NON-DISPLAYED TRADING AND THE PUZZLE OF THE SWISS.
By Robert Cranston, Head Equity Product Management, SIX Swiss Exchange.
We all love a puzzle, at least that is what the hundreds of random math questions popping up on LinkedIn have led me to believe. Leading up to the launch of SwissAtMid in October last year, we spent a lot of time looking at the volumes of Swiss stocks traded in non-displayed pools across Europe, and the patterns that emerged left us with another puzzle.
Formal MTF dark pools were launched, post MiFID I in Europe, towards the end of 2009. Since that time there has been a steady increase in non-displayed trading across Europe, up to around 8-9% of overall market flow trading in non-display pools (according to statistics gathered by Rosenblatt Securities). While there is a natural expectation that this will slow down, especially with the advent of MiFID II, there is an expectation that there will still be a significant amount of flow that will be traded in non-display venues moving forward.
What has come to light is a difference in trading patterns when we started to look at our market compared to some others across Europe. The market with by far the largest shift to non-display trading is London, where the market regularly trades over 12% in non-displayed MTFs. There is also a larger increase in non-displayed trading seen in French instruments where there are regular peaks in trading of over 10%. This compares to the Swiss market where we see around 7-8% of flow trading in dark pools (see Fig. 1). Both the growth in dark trading and peak so far is considerably smaller for Swiss instruments, the question is why?
To find answers, we need to look at why people trade in this way. The easiest answer is that it must be more profitable to trade in the dark for London and Paris stocks versus those from the other markets. What would drive this profitability? To understand this we need to look at the underlying structure of the different markets to try and assess the value of trading in the dark. In this article we look at two key indicators of book quality (pieces of the puzzle) – the average spread and the average volume available at the top of the order-book.
The first piece of the puzzle, the average spread, is a key factor when talking about mid-point trading. The wider the spread, the more value there is in a mid-point execution. However, if this was the only factor driving the decisions, then the Swiss market should expect a much larger percentage of dark trading to take place, as the value of trading at the mid for the main Swiss blue chip index is significantly greater than that of the associated indices from the other European venues. Taking half the spread of a Swiss instrument gives you over 3.6 bps when a similar trade in a CAC40 instrument would only yield 2.1 bps on average (see Fig. 2). So the puzzle is not solved by spread leeway.
The next piece of the puzzle is the queue sizes on the lit book. We do this by looking at the average bid and ask depth for these same indices. This number is very useful as it gives us a view of not only the volumes available at the touch, but also a very strong indication (when looking at highly liquid instruments) of the time it would take for price levels to change and move. The assumption here being that for a highly liquid instrument, a market with considerably more volume available at the touch will see less price moves, and a more stable market. Here we start to see again some very clear differences in book structure, with the Swiss market having much larger volumes at the touch than the other markets being compared (see Fig.3).
These two factors work together. A wide average spread, when driven by wider average tick sizes on the Swiss Exchange (due to using FESE table 2), builds additional liquidity at the touch. The puzzle appears to be taking shape.
However, it also has a secondary impact when it comes to non-display trading. The value of non-display trading to participants is not just price improvement but also the ability to hide their intentions from the market. With an order book that is very dense at the touch (in terms of volume) it is easier to be able to trade out of even relatively large positions without impacting the price on the order book. When there is less volume available, trying to trade without moving the market is much more difficult, and drives you to work harder to try and trade with reduced impact.
With the launch of SwissAtMid last year we are starting to see the difference in non-display volumes between markets narrow for dark trading. By making it much easier to achieve the considerable price improvement available in Swiss instruments, and housing this directly on the same platform as the largest source of Swiss equity liquidity, we are removing a lot of the risk of trying to trade in the dark today. This makes it easier for many firms to achieve the price improvement available at the mid-point with very little technical change. As you can see from the data (see Fig. 4), the growth is significant, and yet we have not seen significant impact to the volumes of Swiss non-display trading carried out by the other public dark pools in Europe. We fully expect the non-display volumes in SwissAtMid to continue to grow, but with the advent of MiFID II we won’t run out of new pieces to fit into the puzzle anytime soon.