BIS: How Did We Get to $6.6 Trillion?

BIS: How Did We Get to $6.6 Trillion?

The Bank for International Settlements’ Triennial Survey of FX Turnover sent the market’s activity benchmark to a new high, but what drove the growth and what does it mean for the industry? Colin Lambert finds out.

The first thing to note about the Bank for International Settlements’ $6.6 billion figure for average daily turnover in FX markets is the degree of scepticism with which it was met in industry circles. Growth from the 2016 survey, which was the first since 2001 to show a decline, was expected, but it was the scale of the rise that took many by surprise, coming in at a very impressive 30% (almost 33% without the impact of exchange rate movements).

So why was the market blindsided and what does $6.6 trillion actually represent? In answer to the first question, the head of FX sales at a bank in London is succinct, “We, as an industry, are obsessed by spot activity – it’s where most of us make our money.”

In isolation that would seem a satisfactory answer, however it does belie the fact that all available data on spot markets pointed to a decline in activity, and that contrasts sharply with the BIS’ 20.3% increase in the latest survey. Taking the semi-annual FX committee surveys from the UK, US and Singapore – the three biggest FX markets in the world – from April 2016 to April 2019 activity actually dropped by 4.4%. Likewise, spot data from EBS, CME (using CME futures as a proxy for spot), FastmatchFX (EuronextFX), HotspotFX (CboeFX) and Thomson Reuters (Refinitiv) indicates an 8.1% decline in activity – and that takes into account the growth of Fastmatch during that period. Even including FXSpotStream which, while it is multi-product was largely a spot business in this period, the level of activity declined by just over 3% – a far cry from a 20% increase.

If it was just a case of the platform seeing lower volumes it could be explained away by competition from new or existing players; the rise of aggregation and dark pools; and the development of bespoke liquidity pools for clients, but the semi-annual surveys are what really confuses the picture. “I’m not sure how [the BIS] arrived at that number,” says a senior executive at a trading platform. “We saw some growth in spot but it was in the low single digits, nothing like this.”

The head of e-FX trading at a bank has one idea as to where the “missing” volume went. “Not only have we seen a real increase in private liquidity rooms, but the dark pool concept has gone well over that time – it could easily account for 60 yards of volume that no-one really sees.”

In notional terms, spot volume according to the BIS went up by $335 billion from April 2016 to April 2019, so if the e-FX trading head is correct just about one-fifth of that could be down to the increasing popularity of dark mechanisms, most prominently BGC’s MidFX, but even that theory still leaves the small matter of $275 billion unaccounted for. At first glance one answer could be Switzerland, which as a centre recorded a $120 billion increase across all FX products, however the BIS notes in its report that the centre changed how it reported data and therefore the 2019 and 2016 are “not fully comparable”.

Another popular suggestion is in emerging markets, as several sources point out, a lot of this data is not captured by the FX committees in the major centres. Individual product data is unavailable, however collectively, the Asian centres (ex-Australia, Hong Kong, Tokyo, New Zealand and Singapore) saw an extra $87 billion in volume in all products, the bulk of which was China, which saw all FX activity increase by $63 billion between the surveys. Elsewhere, Russia saw a decline in activity, South Africa rose by $8 billion and Mexico was unchanged.

In terms of currency pair activity there was a noticeable increase in Asia currency pairs, led, somewhat surprisingly in a month in which the peg wasn’t under attack, by USD/HKD. Turnover in all products in Hong Kong dollars rose by $142 billion per day over the three-year period, far outstripping the gain in USD/CNY which was $77 billion per day. As a whole, turnover in Asian currencies rose by $379 billion per day from April 2016, roughly one quarter of the overall increase.

What muddies the picture of those promoting emerging market growth as the driver is that turnover in EUR rose by $518 billion per day – hardly an emerging market currency. Given the relative lack of volatility in EUR/USD especially in April 2019 it is hard to fathom what drove the rise outside of, perhaps, FX swaps activity as interest rate expectations either side of the Atlantic diverged. Another theory put forward by the head of FX sales at a bank in London is that this number reflects increased activity by sovereign wealth and reserve managers.

“We have seen a lot of interest from various central banks and sovereign wealth funds to diversify out of the dollar and the first stop is the euro,” the FX sales head reveals. “The numbers involved when it comes to reserves, especially in Asia and the Middle East, are staggering and if a portion of that is being switched out of dollars into euros it would have an impact on turnover.”

Again, while being a good observation, there is a lack of evidence to support that supposition in the BIS report, for while activity with the official sector did rise across the board by $30 billion a day (half of which was in spot), this alone does not explain the huge overall growth.


It could be argued then, that the growth in spot FX turnover captured by the BIS but few others, is the result of a lot of smaller incremental gains from numerous sectors, be it dark trading, private liquidity pools or emerging markets, and the same can largely be said for the sources of this increased flow – although again the broad spread of growth only serves to increase the confusion as to how few saw this increase coming.

In what is simply the latest strange reading from a surprising survey, spot activity between Reporting Dealers only fell by 2.3% or $14 billion per day. “I find that strange, because our volumes have shifted from the primary venues where we would normally see the other dealers to venues where we face more client flow,” observes the bank head of e-FX trading. “Without checking the actual numbers I would think our flow with our peers in the top 10 has continued to fall over the past three years, probably by double digit percentages.”

Similarly, there was a healthy increase of $42 billion per day from the Non-Financial Customer segment, which suggests increased hedging by corporates especially, in spite of the apparent lack of volatility.

