Is Simply Achieving EBBO Good Enough?
Short Article(2 pages)
In the UK most trading conducted by retail investors happens on-exchange but 'off- book'. Retail investors, most of whom trade online using broker websites,
will send a Request for Quote (RFQ) to their broker who will use an RSP Network to elicit two-way quotes from up to 40 market makers. These quotes are good for at least 15 seconds so the broker can present the retail investor with the best
price available and give them the option to accept that price and conclude the deal.
Best execution for the UK retail customer therefore relies on two factors. Firstly, by convention,
market makers provide two-way quotes at or inside the best bid/offer and usually with volumes greater than available on the London Stock Exchange.
Secondly, the competitive nature of the RSP
network means that the price offered may not simply match the LSE BBO price, but may actually improve on it.
This is the theory; how about the evidence?
Figures 1 & 2, show spread capture histograms for different sets of retail trades executed in UK stocks during May 2013.
They are created by taking each retail trade and comparing it's execution price with the volume weighted market wide EVBBO price,
i.e. the best price a Smart Order Router (SOR) could aggressively execute that order size, using lit liquidity on the LSE or any lit MTF.
A value of 0 for spread capture means that the price
achieved was the same as aggressive EVBBO, a value of 0.5 means the price achieved was EVBBO mid price, i.e. capturing 50% of the EVBBO spread.
Fig 1, shows spread capture frequencies for around 100,000 retail trades executed on UK FTSE-350 stocks. Fig 2, shows around 50,000 retail trades
executed in small/mid-cap stocks that were reported to the ISDX exchange.
What's striking is that in both cases, execution prices are mainly to the right of the 0 spread capture mark, i.e. inside the EVBBO spread. Averaging over all flow, FTSE-350 stocks are executed about 15% inside the spread, the small/mid cap
ISDX stocks about 25% inside the spread.
Fig 1: Spread capture for FTSE 350 stocks
Fig 2: Spread capture for ISDX reported small / mid cap stocks
This means that a UK retail investor will get price improvement of at least 15% of the EVBBO spread on their aggressive 'deal now' orders (also with a
free 15 second option to trade).
What does this mean from a best execution standpoint? For many people,
best execution for retail size trades often simply means that the trade should execute within lit primary market BBO prices.
What's clear from our results, however, is that matching BBO or even EBBO isn't 'good enough' - there are liquidity providers willing to quote well
inside lit UK BBO prices for retail investors. So, any UK broker offering to simply match EBBO,
with no potential for improvement, would not be achieving the best price available for their customers.
How about other European markets? Countries in Europe have different ways of routing/ internalising/executing retail flow. We'll give two examples, one
for France and one for Germany.
Fig 3, shows spread captures (versus primary market BBO this time) of French retail flow executed on Equiduct's PartnerEx exchange.
Although the flow is executed at exactly EVBBO on PartnerEx, there are still some clear improvements to be had versus sending the same orders
aggressively to the primary market.
Fig 4, shows spread capture of German trades executed on a regional exchange - Stuttgart - versus BBO prices on Xetra.
Again there is clear price improvement with, in this case, a relatively large number of trades matching at Xetra
Fig 3: Spread capture for French equiduct trades
Figs 4: Spread capture for Stuttgart trades
So the pattern is that, using quite different microstructures, aggressive retail flow is typically achieving significant price improvement over prices
that would be obtained by routing aggressively to a primary market. For retail best execution, BBO (or in the UK, EBBO), is not 'enough'.
Finally, we might ask why should these retail focussed trading pools offer apparently better prices than those offered on lit markets to institutions/
professional traders? The most likely answer is that in each of these venues, the counterparties to the retail flow can be fairly confident they are
dealing with low frequency, 'less toxic' flow. As such, the risks of being picked off by HFT firms or suffering the adverse selection risk of trading
the other side of a large, price-moving institutional block are reduced. Hence the liquidity providers can afford to be more generous.
There are interesting parallels here with the role of buyside Dark Pools / BCNs in the institutional space where, by being selective about who can
participate, these venues can sometimes provide better outcomes to those who do participate.