomplete coverage of activity in emerging market and other : Triennial Central bank 16 Strengths/weaknesses, alternative sourcesand accessRestricted Strengths of the data17 Holistic global perspective and very comprehensive dataset accurate snapshot of the market at a given point in time Sheds light on darker niches in the market Bilateratal voice trades vs electronic trades on lit platforms Emerging market currencies Can gauge extent of on-shore (involving a local dealer ascounterparty on at least one side of the trade) vs off-shore tradingand currency internationalization.

9 Rich cross-sectionalinformation (currencies and trading centres) Information on execution methods based on actual trading recordsRestricted Weaknesses of the data18 Frequencyof the data - only conducted every three years Also a few changes in reporting population, breakdowns etc. Time-series / panel data analysis not so easy Difficulty to disentangle structural vs cyclical factors Perspective by currency, not by country Not possible to construct flows between countries Difficult to get a sense of net long/short exposures Dealer-centricsurvey design + focus on wholesale segmentRestricted Alternative data sources19 Semi-annual surveys by regional FX Committees (London, NY, Tokyo etc.)

10 Research with microstructure focus: Data from inter-dealer platforms such as Reuters/EBS ( Chaboud et al 2014, Rime et al 2010) Trading data from dealers ( Menkhoff et al 2016) Additional data available by Swift (trade confirmation messages) and CLS bank (settlement volumes) Looking forward, data collected via trade repositories (TRs) may prove useful, especially for interest rate derivativesRestricted Access to the data20 Results of the 2016 Survey will be published here: The data can be accessed via the BIS Statistic Explorer You can also send an email to our statistics department for further inquiries: Summary21 Very comprehensive and rich dataset (main downside low frequency) Unique data to study evolution and structure of global OTC markets, expecially in foreign exchange Thus far, mostly used for research targeted towards policy publications Not much academic-style research (yet).