“Our customers were not especially busy in April,” says the head of FX sales at a bank in London. “That’s not to say there wasn’t an increase over the three year period – if I remember correctly, April 2016 was a pretty quiet month. We have seen corporate clients want to hedge more this past year but I wouldn’t say it has driven growth in our business.”

With the net change in flow between Reporting Dealers and with Non-Financial Customers amounting to about 10% of the notional increase over the survey period it seems that, as has been the case for the past decade, the source of the growth is in the Other Financial Customers segment. In hard numbers that is indeed the case, with the segment seeing an increase of $306 billion per day over the survey period and it is within this data that the first real hint of what blindsided so many lies. As noted, the official sector increased turnover by $4 billion per day to April 2019, and elsewhere Institutional Investor activity rose by $18 billion. There was a healthy increase in hedge fund and prop trading firm turnover, which increased by $61 billion per day, but it is in two other areas that interest is piqued.

Firstly, there is one data point in the BIS survey that bears further inspection, one that has, sources familiar with the survey say, got the collators baffled. It is simply entitled “Other” and was responsible for $200 billion per day in 2019, approaching treble 2016’s $71 billion per day.

The best guesses in the market suggest this number represents the growth of retail aggregators in FX, however as one source points out, “Why does that not show up in the retail-driven data, which only grew $6 billion?” It could be a simple reporting issue in that reporting dealers see the aggregators distinct from retail traders and indeed several e-trading sources refer to retail aggregator flow as “sharp” to use the word of one, however it does seem strange that flow so obviously associated with the retail segment is not reported thus.

The other area that could explain the source of the unexpected growth in spot activity is from Non-Reporting Banks. This flow, $448 billion per day from $354 billion in April 2016 could be missed in the FX committee surveys, especially if they are, for the purposes of a local survey, a Reporting Dealer (although that would still mean the flow would be counted by the local survey).

Growth in activity with regional banks is not a new story to the FX market, as more flow has been concentrated in the hands of a relative few, the regional banks have gyrated towards one of these players at the expense of the primary dealing venues especially – hence why the rise did not show up in the platform data. One regional player estimates that around 75% of their flow is executed in a private aggregation pool and that number increases sharply if their specialist currency flow is taken out of the equation.

The BIS report also shows that the amount of spot transacted via a prime broker rose by $354 billion to $918 billion, almost half of all spot turnover. Again, with the UK and US FX committees breaking out prime brokerage data, it could have been expected that this figure would have been telegraphed – it was nothing but, however. Whereas, over the three-year period, UK spot volume via a prime broker rose by around $20 billion (a just over 6% rise to $333 billion per day), in the US it fell by 6.7% or $18 billion per day. The two major centres then, in which the vast majority of the prime brokers would report their volume, saw absolutely no growth, however the global survey reports an increase of $354 billion.

One answer to this could be prime-of-prime activity, although it is broadly thought that this would be captured in the prime brokerage number. Equally one or two sources point to that modern day juggernaut of FX markets, Japanese retail, as being behind the growth, although it needs to be pointed out that activity in Japan generally across the period fell by $23 billion per day and yen spot volumes fell $35 billion per day. If it is indeed flow being transacted via prime-of-primes, then this, allied to the growth in the Other segment, would suggest the BIS and local FX committees might want to consider adding another box to their survey.

The Growth Engine

The growth in spot then, remains something of a mystery that can most easily be explained away by noting how it comes from many sources, none of which really dominate. It could be argued, as one senior market source does, that the data needs to be taken with a pinch of salt.

“The data is collated and submitted by people who are not closely associated with the trading business,” the source argues. “That means you will inevitably get some strange readings.”

But while the growth in spot has confounded many, the same cannot be said for the biggest source of growth in the survey – indeed the biggest market in all of FX – swaps. Activity in this product rose in almost every currency and with all client segments except for Institutional Investors, turnover with whom dropped from $278 billion in 2016 to $212 billion in 2019.

Equally, it can be argued that the growth in FX swaps, at more than 33%, was signalled by the multi-dealer platforms. Probably the closest proxy for this activity is that reported by Thomson Reuters/Refinitv which breaks out non-spot volumes and over the three years from April 2016 saw activity rise from $259 billion to $341 billion – a 31% increase.

Similarly, Deutsche Borse’s 360T, a platform with a high percentage of FX swaps activity according to sources, saw activity climb from EUR 54.1 billion to EUR 79 billion. The latter number does include volume from 360T’s GTX which was acquired in 2018, but assuming spot activity in the region of EUR 10 billion, the growth in what is likely to be largely FX swaps activity would have risen by around 30% again.

The FX committee surveys also presaged an increase in FX swaps activity of around a third, with the UK survey rise some 29% from April 2016 and Singapore, another major swaps centre, saw activity jump by just over 27% (the US, not a major swaps centre, saw activity fall back by around 5% and is only just in front of Japan in terms of notional volume – the latter rising across the survey period by around 7%).

The BIS in its preliminary report, does not break out NDF volumes, however it notes that a “large” proportion of the 22% growth in outright forward activity was down to growth in NDFs, particularly in Brazil, Korea and India. This again could account for a reasonable part of the overall growth, as indeed could FX options, which saw activity rise over 15% from 2016.

Notwithstanding the element of mystery over how the industry missed the growth in spot, the FX market has a new benchmark for the next three years at $6.6 trillion – and perhaps the various theories and facts about what drove the growth since April 2016 miss the biggest paradox of all. As the head of e-FX trading puts it, April was a “dire” month for activity. “April was probably the quietest month of the year,” the trading head observes. “Imagine what the BIS number would have looked like had things been busier!”

Julie Ros
Written by

Julie Ros

Similar Posts

Leave a Reply

Your email address will not be published. Required fields are marked *